Toyota FRS Electric Power Steering in an A Body

[jetmech442]

Objective:
Integrate a 2014 toyota FRS Electric Power Steering(EPS) system into my ’69 Cutlass, allowing the steering ECU to vary assist levels based on vehicle speed, mph, torque input and steering angle.
Motivation:
This car has always been more of a journey-not-destination thing for me. And as such I'm at the point where the Grand Jeep Cherokee steering box and stock LQ9 pump with Turn One reducer valve just isn't doing it for me anymore. I'd love to give Turn one or Lee a thousand dollars, but I don't have that kind of money, plus getting the known easy solution isn't my thing for some reason, I guess I just like the challenge of integration. I've got a build thread over on LS1tech https://ls1tech.com/forums/conversions-hybrids/1418981-lq9-t56-1969-old-442-conversion-2.html where I started, if anyone is interested.
I’ll state here that I am not a CANBUS expert, I’m merely a novice with an Arduino and CANshield that is pulling data from multiple sources and doing his best to put the pieces of the puzzle together. I love discussion and feedback and always am eager to learn from the awesome people here at protouring and my goal is to help others on their path.

So this project, as most of mine do, stands on the shoulders of giants. Below is where the majority of my knowledge came from-people freely sharing what they've learned for the love of cars. Hopefully someone someday will use this to do something even cooler(also kind of assuming that this project is successful in the end, lol). I'm a thermal engineer by day, which means I know lots about temperature and not lots about CANBUS or electronics. I dabble in Arduino, but most of what I do with it is based re-purposing other peoples code-guess I'm trying to say that I don't think this is out of anyone’s ability if they have the desire.
CAN equipment
I am using a Seedtudio v1.2. there is a new version out since I’ve had mine for a while, but I don’ think it does anything significantly different.
http://wiki.seeedstudio.com/CAN-BUS_Shield_V2.0/
This stacks on top of a basic Arduino Uno. And plugs into the car using the OBD-connector and DB9-connector that Seedstudio sells.
There are tons of ways though. I have run into a lot of people who been very successful with the SocketCan, which utilizes Linux and a direct port into a laptop. Since I plan to convert my current rpm/VSS into a CAN message and broadcast, the arduino seemed like the best path for me though.


Different ways to get assist from your unit:
There's several way to tackle this. Some just hook up power(no CANBUS) and have full assist at all times. some hook up what I assume is a PWM circuit and dial in what they want which is the same across all speed levels. My contribution will be(hopefully) adding in the CANBUS communication to the toyota steering ECU and letting the algorithms control assist based on VSS, rpm, steering torque and steering angle. So, as we all know, OEM's spend quite a bit of time tuning the EPS for steering feel. I assume that they spend more time on "feel" for the FRS than they do a Yaris or Prius, so I scooped up a 2014 FRS EPS with ECU. I plan to convert the RPM and VSS from the LS1b GM ECU into CANBUS signals and let it do its thing. That's the plan at least!!!! I'll post more as I make progress, hoping this won't take too long. Below is a quick pic of the EPS in the back seat.

159158

Knowledge sources
So to start with, Waid302 posts in vintage mustang what I consider the best starting place for understanding the project. Its a huge read, but worth it. He provided a ton of pics and diagrams which are enormously helpful-especially if you just want a Prius unit to give full assist.
https://forums.vintage-mustang.com/mod-custom-forum/1008722-electric-steering-fail-safe-no-ebay-module-needed-pics-videos.html

dobrostand has a thread in the same forum detailing specifics of the driveability of the system, which led me to believe this might just be worth it.
https://forums.vintage-mustang.com/mod-custom-forum/866473-rubber-hits-road-electric-power-steering-epas-drivability.html

Also on Vintage mustang- Andrewb70, originally known to me as project_GatToGo and who is a huge inspiration to my build, details the specifics of installing in his cougar, but also takes it a step further at getting his dominator to provide VSS to the prius ECU for variable assist. Also pretty much lays out all the details for fab/install that you could ever want,assuming you are skilled with a whizwheelofdeath, lol. Additionaly, he brings up the detail that the ECU needs to run a centering calibration if any of the parts have been mismatched, replaced or otherwise tampered with.
https://forums.vintage-mustang.com/mod-custom-forum/1147680-67-cougar-toyota-prius-epas-install.html


I believe those threads should be sufficient to get a Yaris/Prius or Saturn Vue EPS installed and supplying full assist. The nice thing about the Japanese EPS systems is that they supply full assist in default mode if no CAN signals are received. The Vue uses an aftermarket box that dials in "fixed" assist. Lots of people feel fine with that, I might also, but I like being a nerd so I'm looking into supplying the proper CANBus channels.

The next major source of info is some enginerds at the University of Tulsa has published some papers on CANBUS hacking. They provide the engine speed HexID (2C4), and a version of the vehicle speed(610). I'm not entirely sure ID 610 is what the Steering ECU is looking for based on some info I find down below, but we'll get to that. This paper is awesome at showing the methodology for grabbing canbus data and then deciphering it. These guys are the real deal.
http://tucrrc.utulsa.edu/ToyotaCAN.html

Then, Fabio published a youtube video of a camera synced with CANBus data in a 2010 Camry. This allows you to see the structure of the CAN message for RPM (2C4) and Vehicle speed(both “B4” and “610”).
https://www.youtube.com/watch?v=KHRPmjwXF1U


I would say this entire paper is a must-read and is also the most indepth of anything I've found so far. It talks about great details for ford and Toyota steering/braking Canbus interfaces. Specifically, page 41 shows how ID B4(vehicle speed) is not only structured in the way of what bits mean what, but also how to combine them, and the importance of the "checksum" bit at the end of the messages for toyota. If the different nodes don't see the checksum, they will ignore the message all together. I don't know if this is all OEM's or just Toyota.
http://illmatics.com/car_hacking.pdf

[jetmech442]

so the papers/videos above will give a pretty good background on all the details. My knowledge was zero before I started and is at where it is now based on those. Below are some real details for the toyota system I have worked through but not yet implemented. Please feel free to ask questions or comment on whatever. I'm hoping to wire it up soon and start testing it out...

Breaking down the Toyota RPM CAN message.
below is an example of the rpm message {2C4 8 0E 1B 00 19 00 00 92 A2}
2C4 is the message ID and this one happens to mean RPM.
8 is the message size, it will send 8 additional pieces of info.
0E is the first piece of info. In decimal it means 14 ---Protip-the stock calculator app in windows can be changed to "programmer" which allows you to input HEX numbers and it auto translates into decimal.
1B is the 2nd, and it translates into 27
Now those two bites alone don't help or add up to anything, but if you put the HEX together (0E1B) it translates into 3,611-which is the actual rpm!
I don't know what the "19" or "92" mean yet, but the A2 should be the checksum at the end.
To calculate the checksum, it helps to know that the message ID is actually two bits, IDH = 2C and IDL = 04. So to calculate it, add(in HEX calculator) 2C+04+08+0E+1B+19+92 = 1A2. This technique is found in http://illmatics.com/car_hacking.pdf page 9 and 10.

Breaking down the Toyota Vehicle Speed CAN message.
Currently, I am aware of two ID’s that should contain vehicle speed info 610 and B4. The paper from Tulsa shows 610 containing vehicle speed data. The paper from illmatics identifies B4 as the message used by the “combination meter”-that’s the instrument cluster for the rest of us. We don’t know which message the steering ECU is looking for, so I will need to send both(1 at a time), and see which one reduces steering assist as the speed increases.
So to break it down, on page 34 of illmatics, {00B4 8 00 00 00 00 8D 06 66 B5} is equal to 10 mph. we get this by combining 06 and 66 into 0666 which translates from hex to dec as 1,638. We then multiply that dec number by .0062.
1638 *.0062 = 10.15mph.
ID B4 details

The formula for this is found on pg 41 of illmatics, but I’ve rewritten it slightly here to make it easier to understand(IMO).
IDH: 00, IDL: B4, Len: 08, Data: 00 00 00 00 CN B6 B7 CS  
CN = Counter that iterates from 00-FF -I don’t know what the counter does currently, but it goes from 0-255(dec)
B6 = 6th byte of the message
B7 = 7th byte of the message
CS = Checksum
Speed = (B6B7) * .0062 == MPH

ID 610 details

The example here is taken from Fabio’s video @2:07 and @2:09 where you can see the dash is displaying the speed in Kph. It looks to me that the 3rd byte represents Mph. Two such examples seem to confirm this…
610 8 {20 00 10 E4 C0 00 00 00} = 25 kph -------Hex 10 = Dec 16. 16mph = 25kph
610 8 {20 00 07 E4 C0 00 00 00} = 12kph -------Hex 07 = Dec 07. 7mph = 11.3kph(perhaps digital dashes always roundup?)

Tommorow I'll try to post the WDS sheets and how to feed it info.

[raustinss]

Ummmmm.... yeah .. sure , I'd cough up the cash for the turn one box and prevent my brain from melting lol

[Sbeck09]

It's all a little over my head, but that's because my hands aren't on it. Fascinating project you have here. Funny too that you posted since I was just researching to see if the EPAS from a 2011+ mustang could be controlled in a similar fashion.

I'll be keeping tabs on your progress. I'm definitely hoping you succeed with minimal issue.

[andrewb70]

Looking forward to seeing where this goes. You got this!!!

Andrew

[c4racer2]

Hmm. Not sure what the end game is for all of this effort. I have several vehicles with electric steering - probably some of the best on the market - Porsche, Camaro SS, chevy truck, also had a mini.
I’m not feeling it. None are as good as real good old hydraulic based steering. I’ve owned lots of those cars too - old and new. Rack, boxes etc.
part of what I like about older cars is the connection you get to the road from the steering and brakes.
Not to be a Debbie downer. I appreciate the engineering behind this as a EE geek and all !

[stab6902]

Cool project! Having driven lots of cars new and old, I also prefer the feel of hydraulic power steering over electric, but it seems these column mount EPS conversions are getting more popular.

A couple of things to think about:

- Stock FRS's have about 1500 lbs over the front wheels and have 215 mm wide tires. Your A-body certainly weighs more and looks like it has much wider tires. This means your steering effort will be considerably higher (assuming similar box/rack ratios between the two cars).

- What steering box were you thinking of running? You don't want to run a power steering box because the T-bar (used to meter hydraulic assist) has to hit its mechanical stops before "acting" like a manual box. A hydraulic box used without a power steering pump will have a few degrees of "slop" due to T-bar flex. This of course means a manual box is pretty much necessary, but even the quick ratio manual boxes are slow by power steering standards. Ideally you'd probably want a 12.7:1 manual box, but I don't know that those exist.

- I think it's a great idea to feed a VSS signal to the EPS controller. Having variable effort assist is probably the biggest advantage to an EPS setup in my opinion. I have limited experience with EPS CAN interfaces, but I can confirm that the checksums are important. I'm not sure what the fallback/degradation strategy is for your EPS unit, but it could be that it defaults to "full" assist unless you feed it all the signals/checksums it needs to be "happy". One out of range signal or incorrect checksum (and there may be 5+ signals, even though VSS seems like it would be the only relevant one) may be enough to put the unit into fallback state (full assist). If you figure it out though, you'll be a hero!

[rchaskin]

Ummmmm.... yeah .. sure , I'd cough up the cash for the turn one box and prevent my brain from melting lol

:rolleyes::rolleyes:

HAHAHHAHAHA!!!

[BMR Sales]

I really feel like a Moron after reading all that but I'm interested.

In my Porsche LS6, I had to use a Manual Rack - I would love to drive and not have a Workout at the same time.

[andrewb70]

Hmm. Not sure what the end game is for all of this effort. I have several vehicles with electric steering - probably some of the best on the market - Porsche, Camaro SS, chevy truck, also had a mini.
I’m not feeling it. None are as good as real good old hydraulic based steering. I’ve owned lots of those cars too - old and new. Rack, boxes etc.
part of what I like about older cars is the connection you get to the road from the steering and brakes.
Not to be a Debbie downer. I appreciate the engineering behind this as a EE geek and all !

I actually encouraged Scott to stick with hydraulic steering. Sometimes we do things not because it makes the most sense, but for the challenge of it. I think this is as much the case here as anything else. I don't think there is anything wrong with that motivation.

With my Cougar I had very few options, since I had already installed a TCP manual rack. I liked certain aspects of the manual rack, while hating others. The EPAS has alleviated the negatives of the manual rack, but has also killed off a little bit of what I liked about it. It is definitely a trade off in my opinion. Also keep in mind that my Cougar is a street/strip car. I have 4.5" wide front wheels! So for me, this made a lot of sense. I am taking it a bit further than most people by integrating the speed sensitive variable assist, but this build is way more ambitious.

I also have to mention DCE. They sell what Scott is attempting to pull off. You can buy it right now:
http://www.dcemotorsport.com/Home/EPAS

The issue is that their stuff is not cheap. However, I've had a couple of chats with one of their engineers, and I think they have a pretty good product.

T.C., one of the DCE kits may work very well for you in the Porsche.

Andrew

[67King]

In my Porsche LS6, I had to use a Manual Rack - I would love to drive and not have a Workout at the same time.

I ditched the power steering in both of my race cars (944T and 968), and taking apart my GTO has reminded me how much of a PITA having the hydraulic system in place is, so I'm looking at these if I can find a spare steering column so I don't have to hack up mine (all #'s matching car).

Subscribed to the thread, but when it comes to electronics, the time saved may justify just buying a system rather than doing all of the research to learn how to use these.

[jetmech442]

Ummmmm.... yeah .. sure , I'd cough up the cash for the turn one box and prevent my brain from melting lol
lol, yeah I get it. And I'm certainly not bashing Lee or turnone, I've wanted one of their systems for over 7 years. But the prices are quite high. hopefully, being a thermal engineer I should be able to prevent my brain from melting.:razz:



It's all a little over my head, but that's because my hands aren't on it. Fascinating project you have here. Funny too that you posted since I was just researching to see if the EPAS from a 2011+ mustang could be controlled in a similar fashion.

I'll be keeping tabs on your progress. I'm definitely hoping you succeed with minimal issue.
Yeah I think I jumped into the details way too deep before explaining my motivation. I think all the links above help provide a decent foundation, but maybe I should do a better job of summarizing the takeaways to help others save some time. I was thinking about the EPS rack from the 2017 ZL1. People seem to love them and they can be had for 300-600$ on fleabay. I decided to go the FRS route(EPS on column) first because the zl1 rack requires a big garage-time investment in properly installing to eliminate bumpsteer and such, and also because I got the whole thing for 130$ and there is a lot of info on hacking the Toyota can systems for autonomous driving research. When I get it working, if it’s good,not great, I could look at the power on rack solution.


Hmm. Not sure what the end game is for all of this effort. I have several vehicles with electric steering - probably some of the best on the market - Porsche, Camaro SS, chevy truck, also had a mini.
I’m not feeling it. None are as good as real good old hydraulic based steering. I’ve owned lots of those cars too - old and new. Rack, boxes etc.
part of what I like about older cars is the connection you get to the road from the steering and brakes.
Not to be a Debbie downer. I appreciate the engineering behind this as a EE geek and all !
I get it for sure. I don’t have the expectation that my garage mashup is going to be better than a Porsche or SS. But I think it’s going to be better than what I have now, for not much money. I’m okay with the time investment, I love learning new systems and tuning something to get it just right.
Please send any pointers my way you may see, gearheads with EE backgrounds seem like a rarity.


Cool project! Having driven lots of cars new and old, I also prefer the feel of hydraulic power steering over electric, but it seems these column mount EPS conversions are getting more popular.

A couple of things to think about:

- Stock FRS's have about 1500 lbs over the front wheels and have 215 mm wide tires. Your A-body certainly weighs more and looks like it has much wider tires. This means your steering effort will be considerably higher (assuming similar box/rack ratios between the two cars).

- What steering box were you thinking of running? You don't want to run a power steering box because the T-bar (used to meter hydraulic assist) has to hit its mechanical stops before "acting" like a manual box. A hydraulic box used without a power steering pump will have a few degrees of "slop" due to T-bar flex. This of course means a manual box is pretty much necessary, but even the quick ratio manual boxes are slow by power steering standards. Ideally you'd probably want a 12.7:1 manual box, but I don't know that those exist.

- I think it's a great idea to feed a VSS signal to the EPS controller. Having variable effort assist is probably the biggest advantage to an EPS setup in my opinion. I have limited experience with EPS CAN interfaces, but I can confirm that the checksums are important. I'm not sure what the fallback/degradation strategy is for your EPS unit, but it could be that it defaults to "full" assist unless you feed it all the signals/checksums it needs to be "happy". One out of range signal or incorrect checksum (and there may be 5+ signals, even though VSS seems like it would be the only relevant one) may be enough to put the unit into fallback state (full assist). If you figure it out though, you'll be a hero!
Stab-thanks man! I was initially planning on using the 12.7 power box I currently have, but you’ve confirmed my suspicions about the Tbar as to why I shouldn’t do that. I still have the manual box in the attic, it’s not quick ratio, so Ill have to put some thoughts into a solution here, thanks again.

Also about the front tire forces. Yeah I’m running 285’s up front and have close to 1900lbs over the front tires. The ECU is mostly a brain with a motor controller mated to a thick aluminum heat sink plate. If you over heat, it reduces the assist provided. This will be something I keep my eye on as it may need a small comp fan to help keep it under the temp limits.

Without any can channels, Japanese OEMs seem to default to full assist(or so I’ve heard). I for sure know that the steering ECU needs torsion signal, steering angle, VSS and eRPM. But you may be right that it is looking for others. I’ll keep keep that in mind if it isn’t behaving as expected.


I actually encouraged Scott to stick with hydraulic steering. Sometimes we do things not because it makes the most sense, but for the challenge of it. I think this is as much the case here as anything else. I don't think there is anything wrong with that motivation.

With my Cougar I had very few options, since I had already installed a TCP manual rack. I liked certain aspects of the manual rack, while hating others. The EPAS has alleviated the negatives of the manual rack, but has also killed off a little bit of what I liked about it. It is definitely a trade off in my opinion. Also keep in mind that my Cougar is a street/strip car. I have 4.5" wide front wheels! So for me, this made a lot of sense. I am taking it a bit further than most people by integrating the speed sensitive variable assist, but this build is way more ambitious.

I also have to mention DCE. They sell what Scott is attempting to pull off. You can buy it right now:
http://www.dcemotorsport.com/Home/EPAS

The issue is that their stuff is not cheap. However, I've had a couple of chats with one of their engineers, and I think they have a pretty good product.

T.C., one of the DCE kits may work very well for you in the Porsche.

Andrew
It’s true! Andrew tried to talk sense into me, but the little idiot in the back of my head wouldn’t shut up. I really do enjoy this(and I get bored at work sometimes so this fills in the gap lol). This car has always been about making it better by making me better. 8 years ago I didn’t know anything about LS’s, ECU tuning with EFILive, suspension, brakes, steering-but now some of my favorite things about the car are that I tuned it and the adjustable 3 link I designed and installed. 80% of my knowledge came from the great people on this site, so now I’m going to have some fun and hopefully help some others out along the way--yes I’m out of my thermal wheelhouse so to speak, but what fun is doing only what you know?

I ditched the power steering in both of my race cars (944T and 968), and taking apart my GTO has reminded me how much of a PITA having the hydraulic system in place is, so I'm looking at these if I can find a spare steering column so I don't have to hack up mine (all #'s matching car).

Subscribed to the thread, but when it comes to electronics, the time saved may justify just buying a system rather than doing all of the research to learn how to use these.
Yeah King, having pulled the flow control valve several times to fine tune the orifice I just got tired of it. I’m looking forward to variable assist without the mess. My true goal here is to do the research for you guys and provide the code over on GITHUB that can just be used.

[andrewb70]

As far as I can determine, the "failsafe" level of assist is the medium level of assist, at least using the Prius/Yaris hardware and ECU. Time to read the FRS FSM!

Andrew

[andrewb70]

This might be helpful:
https://www.buyautoparts.com/buynow/1970/Chevrolet/Chevelle/Manual_Steering_Gear_Box/82-70031_NQ?xref=2502319

Andrew

[c4racer2]

oh - you would be surprised - lots of EE gearheads out there!

Good luck and keep us posted - I will pipe in if I can help in any way.

[David Pozzi]

I drove an FRS at Streets Of Willow track and found the variable effort distracting. I don't think I'd want it to vary, but maybe full power would be overboosted? Perhaps the dial adjustable conversion would be best.
There are fast ratio conversions for GM manual steering boxes. 16:1 I think.

[jetmech442]

Thanks for the on- track insight David. Your the first trackrat I've heard provide feedback on the system other than the glowing reviews from motortrend. Would you rather the variance be more "step function" like? Perhaps if the rpms are high and vss is above 50 then it sounds be a constant low assist? I say 50 cause that was a good show for Blackhawk farms. I'll definitely have the ability to lie to the steering ecu about rpm and speed to suit my needs if I like.
BTW, always been a big fan of the killer cars you and Mary create.:cheers:

[67King]

Thanks for the on- track insight David. Your the first trackrat I've heard provide feedback on the system other than the glowing reviews from motortrend. Would you rather the variance be more "step function" like? Perhaps if the rpms are high and vss is above 50 then it sounds be a constant low assist? I say 50 cause that was a good show for Blackhawk farms. I'll definitely have the ability to lie to the steering ecu about rpm and speed to suit my needs if I like.
BTW, always been a big fan of the killer cars you and Mary create.:cheers:

FWIW, a pretty good friend of mine who instructs both with PCA and at the Porsche center in Atlanta has one that has all but become a dedicated track car, and has no issues with it. Granted, his comparator cars were 944's (Turbo and S2), so much lighter (and 50/50 weight distribution). But FWIW, another anecdote.

[David Pozzi]

I'm sure I could get used to it but it wasn't something I found desirable. Effort was changing mid-corner as I was headed towards an apex, which I really don't like. I'd rather have it up the effort at higher speeds, like 80mph up, than vary it at lower speeds.

[67King]

God that sounds awful. Ian's is a BRZ, I know some stuff was different, wonder if steering tuning was?

[BMR Sales]

I'd rather have it up the effort at higher speeds, like 80mph up, than vary it at lower speeds.

Exactly!

[andrewb70]

The assist strategy for the earlier Toyota parts is very simple. The faster you go, the less assist you get. With the Yaris ECU that I am using, there are three levels of assist. Since the speed signal is coming from the Holley Dominator ECU, I can alter the PPM output and alter the MPH where the medium assist kicks in. Right now I have it set to go into medium assist at 35 mph, and high assist kicks in around 50. I can't change the relationship between medium and high assist, because that is programmed into the steering ECU, but I can shift them both up and down.

What Scott is trying to do is on another level!

Andrew

[RobNoLimit]

We have done 6 EPAS swaps, 5 are up and driving. 3 were on Cobra kit cars, 1 Fox body and 1 Nova. For the rack/pinion cars, Sweet makes rack just for this. Large heavy duty inputs and shafts for the boosted input, and your choice of ratio's. We tried a Yaris ecu with the speed input as Andrew is using. A little tricky to get it done but not too bad. For us, mostly Auto-X and Road Course, the assist add was 'goofy' or 'backwards'. And, with no real way to swap the input relations, we finally stripped it back to a manual (pot) tune. One assist level throughout. This gave the driver the most consistent feel, but in our minds, it could be better.
Ideally, we wanted to be able to build a multi layer tune map. with Speed, G-meter, brake, and Steering angle grids. We hoped for a smooth assist gain during slow speed or decel, reduce assist if the brake is on, and progressively reduce assist as the steer angle increases. But, it was not available with the hardware we had. The static set up works very well, although we usually run it on the 'low' side of perfect, too much high speed assist makes the car flighty, so we tune for that and add some muscle at low speeds. Personally, now that I have some experience with them, I prefer the EPAS to the usual hydro units now. So much so, that we are working on a version of our WideRide IFS using an EPAS Rack.
Glad to see this project moving forward, Good luck!

[andrewb70]

Rob,

I am pretty sure that without the CAN box, the Yaris ECU will default to "failsafe" mode. I didn't care for how the car drove in "failsafe" mode, but once I added Blake's CAN box, it is much better. Also keep in mind that my Cougar is not a race car. I am looking for good street manners, which I feel I am getting close to achieving. If you have any pictures of how you mounted the various boxes, I am sure people would love to see.

I am also pretty sure that DCE has what you want in terms of fully programmable steering ECU based on a number of inputs.

Andrew

[rickpaw]

I'm wondering if the EPS would adequate to turn big tires for a 4x4? I just got a 74 FJ40 and been thinking about converting it to power/assisted steering.

[jetmech442]

I'm sure I could get used to it but it wasn't something I found desirable. Effort was changing mid-corner as I was headed towards an apex, which I really don't like. I'd rather have it up the effort at higher speeds, like 80mph up, than vary it at lower speeds.
I'm hoping I don't see that behavior in this system, but it's something I'm going to keep my eye on.


We have done 6 EPAS swaps, 5 are up and driving. 3 were on Cobra kit cars, 1 Fox body and 1 Nova. For the rack/pinion cars, Sweet makes rack just for this. Large heavy duty inputs and shafts for the boosted input, and your choice of ratio's. We tried a Yaris ecu with the speed input as Andrew is using. A little tricky to get it done but not too bad. For us, mostly Auto-X and Road Course, the assist add was 'goofy' or 'backwards'. And, with no real way to swap the input relations, we finally stripped it back to a manual (pot) tune. One assist level throughout. This gave the driver the most consistent feel, but in our minds, it could be better.
Ideally, we wanted to be able to build a multi layer tune map. with Speed, G-meter, brake, and Steering angle grids. We hoped for a smooth assist gain during slow speed or decel, reduce assist if the brake is on, and progressively reduce assist as the steer angle increases. But, it was not available with the hardware we had. The static set up works very well, although we usually run it on the 'low' side of perfect, too much high speed assist makes the car flighty, so we tune for that and add some muscle at low speeds. Personally, now that I have some experience with them, I prefer the EPAS to the usual hydro units now. So much so, that we are working on a version of our WideRide IFS using an EPAS Rack.
Glad to see this project moving forward, Good luck! Yeah a mapped system would be ideal, and depending on how everything goes, I may look into it. My biggest goal out of this was to have a system that works well on the cheap(aren't we all), hence starting with the FRS system. My first reaction was to start making maps in the new arduino FPGA since they're so freaking fast, but then I read some white papers published by Ford on issues they ran into concerning torque ripple and transient gain curves, and figured if I could get a factory system to perform then I'd be miles ahead. The way I see it is right now, it's at least going to be a lot more tuneable and a lot cheaper. It's reassuring to hear though that the heavy hitters have been down this path successfully from a business standpoint.


I'm wondering if the EPS would adequate to turn big tires for a 4x4? I just got a 74 FJ40 and been thinking about converting it to power/assisted steering. I'll let you know how the frs system works on the 285's on my cutlass. I would think it would work great with one caveat- the ECU has the brains but it is also is a large motor controller. The FRS system is circuit breakered at 80 amps(I don't know how much it will actually pull). The manual states that if the system, overheats, it will reduce assist until it cools off. For hot ambients and lots off full assist usage like rock crawling, I would probly mount a small fan to the back of the thick aluminum heat sink. I plan on mounting a thermocouple to mine to see average case temps during regular driving and autoX/roadcourse, but that won't be meaningful until it gets warm again. I'll bet if you grabbed one out of a truck it would be ready to roll.

[jetmech442]

So I made a down n dirty test rig-easy now, this isn't instagram-this is more inline with an episode of roadkill lol. I'll add a system diagram to illustrate, instead of trying to use a bunch of words to explain it.
So the good news is it powers up and when I twist the long shaft the other side is definitely boosted. this was an awesome feeling-I mean, again, standing on the shoulders and such, but it was nice to be able to at least duplicate what others have got working. I have the CANBUS in read mode, thinking that the ECU should be broadcasting at least something, but I'm not picking up anything yet. I need some distraction-free tinker time to figure it out.

In the meantime, I'll detail how I wired it up. the wiring diagram was easily found on ft86club forums. the section is called "EPS". There are 4 connectors. Z1 is the 3 wire connector that comes directly from the column.Z2 is the 12 wire connector. It feeds steering angle and torque sensor values to the ECU. A14 is where you feed high gage power and ground. the terminal spades are .375'' wide, so you know, the big stuff. A15 needs just three wires, switched ignition power to pin6, and CanH to 1 and CanL to 7. That's really it, 5 wires.
159421










159425










This shows the way I wired up the arduino. The only thing not shown is the usb cable feeding into the laptop to printout the can messages received.
159426












159427

[andrewb70]

Solid progress! I am surprised that the steering ECU does not have the resister built in. The Yaris and Prius ECUs act as one of the ends of the CAN bus with the engine ECM being the other end.

What is the "resolver?"

Andrew

[jetmech442]

Solid progress! I am surprised that the steering ECU does not have the resister built in. The Yaris and Prius ECUs act as one of the ends of the CAN bus with the engine ECM being the other end.

What is the "resolver?"

Andrew

Thanks Andrew! MY interpretation of the resolver is to mean the steering angle, makes sense to me as 4 wires for steering angle, and 4 for shaft torque. I'm hoping to prove that through the Can channels-my other assumption is that steeringecu broadcasts at least the steering angle for others to use(like stability control, and self-park in some models).

speaking of CAN channels-I plugged in and loaded the "receive blink" example for the arduino, and sure enough the ECU was broadcasting ID Ox370 . 8 bits in length, the 3rd bit was consistently 26, and the 6th bit varied, but did not really correspond to steering angle or torque. For example, the message looked like 0x370 0 0 26 0 0 11 0 0. where the "11" kinda bounced around.

I think I'm ready to broadcast out some mocked up speed and rpm messages to see if assist reduces like I hope it does, but I don't have a definitive way of measuring if it actually reduces.

I was thinking about hanging a 10lb weight off the vise grips on the input haft, and then measure the other side compressing an old valve spring....? Trying not to let myself get distracted in some complicated measurement setup, I'd like to keep moving forward with thte momentum I've got.--I'm open to any suggestions.

[jetmech442]

So I can confirm that 0x370 is shaft torque. I don't know what the magnitudes or units are, but the graph below definitely corresponds to the force I was exerting on the shaft. Also, I don't see anything in the message that talks about torque direction, because the first hump was clockwise and the second CC, but the magnitudes are the same. This is really cool!

I also found another Can ID that is sent not as frequently, it is 0x4C8. Not sure about this guy yet. My assumptions were that the ECU would broadcast both steering angle and torque, so I'm hoping the 2nd is angle.

I don't necessarily need to fully understand these channels, but it is helping me a lot to understand the system.

159498

[andrewb70]

It sure would be helpful if there was a document that fully described the Toyota steering control strategy. LOL

Solid progress!!!

Andrew

[raustinss]

Again ... my brain hurts , not sure why lol

[jetmech442]

It sure would be helpful if there was a document that fully described the Toyota steering control strategy. LOL

Solid progress!!!

Andrew
Sigh. I know right?!


Again ... my brain hurts , not sure why lol Mine is starting to hurt as well, lol.

I realized that my power supply displays amps, and I've been using that to judge assist. The amperage reading, while not accurate i'm sure, is at least repeatable. So I've clamped the output to a locked position, and the input, I am hanging 7.5 lbs off an 8'' pair of vise grips. I've been flashing in updated CAN ID's and watching amperage, and so far none of them have had an effect. For transparency to the guy who finds this thread in 4 years(like me who has found so many from 2014ish), the Can messages I have sent using the "send" example are as follows:

CAN.sendMsgBuf(0xB4, 0, 8, stmp);

//unsigned char stmp[8] = {0, 0, 0, 0, 0xA0, 6, 102, 200}; //10.1mph- I've tried with and without the 5th byte which is supposed to be some sort of counter
//unsigned char stmp[8] = {0, 0, 0, 0, 0, 0xF, 0xC0, 0x8B}; //24.99 mph
//unsigned char stmp[8] = {0, 0, 0, 0, 0, 0x1C, 0x5A, 0x32}; //44.99 mph


Also tried:
CAN.sendMsgBuf(0x610, 0, 8, stmp);
//unsigned char stmp[8] = {0, 0, 0x90, 0, 0, 0, 0, 0xAE}; // supposedly the 3rd bite is straight speed

I'm starting to wonder if the vehicle CAN ID for the 2014 FRS is different than the Prius/Camry values I've been able to scavenge from the sources in my first post. I have a friend at work who has an FRS that I can try to get a CAN dump from, but won't be back at work until the new year, so I"m going to keep searching and see if I can find any other values other than 610 or B4 to try.

[andrewb70]

Scott,

I think we may have to go down the rabbit hole of looking for Subaru CAN IDs. The FRS/BRZ are clones...

There are a couple of guys on ls1tech that may be able to help, but they are going down their own rabbit hole of developing a free gen 3 ECU flashing software.

Andrew

[jetmech442]

Scott,

I think we may have to go down the rabbit hole of looking for Subaru CAN IDs. The FRS/BRZ are clones...

There are a couple of guys on ls1tech that may be able to help, but they are going down their own rabbit hole of developing a free gen 3 ECU flashing software.

Andrew

Blindsided by the joint-venture,lol. um, yeah I think I may need to descend a bit further down the hole. I guess if it only took a non-electrical guy a few weeks to do it, then everyone would have it right? time to put google back to work.....

[AE95]

Andrewb70 contact me a while ago about my eps conversion I did back in 2012. I own an 1992 toyota corolla alltrac wagon. These rare Toyota wagon are awd. I swapped a 1989 toyota mr2 supercharged 4 cyc into the wagon. Problem was the mr2 did not have power steering and the supercharger is mounted where the power steering pump would have gone. I really wanted to keep power steering. I had worked for lexus dealership back then and knew a lot about both EPS and CAN system. After some research I was able to find a few people that had used Yaris and Prius Electric power steering columns. So took a chance and bought some Prius columns and a yaris non ABS ecu. While the can Ecu will go right into fail safe with no can signal, I wanted to get the whole system working. Easier to just use the speed sensor signal that my Toyota already has and not try to duplicate it into the can network. Now one area I want to address is people saying that the system if functioning at 100% assist when in fail safe. That is not true. If there is a malfunction in the CAN network or speed signal then the system goes into default and this results in assist being fixed at the same assist has it has at 43 MPH which is the min assist.
FAIL-SAFE CHART
When a problem occurs in the electronic power steering system, the P/S warning light on the combination meter comes on, and the ECU stops the steering power assist, maintains the amount of power assist constant, or reduces the amount of power assist to protect the system.
DTC NO. MALFUNCTION FAIL-SAFE
C1511/11
C1512/12
C1513/13
C1514/14 Torque sensor malfunction Assist force restricted
C1517/17 Torque sensor malfunction Power assist stopped
C1524/24 Motor malfunction Power assist stopped
C1531/31
C1532/32 Power steering ECU malfunction Power assist stopped
C1533/33 Temperature sensor malfunction in power steering ECU Power assist stopped
C1541/41
C1542/42 Speed sensor malfunction Amount of power assist remains constant at speed of 43 mph (70 Km/h)
C1551/51 IG power source voltage error Assist force restricted
C1552/52 PIG power source voltage error Assist force restricted
C1554/54 Power source relay malfunction Power assist stopped
C1555/55 Motor relay Malfunction Power assist stopped
U0073 CAN bus malfunction Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
U0105 ECU communication error Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
U0121 ABS ECU communication error Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
- Extremely high temperature in ECU Assist force restricted until normal ECU temperature recovers
- Power source voltage drop Assist force suspended until voltage recovers

This is the same for the FRS ecu. Other issue that you will have tell you can get the CAN working it will not be able to program zero point calibration for the steering wheel center. The zero point I have only be able to do the calibration with a scan tool. There are instruction to do it with out a scan tool but you need to be able to see the P/S light on the instrument cluster which is done over the CAN. Scan tool I have but had to figure out how to duplicate the RPM signal on the can which is the CAN signal that EPS ECU looks for has a wake up signal. I found a guy that post a youtube video recording of the Yaris CAN data. I found a module that was programmable for CAN transmitting (cando auto module) . Mine you not has much info or Arduino CAN info back then. By wiring the module in connecting my scan tool to the CAN network then start to program the HEX data in tell I came out the correct one for RPM. Then I was able to program the Zero point using the scan tool and the system was now fulling operational. I can confirm that it is a different then when the system is just in fail safe.
Looking at information on the FRS EPS. This system is the newest version of column EPS system most likely was made my jtekt part of the Koyo Bearings company. They make most of Toyota steering components. Like Racks and EPS. It looks like the FRS system is a jtekt made one. It has the look of a Toyota style. But the Toyota part number has an SU at the beginning which may show signs of it having some Subaru influence but after some looking at the information I could find I am inclined to believe it is Toyota protocol. But it is running on the newer Toyota can protocol and since several companies and even and MIT student had hack the setup in the prius for remote and autonomous vehicle. So I would not be surprised if Toyota changed things up on EPS protocol that is not required universal OBD2 data like the RPM signal is. The other change is the Yaris system runs on 12 volts. Both to the ecu and to the electric motor. The FRS eps esu setups the 12 volts up to 35 volts. We saw this same thing on some of the Lexus EPS systems. One Thing is the connection for the ECU is the same on the scion TC and a few others so if we end determining that the CAN protocol is Subaru based you may be able to pick up a different ECU to play with.
Just for fun here are some pic of my car and setup!

https://static1.pt-content.com/images/pt/2019/01/photo_zps309483ee-1.jpg

4AGZE Engine has since been removed. Currently swapping a Camry 1mzfe v6 with TRD supercharger.

https://static1.pt-content.com/images/pt/2019/01/photo_zps13543775-1.jpg

Prius column top and 92 corolla tilt column lower. Note that is also a break away column if in an accident. I wanted to keep that function.

https://static1.pt-content.com/images/pt/2019/01/photo_zpsef98482c-1.jpg

Both disassembled and just the parts I need laid out.

https://static1.pt-content.com/images/pt/2019/01/photo_zpsa0398a17-1.jpg

Also was greatly surprised to find out the u-joint from a 92 and a 2004 would be the same size. I was able to grind off the stack marks , press the caps off and make one unit that would attach the EPS to the Stock Rack and pinion. The rack and pinion I looped the lines and weld up the hydraulic slip joint.

https://static1.pt-content.com/images/pt/2019/01/photo_zps6a6fbc17-1.jpg

https://static1.pt-content.com/images/pt/2019/01/photo_zps3bf90e15-1.jpg

Cut weld shafts together and the end result.

https://static1.pt-content.com/images/pt/2019/01/photo_zps68920aa1-1.jpg

Then Installed into vehicle. And no there is plenty of space between the motor and gas pedal.

https://static1.pt-content.com/images/pt/2019/01/photo_zps9a36d8d4-1.jpg

Cando auto module

https://static1.pt-content.com/images/pt/2019/01/photo1_zpsad8c210a-1.jpg

The completed Build of the vehicle minus any info on the v6 swap currently under way.

http://forum.alltracwagon.com/cgi-bin/forum/Blah.pl?m-1347147405/s-105/

more information on my setup and also on a couple of others from some Australian’s

http://www.toymods.org.au/forums/threads/79652-Electric-power-steering

My current Daily Driver and weekend desert runner Another AWD corolla wagon with my own custom 4 inch lift and 27 inch tires. That will also be getting a EPS install some time down the road when it also gets a v6 swap from a Lexus RX300. Already have all the parts just need to finish the other wagon first. I also have an 87 supra with a built 7mgte. I have the EPS rack from an IS350 and ecu I would love to try and mount in the supra. So I am a little keen to get this all working since its ecu is a CAN for both the RPM and Speed signal.

https://static1.pt-content.com/images/pt/2019/01/image1_zpsp3gzi6zu-1.png

[andrewb70]

Jared,

Thanks for posting!!!

Andrew

[jetmech442]

Andrewb70 contact me a while ago about my eps conversion I did back in 2012. I own an 1992 toyota corolla alltrac wagon. These rare Toyota wagon are awd. I swapped a 1989 toyota mr2 supercharged 4 cyc into the wagon. Problem was the mr2 did not have power steering and the supercharger is mounted where the power steering pump would have gone. I really wanted to keep power steering. I had worked for lexus dealership back then and knew a lot about both EPS and CAN system. After some research I was able to find a few people that had used Yaris and Prius Electric power steering columns. So took a chance and bought some Prius columns and a yaris non ABS ecu. While the can Ecu will go right into fail safe with no can signal, I wanted to get the whole system working. Easier to just use the speed sensor signal that my Toyota already has and not try to duplicate it into the can network. Now one area I want to address is people saying that the system if functioning at 100% assist when in fail safe. That is not true. If there is a malfunction in the CAN network or speed signal then the system goes into default and this results in assist being fixed at the same assist has it has at 43 MPH which is the min assist.
FAIL-SAFE CHART
When a problem occurs in the electronic power steering system, the P/S warning light on the combination meter comes on, and the ECU stops the steering power assist, maintains the amount of power assist constant, or reduces the amount of power assist to protect the system.
DTC NO. MALFUNCTION FAIL-SAFE
C1511/11
C1512/12
C1513/13
C1514/14 Torque sensor malfunction Assist force restricted
C1517/17 Torque sensor malfunction Power assist stopped
C1524/24 Motor malfunction Power assist stopped
C1531/31
C1532/32 Power steering ECU malfunction Power assist stopped
C1533/33 Temperature sensor malfunction in power steering ECU Power assist stopped
C1541/41
C1542/42 Speed sensor malfunction Amount of power assist remains constant at speed of 43 mph (70 Km/h)
C1551/51 IG power source voltage error Assist force restricted
C1552/52 PIG power source voltage error Assist force restricted
C1554/54 Power source relay malfunction Power assist stopped
C1555/55 Motor relay Malfunction Power assist stopped
U0073 CAN bus malfunction Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
U0105 ECU communication error Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
U0121 ABS ECU communication error Amount of power assist remains constant at speed of 43 mph (70 km/h) note for FRS is 62 mph
- Extremely high temperature in ECU Assist force restricted until normal ECU temperature recovers
- Power source voltage drop Assist force suspended until voltage recovers

This is the same for the FRS ecu. Other issue that you will have tell you can get the CAN working it will not be able to program zero point calibration for the steering wheel center. The zero point I have only be able to do the calibration with a scan tool. There are instruction to do it with out a scan tool but you need to be able to see the P/S light on the instrument cluster which is done over the CAN. Scan tool I have but had to figure out how to duplicate the RPM signal on the can which is the CAN signal that EPS ECU looks for has a wake up signal. I found a guy that post a youtube video recording of the Yaris CAN data. I found a module that was programmable for CAN transmitting (cando auto module) . Mine you not has much info or Arduino CAN info back then. By wiring the module in connecting my scan tool to the CAN network then start to program the HEX data in tell I came out the correct one for RPM. Then I was able to program the Zero point using the scan tool and the system was now fulling operational. I can confirm that it is a different then when the system is just in fail safe.
Looking at information on the FRS EPS. This system is the newest version of column EPS system most likely was made my jtekt part of the Koyo Bearings company. They make most of Toyota steering components. Like Racks and EPS. It looks like the FRS system is a jtekt made one. It has the look of a Toyota style. But the Toyota part number has an SU at the beginning which may show signs of it having some Subaru influence but after some looking at the information I could find I am inclined to believe it is Toyota protocol. But it is running on the newer Toyota can protocol and since several companies and even and MIT student had hack the setup in the prius for remote and autonomous vehicle. So I would not be surprised if Toyota changed things up on EPS protocol that is not required universal OBD2 data like the RPM signal is. The other change is the Yaris system runs on 12 volts. Both to the ecu and to the electric motor. The FRS eps esu setups the 12 volts up to 35 volts. We saw this same thing on some of the Lexus EPS systems. One Thing is the connection for the ECU is the same on the scion TC and a few others so if we end determining that the CAN protocol is Subaru based you may be able to pick up a different ECU to play with.
Just for fun here are some pic of my car and setup!


Dude I coulda swore I recognized your username and some of those pics, then I realized I have your thread bookmarked from toymods. Your explanation of the Canbus termination actually made sense after having read a half dozen or so word-only explanations, so thanks for taking the time to post-especially the above. I've read the above twice now,and I'm sure I'll continue to absorb more as I re-read, but a few things already stuck out:
Centering the system. The FRS uses a separate Steering Angle Sensor(SAS) that I do not have, it is quite pricey when new at 250$. I'm not entirely sure what the "resolver" does in the wiring diagram of the Steering Control Unit(SCU), but my SCU does not broadcast angle, so I'll have to figure something out there. Im hoping to map a much cheaper sensor to the right Can message. Interestingly, Can ID 0x18, which I'm convinced is Steering angle on the 500kbps Bus, does seem to have a static angle stored in the 3rd Bite that never changes, even from day to day. I think I may be able to center the system when I broadcast 0x18 by controlling that bite. IF not, then a friend lent me his TechStream cable and disc, which can do the programming, I'm just not sure if it will complete the alignment since it also tries to reset the YAW sensor, which I don't have(or need). also not sure if it needs to communicate with the rest of the Bus. But anywya, when the time comes, I'll try it out and tell you how it went.

The fact that the TechStream is a toyota service tool tells me SCU should be toyota CAN, but I am finding a lot of ID's familiar to the WRX guys over on NASIOC (great resource and those guys take nerdy Can hacking/HEX editing to the next level!!!!). Thats really good insight though that the TC has the same wiring as the FRS, if the values that Tulsa and Fabio and the MIT guys would work on the TC module, then half the battle would be over.

I love the wagon by the way, what a baller car. I'll bet with the supercharger its just a killer driver. Again man, thanks for the in depth post, if you see me heading down a path you've already tried feel free to steer me right. .

[dhutton]

You have to admire guys who can bring tech at this level to their builds.

Don

[andrewb70]

This project makes my EPAS install look like child's play...I am very curious to see where it goes.

Andrew

[AE95]

Dude I coulda swore I recognized your username and some of those pics, then I realized I have your thread bookmarked from toymods. Your explanation of the Canbus termination actually made sense after having read a half dozen or so word-only explanations, so thanks for taking the time to post-especially the above. I've read the above twice now,and I'm sure I'll continue to absorb more as I re-read, but a few things already stuck out:
Centering the system. The FRS uses a separate Steering Angle Sensor(SAS) that I do not have, it is quite pricey when new at 250$. I'm not entirely sure what the "resolver" does in the wiring diagram of the Steering Control Unit(SCU), but my SCU does not broadcast angle, so I'll have to figure something out there. Im hoping to map a much cheaper sensor to the right Can message. Interestingly, Can ID 0x18, which I'm convinced is Steering angle on the 500kbps Bus, does seem to have a static angle stored in the 3rd Bite that never changes, even from day to day. I think I may be able to center the system when I broadcast 0x18 by controlling that bite. IF not, then a friend lent me his TechStream cable and disc, which can do the programming, I'm just not sure if it will complete the alignment since it also tries to reset the YAW sensor, which I don't have(or need). also not sure if it needs to communicate with the rest of the Bus. But anywya, when the time comes, I'll try it out and tell you how it went.

The fact that the TechStream is a toyota service tool tells me SCU should be toyota CAN, but I am finding a lot of ID's familiar to the WRX guys over on NASIOC (great resource and those guys take nerdy Can hacking/HEX editing to the next level!!!!). Thats really good insight though that the TC has the same wiring as the FRS, if the values that Tulsa and Fabio and the MIT guys would work on the TC module, then half the battle would be over.

I love the wagon by the way, what a baller car. I'll bet with the supercharger its just a killer driver. Again man, thanks for the in depth post, if you see me heading down a path you've already tried feel free to steer me right. .

The resolver is the Rotation angle sensor. so no need to worry about getting something else or speeding any more time on it has it already works. Toyota has more then one zero point calibration. Yes there is one for the yaw rate sensor, but the one I am referring to is zero point for the torque sensor and at the same time it will sset the steering angle sensor for center. If you have the cable and techstream then the calibration will be in the EPS module then on the left side bullets you will select the utilities and under there with either say initialization or zero point. then follow the instruction. something like center the steering wheel and stay still for 6 seconds. then it is calibrated. ECU know where center is for the steering wheel and torque sensor neutral position. The only 2 things you need to focus on is one sending the RPM over can which you already have and then sending the speed signal over can. This was I the main reason I went with the non ABS yaris ecu has you not only need to figure out the HEX code for speed. but the write a code that will take the analog signal and convert to the correct HEX code. But I will love to see someone get it worked out has it would open a lot more option of EPS system. Like the Lexus IS350 EPS rack and pinion that I want to put on my 87 Toyota supra.

https://static1.pt-content.com/images/pt/2019/01/EPS20rack_zps8i5iyxla-1.png

one other thing you should do is go to techinfo.toyota.com you can sign up for a 2 day for like 20 bucks. print out all the repair manual pages for the EPS system. but also click on the 'new car features" will just be the abbreviation "NCF". for the 2014 scion FR-S. This will give you a way better explanation of how the system is built and works.

[jetmech442]

The resolver is the Rotation angle sensor. so no need to worry about getting something else or speeding any more time on it has it already works. Toyota has more then one zero point calibration. Yes there is one for the yaw rate sensor, but the one I am referring to is zero point for the torque sensor and at the same time it will sset the steering angle sensor for center. If you have the cable and techstream then the calibration will be in the EPS module then on the left side bullets you will select the utilities and under there with either say initialization or zero point. then follow the instruction. something like center the steering wheel and stay still for 6 seconds. then it is calibrated. ECU know where center is for the steering wheel and torque sensor neutral position. The only 2 things you need to focus on is one sending the RPM over can which you already have and then sending the speed signal over can. This was I the main reason I went with the non ABS yaris ecu has you not only need to figure out the HEX code for speed. but the write a code that will take the analog signal and convert to the correct HEX code. But I will love to see someone get it worked out has it would open a lot more option of EPS system. Like the Lexus IS350 EPS rack and pinion that I want to put on my 87 Toyota supra.



one other thing you should do is go to techinfo.toyota.com you can sign up for a 2 day for like 20 bucks. print out all the repair manual pages for the EPS system. but also click on the 'new car features" will just be the abbreviation "NCF". for the 2014 scion FR-S. This will give you a way better explanation of how the system is built and works.

Okay, I assumed the resolver would be the angle, but I also assumed it would broadcast the angle from the SCU out on the CANBus. But when I only saw 0x370 and 0x4C8, I figured my assumptions were wrong and the resolver must not be what I thought it was, but I hear what your saying, so I'll concentrate on speed and RPM first. Thanks for the description with the techstream, I hadn't actually wired it up and played with it yet, I was just going off the zeroing instructions I found online, so I'll try to get to that tomorrow morning. Ill defiinitly go to techinfo.toyota.com and grab the extra info, I didn't know that was another option.

I've been working on decoding everything at night when the tiny ones go to bed, but it's slower going than I'd hoped. I'll post up in a few what I have.

[jetmech442]

0xD4 wheel speed
Little endian(that just means put the higher order bite first:so if you use excel, RPM=hex2dec(concatenate(B1,B0))*.03495
or if you already converted the hex to decimal, the equation below will work just the same.
(B1*256 + B0) *.03495
example of message contents: {0xD9, 0x7, 0xDC, 0x7, 0xE2, 0x7, 0xE1, 0x7} is an example of 70.2 mph.
each group of 2 is a certain wheel. I don't know really which is which...don't really care right now I guess.

s160202





0x140 has a bunch of stuff in it, I care most right now about engine RPM
B2 and B3 are RPM.
little endian
example {0x1E, 0x9, 0x8E, 0xF, 0x1E, 0x1E, 0x22, 0x20}; this message content is for 3982 RPM.

PROTIP: if your at idle, then B2 has a "4" added to the hex number which throws off the reading. They may use the 4 for a purpose, like to signal if clutch was used, I don't know yet. For example, B2 = 43 and B3 = E8. You need to strip the "4" off to properly have E83, which equals 1000rpm. otherwise you have E843 which is 17384 rpm.
160203






0x18 has to do with Steering angle. I've been working on deciphering it, but I need some legit time to get nerdy and be able to think, so maybe next year lol.
For the graphs below, the car was turned on and idle. then we only turned the wheel left to the stops, then right to the stops, then back to center.
B0 I think has to do with a way to tell if turning cw or ccw, maybe velocity also? not sure.
B1
B2 is static at 112. I think this is programmed as the "zero" when you use the techstream. I might have chance to get a dump from a BRZ, this number would be different if it is what I think it is.
B4,B5,B6 are empty
B7 a much noisier signal that follows the movements of the wheel.
160205

160206

160207

160204

[andrewb70]

We are diving deep...love it.

Andrew

[jetmech442]

Okay. I've taken all 18 messages I captured on the high speed bus (500kbps), and re-broadcast them to my test rig, but I'm not seeing any reduction in current. I have a few thoughts I need to chase down, I'd be more than happy to entertain comments/suggestions or questions-anything that I havent considered. I am currently applying a torque that demands 2.5 amps, and broadcasting the messages taken from a segment where we were doing 70mph. I assume there is much less assist at 70mph(thats the whole point of this at least).

1)Assist reduction doesn't occur the way I think it does.
If it doesn't reduce assist statically, then I'm not entirely sure how I would try to measure a dynamic torque with repeatable accuracy. Open to suggestions.

2) My arduino code is wrong.
Highly possible-I've cleaned it up as much as I can to reduce any errors. And later today I'm going to post over on seedstudio to see if someone feels like belittling me for my poor excuse of coding lol. I don't want to post the code here yet because it doesn't work. If anyone wants it though to help I'm happy to PM them.



3) my arduino is not properly re-broadcasting the messages.
I rebroadcast RPM to my G8GT so I think it's okay, but I can't be entirely sure what happens when I try to send all 18 messages. do I need a delay? Does the delay mess up the sync with the bus? I dunno. I could get another Uno and CanShield and have it read in what I think I'm broadcasting. wouldn't hurt I guess, as I could use the extra setup on the G8 later on.

4) 0x18 was captured from the FRS and is definitely lots of stuff for steering angle. My SCU doesn't broadcast that angle, but it does have a resolver which I believe(especially after talking with AE95) is also a steering angle. Maybe 0x18 and the internal resolver have to match somehow? maybe its a fail safe? I don't know, just thinking of possible causes.

5) some of the channels, like 0x18 have an incrementer on the last byte that increases everytime the message is sent. It's not steady. I wonder how precise the thing needs to be, or if it syncs with an internal clock or something weird like that.


Okay, thats what I got. The journey continues.

[jetmech442]

Okay, so I bought a book. yes-it's getting that bad, I know. "Controller Area Network Prototyping with Arduino" is from Copperhill Technologies and he covers the J1939 standard and how the arduino fits into it, as well as capabilities/limitations, and he provides libraries to interface with CAN.
One of the takeaways is that arduino is not quite fast enough to read EVERY message and post it to the screen via Serial.print(and definitely not fast enough to store them in the buffer and the read/write to an SD card. In order to get the data that I showed above, it strips off a lot of data and just leaves you with the ID and the 8 byte data set.

Now, this still may be my go-to solution once I've completely figured out what channels are important and how to exactly replicate them(including the Cyclic Redundancy Counter). But in the meantime I need a platform that has troubleshooting software built in and has play-back capabilities. So, in comes Raspberry Pi and the PiCan2 module! I have them both on my desk right now downloading all the setup files. :) So next steps are to re-record the FRS at 70mph and doing turning stuff, probably start next week because my buddies FRS is under the knife for the valvespring recall, but I'm hopeful that once I get a clean scan on the PiCan2 and use the playback feature to isolate the channels that the SCu cares about.

[Conundrum]

I hope you have break through, my programming or my electronic skill is not as strong as yours.

Have you contacted this guy?

https://www.geraldjustprojects.com/brzfrs-ecu-cluster-canbus-upadte/

Here is what he posted in his blog:

As for power steering, This system operated in reverse. When the ECU is working correctly, It will read the RPM signal from the engine. if the engine is not running. it will not turn on power steering. Once the engine has been started it will turn it on as usual. But in a event of a emergency of something is wrong with the system. it can determine if there is no signal coming from the engine what so ever and power has been applied to it it will default to ON, then broadcast a fault code.

[andrewb70]

Most systems call that "failsafe" mode. So in case engine dies while driving, PS is still functional.

Andrew

[jetmech442]

I hope you have break through, my programming or my electronic skill is not as strong as yours.

Have you contacted this guy?

https://www.geraldjustprojects.com/brzfrs-ecu-cluster-canbus-upadte/

Here is what he posted in his blog:

As for power steering, This system operated in reverse. When the ECU is working correctly, It will read the RPM signal from the engine. if the engine is not running. it will not turn on power steering. Once the engine has been started it will turn it on as usual. But in a event of a emergency of something is wrong with the system. it can determine if there is no signal coming from the engine what so ever and power has been applied to it it will default to ON, then broadcast a fault code.

Hey conundrum, thanks for the link! i'm going to spend my usually uneventful friday going through it. Right after my last post my wife decided we needed to move. So we had to complete all the open projects in the house, re-install the brakes I just pulled off the cutlass, well the house, buy the new house. All that went cool, then the 3rd week in the new house a 70 foot white Burr Oak came down for no reason. Quoted 2K to turn it into firewood-but what a waste! so I did the only responsible thing, which is buy an 800 dollar chainsaw and an alaskan chainsaw mill. So now I have more lumber than I know what to do with lol.

169405169411169410
But anyways, I just finished a workbench in the new garage(guess what I made it out of), so now I can get back to work on the steering system.

[andrewb70]

I bet some woodworker would give you good money for that wood.

Andrew

[dhutton]

Seal the end grain or it will split and check imho.

Don

[jetmech442]

I bet some woodworker would give you good money for that wood.

Andrew

Like enough for a supercharger? I've got some 4" thick slabs about 7 feet long that could make ahelluva table.

[jetmech442]

Seal the end grain or it will split and check imho.

DonThanks Don, I sealed it with a Primer /latex combo then the rest of a can of mothers carnuba wax [emoji4]. Stacked, stickered and covered. I'm definitely new and excited to make some beautiful furniture so any other advice is more than welcome.

[BMR Sales]

Hey conundrum, thanks for the link! i'm going to spend my usually uneventful friday going through it. Right after my last post my wife decided we needed to move. So we had to complete all the open projects in the house, re-install the brakes I just pulled off the cutlass, well the house, buy the new house. All that went cool, then the 3rd week in the new house a 70 foot white Burr Oak came down for no reason. Quoted 2K to turn it into firewood-but what a waste! so I did the only responsible thing, which is buy an 800 dollar chainsaw and an alaskan chainsaw mill. So now I have more lumber than I know what to do with lol.

169405169411169410
But anyways, I just finished a workbench in the new garage(guess what I made it out of), so now I can get back to work on the steering system.

Great Idea! Since mine fell on my Mancave/Workshop, I had to hire people to remove it. looks like our Trees were related in Root-Rot

https://i.hmjimg.com/images/2019/11/21/73210208_10220059114178576_1458912820363001856_o.j pg

https://i.hmjimg.com/images/2019/10/24/20191023_0917161-002.jpg (https://www.hostmyjunk.com/image/BBaHG)

[jetmech442]

Great Idea! Since mine fell on my Mancave/Workshop, I had to hire people to remove it. looks like our Trees were related in Root-Rot

https://i.hmjimg.com/images/2019/11/21/73210208_10220059114178576_1458912820363001856_o.j pg

https://i.hmjimg.com/images/2019/10/24/20191023_0917161-002.jpg (https://www.hostmyjunk.com/image/BBaHG)
Oh man!Looks like you built you man cave like you build your suspension pieces-tougher than the right side of a rhino's @$$! Looks like you coulda saved some money and left it there to use a an engine hoist lolol.

[Fab]

Hi All,

I've find this thread googoling for EPS conversion steering...I have Impreza WRX where I change from hydraulic to EPS the steering system. I made this using a Corsa C EPS...this sistem work "well" with street tire (this car is only 1250kg) but whit semislick tires the Corsa EPS doesn't work. The system go il failsafe when the car is stopped (like parking) and during cruise isn't 100% reliable....
This because the motor is small...
Reading this good thread (and the thread of MWP on other forum) seems the BRZ/FR EPS is good. The BRZ/FR is little bit lighter then my Impreza, around 1180kg...but seem the system work well in a Cutlass that weight around 1500kg...
Is this right?

I looking for use also a RAV4 EPS that have a very big motor, and it's used on a car of 1500kg...I think the system is tested/checked for more kilos....

Thanks for reply, and apologise for my bad writing!

Bye
Fabry

[Vimes]

Glad I found this thread, have it saved. I'm doing the same thing and have a 2010 Corolla box in my shop right now. Haven't started playing with it yet though. I went with the Corolla because it's a FWD unit, and I figure turning wheels under power is going to be a lot harder to do than turning wheels that are just along for the ride. And as Toyota uses a lot of the same parts across the board, your FRS probably has the same unit, just with different programming for a RWD car.

So, any luck with getting this working? I tried reading that canbus programming language, and I can't wrap my head around it. I have a hard enough time with English and it's supposed to be my primary language.

[jetmech442]

Hi All,

I've find this thread googoling for EPS conversion steering...I have Impreza WRX where I change from hydraulic to EPS the steering system. I made this using a Corsa C EPS...this sistem work "well" with street tire (this car is only 1250kg) but whit semislick tires the Corsa EPS doesn't work. The system go il failsafe when the car is stopped (like parking) and during cruise isn't 100% reliable....
This because the motor is small...
Reading this good thread (and the thread of MWP on other forum) seems the BRZ/FR EPS is good. The BRZ/FR is little bit lighter then my Impreza, around 1180kg...but seem the system work well in a Cutlass that weight around 1500kg...
Is this right?

I looking for use also a RAV4 EPS that have a very big motor, and it's used on a car of 1500kg...I think the system is tested/checked for more kilos....

Thanks for reply, and apologise for my bad writing!

Bye
Fabry

Hey Fabry,
I'm glad the thread is of some use to you. I admit a lot of the knowledge was pieced together from the threads you subaru guys posted. :) The reason behind the FRS unit and not something out of an SUV, was that I thought the programming/ calibration of the FRS was great. My hope was to see how well that stock system would work if I fed it factory style inputs.

It hasn't been installed yet. I realized that the arduino Can-shield isn't fast enough to collect the entire message and broadcast what I need. I switched over to a PiCan2 /Raspberry PI setup and then got buried by work/home/ and now quarantine.

As far as your failsafe issue, you may try to feed some cool air to the motor controller and/or the motor itself. I'm not sure if the motor temperature is actually measured or if it is predicted based on usage. If your under-dash temps are hotter than the Corolla then that might explain it...

[jetmech442]

Glad I found this thread, have it saved. I'm doing the same thing and have a 2010 Corolla box in my shop right now. Haven't started playing with it yet though. I went with the Corolla because it's a FWD unit, and I figure turning wheels under power is going to be a lot harder to do than turning wheels that are just along for the ride. And as Toyota uses a lot of the same parts across the board, your FRS probably has the same unit, just with different programming for a RWD car.

So, any luck with getting this working? I tried reading that canbus programming language, and I can't wrap my head around it. I have a hard enough time with English and it's supposed to be my primary language.

There's a ton to be learned from the arduino/Can-shield setup. It may be possible to get just that system to work live on your car, but I don't think I'm good enough to figure it out. The Raspbertry Pi/ PiCan2 seems promising, but I got bogged down trying to get the lcd screen to work so I wouldn't have to bring my laptop in the car each time I wanted to record data of my buddies FRS.

If you figure anything or have any questions let me know and maybe one of us will crack this thing.

[David Pozzi]

I made this a Sticky so it can be more easily found. It has some great tech, (except for the Tree part)...

[Vimes]

There's a ton to be learned from the arduino/Can-shield setup. It may be possible to get just that system to work live on your car, but I don't think I'm good enough to figure it out. The Raspbertry Pi/ PiCan2 seems promising, but I got bogged down trying to get the lcd screen to work so I wouldn't have to bring my laptop in the car each time I wanted to record data of my buddies FRS.

If you figure anything or have any questions let me know and maybe one of us will crack this thing.

I'm going to give this a shot.

http://myraceshop.com/Legacy_Power_Steering_Controlle.html

Unfortunately it will be a while before I can post results, like a year or so. Good luck on this, cracking the CAN would allow people to use the more modern power racks.

[jetmech442]

I'm going to give this a shot.

http://myraceshop.com/Legacy_Power_Steering_Controlle.html

Unfortunately it will be a while before I can post results, like a year or so. Good luck on this, cracking the CAN would allow people to use the more modern power racks.

Looks like that product controls an electric fluid pump (preferred from an MR2) base don vehicle speed only. Its probably the easiest way for variable assist, but is not the same as an electric-only(no fluid) power assist. Looks like Fleabay has the MR2 pump for around 300$ so not cheap for an older, used pump, and if it can be mounted to the engine or frame easily then it would probly be a pretty good option. I've go a lot of open projects open at the moment, mostly car, but also some more house/family stuff(yes some more trees had to be slabbed lol), but I plan to pick this back up pretty aggressive as winter nears.

- - - Updated - - -


I made this a Sticky so it can be more easily found. It has some great tech, (except for the Tree part)...

Honestly David, I'm honored. I know that probly sounds silly, but I'm more motivated now to be able to add a "Solved" tag at the end of the title now more than ever.

[macostech]

Objective:
Integrate a 2014 toyota FRS Electric Power Steering(EPS) system into my ’69 Cutlass, allowing the steering ECU to vary assist levels based on vehicle speed, mph, torque input and steering angle.
Motivation:
This car has always been more of a journey-not-destination thing for me. And as such I'm at the point where the Grand Jeep Cherokee steering box and stock LQ9 pump with Turn One reducer valve just isn't doing it for me anymore. I'd love to give Turn one or Lee a thousand dollars, but I don't have that kind of money, plus getting the known easy solution isn't my thing for some reason, I guess I just like the challenge of integration. I've got a build thread over on LS1tech https://ls1tech.com/forums/conversions-hybrids/1418981-lq9-t56-1969-old-442-conversion-2.html where I started, if anyone is interested.
I’ll state here that I am not a CANBUS expert, I’m merely a novice with an Arduino and CANshield that is pulling data from multiple sources and doing his best to put the pieces of the puzzle together. I love discussion and feedback and always am eager to learn from the awesome people here at protouring and my goal is to help others on their path.

So this project, as most of mine do, stands on the shoulders of giants. Below is where the majority of my knowledge came from-people freely sharing what they've learned for the love of cars. Hopefully someone someday will use this to do something even cooler(also kind of assuming that this project is successful in the end, lol). I'm a thermal engineer by day, which means I know lots about temperature and not lots about CANBUS or electronics. I dabble in Arduino, but most of what I do with it is based re-purposing other peoples code-guess I'm trying to say that I don't think this is out of anyone’s ability if they have the desire.
CAN equipment
I am using a Seedtudio v1.2. there is a new version out since I’ve had mine for a while, but I don’ think it does anything significantly different.
http://wiki.seeedstudio.com/CAN-BUS_Shield_V2.0/
This stacks on top of a basic Arduino Uno. And plugs into the car using the OBD-connector and DB9-connector that Seedstudio sells.
There are tons of ways though. I have run into a lot of people who been very successful with the SocketCan, which utilizes Linux and a direct port into a laptop. Since I plan to convert my current rpm/VSS into a CAN message and broadcast, the arduino seemed like the best path for me though.


Different ways to get assist from your unit:
There's several way to tackle this. Some just hook up power(no CANBUS) and have full assist at all times. some hook up what I assume is a PWM circuit and dial in what they want which is the same across all speed levels. My contribution will be(hopefully) adding in the CANBUS communication to the toyota steering ECU and letting the algorithms control assist based on VSS, rpm, steering torque and steering angle. So, as we all know, OEM's spend quite a bit of time tuning the EPS for steering feel. I assume that they spend more time on "feel" for the FRS than they do a Yaris or Prius, so I scooped up a 2014 FRS EPS with ECU. I plan to convert the RPM and VSS from the LS1b GM ECU into CANBUS signals and let it do its thing. That's the plan at least!!!! I'll post more as I make progress, hoping this won't take too long. Below is a quick pic of the EPS in the back seat.

159158

Knowledge sources
So to start with, Waid302 posts in vintage mustang what I consider the best starting place for understanding the project. Its a huge read, but worth it. He provided a ton of pics and diagrams which are enormously helpful-especially if you just want a Prius unit to give full assist.
https://forums.vintage-mustang.com/mod-custom-forum/1008722-electric-steering-fail-safe-no-ebay-module-needed-pics-videos.html

dobrostand has a thread in the same forum detailing specifics of the driveability of the system, which led me to believe this might just be worth it.
https://forums.vintage-mustang.com/mod-custom-forum/866473-rubber-hits-road-electric-power-steering-epas-drivability.html

Also on Vintage mustang- Andrewb70, originally known to me as project_GatToGo and who is a huge inspiration to my build, details the specifics of installing in his cougar, but also takes it a step further at getting his dominator to provide VSS to the prius ECU for variable assist. Also pretty much lays out all the details for fab/install that you could ever want,assuming you are skilled with a whizwheelofdeath, lol. Additionaly, he brings up the detail that the ECU needs to run a centering calibration if any of the parts have been mismatched, replaced or otherwise tampered with.
https://forums.vintage-mustang.com/mod-custom-forum/1147680-67-cougar-toyota-prius-epas-install.html


I believe those threads should be sufficient to get a Yaris/Prius or Saturn Vue EPS installed and supplying full assist. The nice thing about the Japanese EPS systems is that they supply full assist in default mode if no CAN signals are received. The Vue uses an aftermarket box that dials in "fixed" assist. Lots of people feel fine with that, I might also, but I like being a nerd so I'm looking into supplying the proper CANBus channels.

The next major source of info is some enginerds at the University of Tulsa has published some papers on CANBUS hacking. They provide the engine speed HexID (2C4), and a version of the vehicle speed(610). I'm not entirely sure ID 610 is what the Steering ECU is looking for based on some info I find down below, but we'll get to that. This paper is awesome at showing the methodology for grabbing canbus data and then deciphering it. These guys are the real deal.
http://tucrrc.utulsa.edu/ToyotaCAN.html

Then, Fabio published a youtube video of a camera synced with CANBus data in a 2010 Camry. This allows you to see the structure of the CAN message for RPM (2C4) and Vehicle speed(both “B4” and “610”).
https://www.youtube.com/watch?v=KHRPmjwXF1U


I would say this entire paper is a must-read and is also the most indepth of anything I've found so far. It talks about great details for ford and Toyota steering/braking Canbus interfaces. Specifically, page 41 shows how ID B4(vehicle speed) is not only structured in the way of what bits mean what, but also how to combine them, and the importance of the "checksum" bit at the end of the messages for toyota. If the different nodes don't see the checksum, they will ignore the message all together. I don't know if this is all OEM's or just Toyota.
http://illmatics.com/car_hacking.pdf


Sorry to jump into this thread asking something not 100% related but I can't find a way to message you privately.
I own a FRS and I'm trying to find a way to make the steering heavier (i.e. reduce power steering) and I think the only way to do so is to trick the controller by modifying the signal coming from the torque sensor so that it will "think" you're applying less torque to the steering wheel therefore it should reduce power steering force.
Do you know how can I achieve this?
I've read that the torque sensor 4 wires connect to the 12 pin connector but what do I have to do with these?
Put a resistor in line with a specific cable out of this connector? But I don't think it's as easy as that, is it?
Thank you so much in advance.

[jetmech442]

Macostech,
For what you are describing, you would need to Intercept the CAN signal as it is broadcast from the source(before it reaches anywhere else on the bus).Then your CAN controller would need to modify that signals contents to what you want(lower torque), and then let that modified signal carry on the the rest of the car. I think 0x370 is the correct channel name for steering torque(had to go back to pg 2 to refresh my memory).

So to clarify, you would need to splice into the can wires that come out of the steering ECU and insert a CAN controller. based on my experience above, I would skip the arduino, and go straight to the raspberry pi. You will need to listen for 0x370 and experiment with your car to see how the signal reacts. This thread should have enough to get you started, but it's definitely not easy. If you make any headway let me know for sure, we'll need all the we can get to solve this.

[andrewb70]

Would putting wider tires accomplish the same thing?

[Vimes]

Looks like that product controls an electric fluid pump (preferred from an MR2) base don vehicle speed only. Its probably the easiest way for variable assist, but is not the same as an electric-only(no fluid) power assist.

Yeah, but I figure the controller won't know whether it's working with an MR2 pump or an electric drive unit. I've since picked up a 2010ish Corolla electric column drive power steering unit. It looks compact enough to fit under my dash invisibly, and power assist can be controlled with a POT feeding an input on the control board. I'll be using this board to control that input instead of a dash-mounted POT. Going to be a bit longer than I thought on getting it in though as I've only just in the last week (FINALLY!!!) gotten a shop to commit to my project but he won't have an opening in his shop until April 2021 at the earliest.

[stab6902]

Yeah, but I figure the controller won't know whether it's working with an MR2 pump or an electric drive unit. I've since picked up a 2010ish Corolla electric column drive power steering unit. It looks compact enough to fit under my dash invisibly, and power assist can be controlled with a POT feeding an input on the control board. I'll be using this board to control that input instead of a dash-mounted POT. Going to be a bit longer than I thought on getting it in though as I've only just in the last week (FINALLY!!!) gotten a shop to commit to my project but he won't have an opening in his shop until April 2021 at the earliest.

Since you'll be manipulating steering feel based on vehicle speed, make sure you have a good reference vehicle speed signal, or at least consider what will happen in highly dynamic circumstances (i.e. wheel lockup or wheelspin). You need more than just one wheel speed sensor as an input to do this correctly.

Also, for anyone considering these "small car" column mount EPS systems for pro-touring cars, remember they're designed for ~3000 lb cars with skinny all season tires. I'd have durability concerns if I was considering putting one in a 4000 lb pro-touring car with sticky 275+ tires, especially if you're thinking autocross or track use.

[jetmech442]

Since you'll be manipulating steering feel based on vehicle speed, make sure you have a good reference vehicle speed signal, or at least consider what will happen in highly dynamic circumstances (i.e. wheel lockup or wheelspin). You need more than just one wheel speed sensor as an input to do this correctly.

Also, for anyone considering these "small car" column mount EPS systems for pro-touring cars, remember they're designed for ~3000 lb cars with skinny all season tires. I'd have durability concerns if I was considering putting one in a 4000 lb pro-touring car with sticky 275+ tires, especially if you're thinking autocross or track use.

That's a great point Stab about the heat generation(and the speed sensor too lol). Since our cars are heavier and being used for HPDE events and coupled with generally higher ambient(no AC and close to the firewall) this is something that would need to be tested. I'd be surprised if one of the CAN channels didn't report temperature for de-rate purposes.

After I got it working properly, I had planned on thermocoupling the SCU and motor and boxing in my test bench so I could cyclically load it to represent a track session for 20 minutes. My life is overloaded right now, so anyone working on this would be wise to try and duct some airflow to the motor or have a way to actively monitor temps so it doesn't shut down unexpectedly during a session.

[Vimes]

Also, for anyone considering these "small car" column mount EPS systems for pro-touring cars, remember they're designed for ~3000 lb cars with skinny all season tires. I'd have durability concerns if I was considering putting one in a 4000 lb pro-touring car with sticky 275+ tires, especially if you're thinking autocross or track use.

I actually did think about that. My thinking is that the Corolla has to be able to carry its own 2800lbs, tow 1500lbs and carry as many as 5 adults at 750ish lbs. In addition, the steering system has to be able to force the drive axles to turn even under WOT while carrying this weight. And this is where I think the key is - while my truck will weigh more, the front wheels will be unpowered. I figure that no engine torque on the front wheels should allow the EPS to control a slightly heavier vehicle. I wouldn't consider this on a heavy truck, but mine weighs right at 4000lbs stock with no driver and a full tank of gas, and I've not started weight reduction at all. When it's all said and done, my target weight for my truck will be 3600lbs. Dropping 400lbs off the truck shouldn't be too much of a deal with the work I have planned, so the EPS should be able to handle it.

[stab6902]

I think tire size/compound, vehicle weight, and lateral accel are much bigger contributors to EPS load than torque steer. Lots of people are having issues with their Toyota FRS/BRZ's EPS systems on the track with only the addition of stickier tires: https://www.ft86club.com/forums/showthread.php?t=90718

I'm not trying to pick on any particular OEM/supplier either, as lots of platforms have similar issues -
BMW: https://f80.bimmerpost.com/forums/showthread.php?t=1383918
VW: https://www.vwvortex.com/threads/electric-power-steering-overheating-need-advice.4302388/
Ford: https://www.focusrs.org/threads/powersteering-overheating-track.117784/
etc.

Overall I like EPS, especially the newer rack mount units, and I think they're getting better every year and generally have fewer issues than hydraulic power steering systems. Since EPS units come from more modern and highly integrated cars, however, you have to be extra careful to get the electrical integration and sizing right. I think you guys are on the right track and wish you luck in getting it figured out!

[67King]

Tangential, but apparently the early Mustang stuff isn't as integrated into the PCM. And the software is not really secure. I am way out of my specialty here, but talking to a friend from when I was at Ford (engine development), and he is relaying some stories to me, this being one. I believe it was 2011 for the first year, seems we never had to protect for the PS pump when I was working on the Coyote.

[jetmech442]

Tangential, but apparently the early Mustang stuff isn't as integrated into the PCM. And the software is not really secure. I am way out of my specialty here, but talking to a friend from when I was at Ford (engine development), and he is relaying some stories to me, this being one. I believe it was 2011 for the first year, seems we never had to protect for the PS pump when I was working on the Coyote.


Hey 67king that's really good info! I have a feeling that 'yota put more effort into security than others at the time due to the"run away pedal" in the... Prius I think? I've never had the opportunity to drive a mustang but I've ridden in them. Are you a fan of the electric steering feel in them? They'd obviously be better suited to the heft of the a bodies so someone may find that a good option and much easier to integrate it sounds like.

[67King]

Who knows? This was going on when headcount was horrendously low. I started in 2000, left in the middle of 2007. There were separations in October, 2006, and then again a few months after I left. By then, 50% of the engineering staff had left from the time I started. So it may have just been an accepted risk given that there wouldn't be any risk to emissions or whatever.

Sadly, I have still not driven one of the Mustangs. So I dont' know. I can just share that my father-in-law, who works for Porsche at the PEC in Atlanta as a driver, absolutely loves the GT350......and he hates everything that isn't made by Porsche. He loves everything about it. I'd love to try it, but I don't have the skills with PCM stuff that you do. I'd like to just get my car a little further along, too!

[v413nc3]

I'm currently working on getting this to function and if you want to share some information I'd be happy to report back. I am using a significantly more robust system on the CAN side (all MoTeC) so I can do some capture, test, and send that is a little easier to work with and fast enough that nothing gets dropped. As far as I know the checksums from the ECU out to the CAN bus don't exist for at least RPM. Vehicle speed however I think may come from the ABS system so I'm not sure just yet if it does or does not have them.

[Fab]

Hey Fabry,
I'm glad the thread is of some use to you. I admit a lot of the knowledge was pieced together from the threads you subaru guys posted. :) The reason behind the FRS unit and not something out of an SUV, was that I thought the programming/ calibration of the FRS was great. My hope was to see how well that stock system would work if I fed it factory style inputs.

It hasn't been installed yet. I realized that the arduino Can-shield isn't fast enough to collect the entire message and broadcast what I need. I switched over to a PiCan2 /Raspberry PI setup and then got buried by work/home/ and now quarantine.

As far as your failsafe issue, you may try to feed some cool air to the motor controller and/or the motor itself. I'm not sure if the motor temperature is actually measured or if it is predicted based on usage. If your under-dash temps are hotter than the Corolla then that might explain it...


Hi Jetmech,

sorry for this long time to reply, but i'm very busy by my job...I don't have time for power on my Subie!

I don't remenber if in this thread was mentioned, but the FRS\BRZ EPS you can find a issue on the elastic jonier (copuler) inside the motor (join the motor with the screw) you can find info about this on this link:

https://www.ft86club.com/forums/showthread.php?t=133606


So, about my corsa EPS, I think the limit is the motor and the ECU...DC Electronics from UK use my system with diffeert controllwer, controller made by DC, but for what I've seen this system is for lightwheight car...

So, I'v buyed a RAV4 EPS and a FRS\BRZ EPS...both work great whitout CAN-BUS signals, as mentioned by someone they work in a FailSafe mode...yesterday nigth I've tried to keep some info about CAN-BUS signal for this EPS...and I've find some docs that maybe can be usefull...or I hope so!


Attacched you can find the docs!

183967
183968
183969
183970
183971
183972



So, for now I tested the EPS only on the bech...need to finish some things on my Subie and then try to mount...I like to have a collapsible sistem like the oem for safety...and if is possible also the tilt system for more confort...(but I think a fixed system is more strong)

Bye
Fabry

[jetmech442]

I'm currently working on getting this to function and if you want to share some information I'd be happy to report back. I am using a significantly more robust system on the CAN side (all MoTeC) so I can do some capture, test, and send that is a little easier to work with and fast enough that nothing gets dropped. As far as I know the checksums from the ECU out to the CAN bus don't exist for at least RPM. Vehicle speed however I think may come from the ABS system so I'm not sure just yet if it does or does not have them.

We've chatted through PM a couple times and it's got me thinking. I've made some assumptions about the system, that at this point aren't backed up. The biggest one is that "assist vs torque" is reduced at vehicle speed. The system clearly has a torque sensor, perhaps it provides assist based on need. I'm not sure, but will try to figure that out. Theres a new guy at work with an FRS and once the snow melts he'll let me grab some more data, hopefully I can get the PiCan2 properly to capture the entire channel.

Additionally, a guy contacted me through WheelWell about this. he's got what looks to be a wicked maroon-red time attack Impreza with 315's on all corners and the frs column spliced in using "default" mode. he reported that the wheel self centered nicely, and that the feel was really good. Again, I'm wondering if the FRS still lets the torque sensor function properly in default mode, and perhaps the prius system either doesn't have a torque sensor or doesn't let it function in default. He didn't use techstream either to re-zero wheel position.

Now, maybe his steering forces are overcoming what the SCU is trying to do since he has some pretty high camber/caster and ultra wide tires. This would be great since most of us are trying to install it on vehicles with similar setups(big tires, high caster...etc).

He did mention the SCU getting extremely hot, as @stab6902 also mentioned. Luckily thermal performance and simulation is my actual day job, so I will be able to evaluate a mitigation prior to install.

[andrewb70]

Still following along. Good stuff Scott!

Andrew

[Vimes]

I've come across a potential electrohydraulic system, for anyone who is interested. MOPAR part number 52124998AJ is an electrohydraulic power steering pump, and came on 2011-2013 Dodge Durangos and Jeep Grand Cherokees. This is designed from the factory to manage a 5000lb RWD or AWD vehicle using rack and pinion steering, so it's going to easily handle just about anything this forum could throw at it. I can see this being used by someone who needs the space where the power steering pump is located to run

https://static1.pt-content.com/images/pt/2021/05/23152b357e0160950d60b0085055a16f-1.png

I still plan to use a Toyota column drive unit for my project, which is already sitting on my shelf, but I figured someone else might be able to make use of this.

[TanMan]

Forgive me if the question has been answered already...

With all of the engine swaps and challenges with front accessory drives, why don’t we see more electric power steering pumps around? The traditional belt driven pump probably doesn’t steal a ton of HP from the engine, but it seems like a dedicated electric pump like the one above could be mounted most anywhere and supply hydraulic fluid pressure to a steering box without interacting with the engine. I see that as a benefit but it is clearly not the norm with most steering box vehicles.

I imagine the current draw is significant, and I realize that puts a greater load on the alternator (thus stealing HP). But times of maximum steering effort probably don’t occur when the engine is heavily loaded. I’m thinking diving into the turn while you decelerate is a time when you can give away some HP.

Or is there a safety/reliability factor, and you wouldn’t want to rely on an electrical system that could leave you [all of the sudden] without power steering? Or are they just expensive to manufacture?

Side note- it sure would make an LT swap into a muscle car more desirable.

[Vimes]

Forgive me if the question has been answered already...

With all of the engine swaps and challenges with front accessory drives, why don’t we see more electric power steering pumps around? The traditional belt driven pump probably doesn’t steal a ton of HP from the engine, but it seems like a dedicated electric pump like the one above could be mounted most anywhere and supply hydraulic fluid pressure to a steering box without interacting with the engine. I see that as a benefit but it is clearly not the norm with most steering box vehicles.

I imagine the current draw is significant, and I realize that puts a greater load on the alternator (thus stealing HP). But times of maximum steering effort probably don’t occur when the engine is heavily loaded. I’m thinking diving into the turn while you decelerate is a time when you can give away some HP.

Or is there a safety/reliability factor, and you wouldn’t want to rely on an electrical system that could leave you [all of the sudden] without power steering? Or are they just expensive to manufacture?

Side note- it sure would make an LT swap into a muscle car more desirable.

The biggest reason people don't do it is electrohydraulic PS pumps are damned near impossible to find, and the ones that have been available have been for tiny cars like the Toyota MR2. And, when the OEMs started going electric they also started using CAN data to control them, which if you've read this thread you've seen the CAN is not the easiest thing to defeat/work around. Plus, the assist is either on/off, or uses a potentiometer to manually adjust but that either requires you to keep fiddling with it, or find one setting that sorta works and you just work around it. Then there's the aftermarket systems, which cost a grand or more and don't have the greatest reviews.

I've found a board (http://myraceshop.com/Legacy_Power_Steering_Controlle.html) that may let me get speed-adjusted assist from my Toyota unit, but at the rate I'm going it'll be a couple of years before I find out. As of now, I have no idea if it'll work or not.

[TanMan]

That all makes sense. But good on you for taking the time to experiment, and potentially make it work. When things get too deep with software and CAN bus control, I get spooked. So I’d love to hear how it goes.

[Fab]

Hi there,

after some time i restart to work to install my BRZ EPS in to my Subaru replacing the Corsa C EPS I still use. After some research, trying to understand better the CAN-BUS communication for make my own controller, I decide to use a controller on the market, with POT for decide the assist...
After I had the controller, I decide to "look" inside using a sniffer, and I find the data attached in the TXT files.
Hope this can help someone trying to use this EPS!



193824

193822

193823



Byebye
Fabrizio

[neongreen]

Has anybody made any further progress on CAN messages for EPS?