Enter your username:
Do you want to login or register?
  • Forgot your password?

    Login / Register



    Results 1 to 20 of 20
    1. #1
      Join Date
      Dec 2009
      Posts
      261
      Country Flag: Canada

      Magnetorheological Shocks, some parts and ideas

      I was hoping to actually be able to post doing this to my car this winter. But it seems that I'll be heading out to the hinterlands for work this year. So rather than keep ahold of the idea, I figure I'll put it out there for any whom are curious.

      I've been looking at Magnetorheological shock absorbers for quite a while. They represent a number of interesting possibilities to mess with transients, or hell, even just to make spring changes that much easier.

      The biggest barrier to entry has been the price of the shocks themselves however, as the corvette mag shocks and similar mag vehicles start at the 5-600 range per corner. Now while that was about what I was looking to spend in the non mag range (say afco doubles), it's a little steep just to test the envelope of damping I can get away with in a given shock.

      Enter cadillac, who decided that the escalade EXT really needed to have a magnetorheological shock system, and luckily, that they could sell it much cheaper. ACDelco sells them for a scant 270 a piece for rear shocks:
      https://www.acdelcodirect.com/produc...truts/25871432
      That's a cheap entry point to see whether they'll have the damping range that can suit the critical damping ranges that we'll be after. Plus a large truck is hopefully already higher in terms of force rates than a typical passenger car, given the weights involved.
      Finding out what we can get away with will take a basic shock dyno, something this:
      http://www.circletrack.com/howto/828...ng_shock_dyno/
      Though finding out an accurate initial value could prove trying. Alternatively you could build something like this:
      http://honda-tech.com/showthread.php?t=2325055
      Though that is mainly overkill, since we will be able to vary the velocity response at will, our main concern is seeing just how much resistance the shock can generate between 0-5v.



      That brings us to the next big problem, controlling the shocks themselves. They need a 5v pulse width modulated signal which controls the damp rate of the shock. Increasing the duty cycle increases the damping rate, luckily the shock response is apparently pretty linear, so there shouldn't be any annoying peaks or dips in shock response to program around. As for what to program, I'm unsure on that point.

      I've been thinking of using something along the lines of an Arduino, which has the pwm capabilities, but can only supply about 40ma of current. Now I think that's where you'd use say a tip120 transistor to act as a relay and provide the dampers with the current they need. I believe at that point you'd need an additional 5v source to run thorough the tip120, not really sure though. An arduino leonardo would provide more than enough output and input pins, I'm just a little concerned about the duty cycle and general hardiness of these. They seem to be pretty tough, it's just not something you'd want to fail at an unexpected moment.
      http://www.adafruit.com/products/849#Description
      http://www.digikey.ca/product-detail...120-ND/1052441
      http://www.ebay.com/itm/DC-DC-Conver...-/171002548980

      As for inputs the sky is the limit really, you could pretty much add just about anything you want since an arduino like the leonardo has a dozen input pins. start with the requisite potentiometers on each corner to measure suspension movement (gm stock ones would likely work, but you could put something less robust up and out of the way with pushrods) and add whatever you want to use as a variable for the damping rate.

      Here's a Delphi presentation on the subject:
      https://docs.google.com/viewer?a=v&q...-FRxUrbAXaEb6g
      and an interesting thesis on mag shocks which goes into some of the algorithms and even shock design:
      https://docs.google.com/viewer?a=v&q...h_5LlqNjmsKGGA
      Who knows, maybe this'll help someone else with too much time on their hands
      Last edited by ace_xp2; 10-15-2013 at 11:56 AM. Reason: spacing


    2. #2
      Join Date
      May 2010
      Location
      kitchener,Ontario,Canada
      Posts
      2,336
      Country Flag: Canada
      Holy crap you've done some home work there eh ? What's your car and where in Canada are you... it is the second largest country in the world and has more square feet of awesomeness....sorry Americans..its from a beer commercial. bit of a joke lol
      Ryan Austin

    3. #3
      Join Date
      Oct 2012
      Posts
      434
      Country Flag: United States
      I can tell by this just scanning this post that it deserves more focus than I can generate right now. I will revisit in the morning, it seems to be the beginning of a great thread.

    4. #4
      Join Date
      Dec 2009
      Posts
      261
      Country Flag: Canada
      Quote Originally Posted by raustinss View Post
      Holy crap you've done some home work there eh ? What's your car and where in Canada are you... it is the second largest country in the world and has more square feet of awesomeness....sorry Americans..its from a beer commercial. bit of a joke lol
      Ryan Austin
      I've been working on it for a bit, yeah
      I'm building an LS swapped 65 mustang in the Okanagan, BC.

    5. #5
      Join Date
      May 2010
      Location
      kitchener,Ontario,Canada
      Posts
      2,336
      Country Flag: Canada
      Well please keep this updated...I'd love to run mag shocks. Good luck with the half breed ls mustang...good news as both ford and Chevy guys will like you. As for the dodge boys...well...too bad lol

    6. #6
      Join Date
      Sep 2008
      Location
      Jacksonville, Florida
      Posts
      629
      Country Flag: United States
      Sounds very interesting, but without a 4 post shaker and a team of engineers, I think programming would be daunting. I'm very interested to see if you could get this working.
      I get confused with double adjustable coil overs. Cant tell you how many times in a row I reset back to zero and count clicks only to lose track and back to zero....
      Craig Scholl
      CJD Automotive, LLC
      Jacksonville, Florida
      904-400-1802
      www.cjdautomotive.com

      "I own a Mopar, I already know it won't be in stock, won't ship tomorrow, and won't fit without modification."

    7. #7
      Join Date
      Nov 2012
      Location
      Sacramento, CA
      Posts
      1,918
      Country Flag: United States
      Quote Originally Posted by sccacuda View Post
      Sounds very interesting, but without a 4 post shaker and a team of engineers, I think programming would be daunting. I'm very interested to see if you could get this working.
      I get confused with double adjustable coil overs. Cant tell you how many times in a row I reset back to zero and count clicks only to lose track and back to zero....


      Haha ! That's funny because it is way too true!

      Craig, all my race teams write down the click changes in a super detailed run sheet EVERY session ... and we still get off sometimes.
      If a car does something different than what it was supposed to ... the first things I check are the tire durometer and recount the shock settings.

      I can't tell you how many times the problem has been the shock settings are off from human error.


    8. #8
      Join Date
      Dec 2005
      Location
      Birmingham, AL
      Posts
      3,356
      Country Flag: United States
      While this has an extremely high cool factor, I didnt get enough of my family's engineering genes to even consider this undertaking. I applaud you for going for it. Keep us posted.
      Stephen

    9. #9
      Join Date
      Dec 2010
      Posts
      169
      Quote Originally Posted by ace_xp2 View Post
      I was hoping to actually be able to post doing this to my car this winter. But it seems that I'll be heading out to the hinterlands for work this year. So rather than keep ahold of the idea, I figure I'll put it out there for any whom are curious.

      I've been looking at Magnetorheological shock absorbers for quite a while. They represent a number of interesting possibilities to mess with transients, or hell, even just to make spring changes that much easier.

      The biggest barrier to entry has been the price of the shocks themselves however, as the corvette mag shocks and similar mag vehicles start at the 5-600 range per corner. Now while that was about what I was looking to spend in the non mag range (say afco doubles), it's a little steep just to test the envelope of damping I can get away with in a given shock.

      Enter cadillac, who decided that the escalade EXT really needed to have a magnetorheological shock system, and luckily, that they could sell it much cheaper. ACDelco sells them for a scant 270 a piece for rear shocks:
      https://www.acdelcodirect.com/produc...truts/25871432
      That's a cheap entry point to see whether they'll have the damping range that can suit the critical damping ranges that we'll be after. Plus a large truck is hopefully already higher in terms of force rates than a typical passenger car, given the weights involved.
      Finding out what we can get away with will take a basic shock dyno, something this:
      http://www.circletrack.com/howto/828...ng_shock_dyno/
      Though finding out an accurate initial value could prove trying. Alternatively you could build something like this:
      http://honda-tech.com/showthread.php?t=2325055
      Though that is mainly overkill, since we will be able to vary the velocity response at will, our main concern is seeing just how much resistance the shock can generate between 0-5v.

      That brings us to the next big problem, controlling the shocks themselves. They need a 5v pulse width modulated signal which controls the damp rate of the shock. Increasing the duty cycle increases the damping rate, luckily the shock response is apparently pretty linear, so there shouldn't be any annoying peaks or dips in shock response to program around. As for what to program, I'm unsure on that point.

      I've been thinking of using something along the lines of an Arduino, which has the pwm capabilities, but can only supply about 40ma of current. Now I think that's where you'd use say a tip120 transistor to act as a relay and provide the dampers with the current they need. I believe at that point you'd need an additional 5v source to run thorough the tip120, not really sure though. An arduino leonardo would provide more than enough output and input pins, I'm just a little concerned about the duty cycle and general hardiness of these. They seem to be pretty tough, it's just not something you'd want to fail at an unexpected moment.
      http://www.adafruit.com/products/849#Description
      http://www.digikey.ca/product-detail...120-ND/1052441
      http://www.ebay.com/itm/DC-DC-Conver...-/171002548980

      As for inputs the sky is the limit really, you could pretty much add just about anything you want since an arduino like the leonardo has a dozen input pins. start with the requisite potentiometers on each corner to measure suspension movement (gm stock ones would likely work, but you could put something less robust up and out of the way with pushrods) and add whatever you want to use as a variable for the damping rate.

      Here's a Delphi presentation on the subject:
      https://docs.google.com/viewer?a=v&q...-FRxUrbAXaEb6g
      and an interesting thesis on mag shocks which goes into some of the algorithms and even shock design:
      https://docs.google.com/viewer?a=v&q...h_5LlqNjmsKGGA
      Who knows, maybe this'll help someone else with too much time on their hands

      You could do this in real time using a PIC Micro and assembly language with a channel for each shock and a look up table for pulse rate depending on your selection for input a 16F-877 would work perfectly and give you enough input and output to have 256 separate shock settings plus an LCD read out showing the current selection. The TIP120 would provide all the drive you would need and you could build a very robust 25 amp 5 v regulated power supply for all four TIP 120's. The PIC is has plenty of non volatile memory so you could have almost unlimited shock settings at all four wheels. You could set up a variable pulse width modulation system to power 1 shock on the dyno and then plot your curves with various pulse widths to determine how to program the shocks for your spring rate, unsprung weight and corner weights. I'll bet you could do the whole electronic package for under $100. I would look at accelerometers at all four corners on the chassis and at the shock location to evaluate the system, they are cheap as dirt now and can be used with the PIC A to D converter. You might even be able to make an automated shock setting system after you have enough data. We built a system to evaluate conventional mechanical shocks at various settings of rebound and bump using a PIC micro and accelerometers writing to a memory card and a simple Visual Basic program to plot the results from changing settings over the same course multiple times. Not really very elegant but cheap and it worked. I use PICs directly because they are cheap and easy to program and assembly language is really fast.

    10. #10
      Join Date
      Dec 2009
      Posts
      261
      Country Flag: Canada
      Could I deadbug a 16f877? or would I have to buy a pic board as well?
      Are you recommending it because of robustness/speed concerns? Or is it a familiarity thing?
      I ask because I'm willing to tackle a new language if necessary, but the "any idiot can" nature of arduino programming is appealing for its ease. And from what I've been reading a 16f877 has 8 D/A channels, so I'd only be able to run 4 pwm channels and 4 potentiometers, right? or can you break those channels up somehow? Could I daisy chain multiple PICs together if I needed more channels?
      The 16f877 runs 5MIPS, so if that's fast enough I'm guessing the 16MIPS the ATmega32u4 on the arduino runs should be enough at least. It's weird realising just how fast 16mhz is when you've become accustomed to 2ghz phones that need multiple cores.

      The plan is to start with a simple velocity to voltage table, so that it looks at the speed and direction the suspension is moving and then chooses the appropriate voltage based on that.That shouldn't take up much memory and I can do the calcs to convert %critical damping to voltage on a seperate pc.
      From there you could add in all kinds modifiers, Like say an input off the launch control button which puts the rear shocks at full stiff until the car hits a given speed. Or in keeping with fairly basic and easy modifiers, you could add a steering potentiometer and use a button press to tell the suspension to stiffen up the right rear for half a second after a left hander has been initiated to take the push out of that one pesky turn on the track.

    11. #11
      Join Date
      Dec 2010
      Posts
      169
      Yes you are right, I am very familiar with PIC devices, all I have is their PICSTART plus programer. I build all my own boards etc They have free software that gives you debugging capabilities and a complete programming environment. I often burn over a hundred program variations while building a project and you can do it in circuit if you want to. Some of my projects are traction control, A shifting system for unlimited light hydroplanes, a ballistic program that calculates scope clicks based on wind, altitude and distance using a wireless wind direction and speed,
      a wireless (5)segment timer for autox, phone based portable trailer alarm, Li battery charger, 30mw infrared laser for timing systems, and many others. When you write programs in assembly a 20 MHZ clock will smoke higher level programs because they are very inefficient. You should use what you are familiar with because it will save time, I thought you were just starting out. Microchip makes processors that have much higher speed and complexity but I like to use the simplest one I can get buy with. I am not a big fan of linear measuring devices because the cheap ones are junk and the good ones are very expensive. I like the accelerometers because they can give you speed and distance with a time base too. The Ardino and Propeller devices are one size fits all and usually have more than you need, that is why I went off on my own with PIC stuff. My 5 segment timing system uses 16 separate processors so it could get very expensive using Ardino stuff. I start with a 20 MHZ crystal and use a 12f 629 to break that down to a 1000 cps and then custom tune that program so that I get 1000 PPS over an 8 hour period or 28,800,000 pulses with an accuracy of +/- 10 pulses for under $2, hard to beat.

    12. #12
      Join Date
      Dec 2009
      Posts
      261
      Country Flag: Canada
      A typical rotary potentiometer is usually not bad though right? I mean, that's how GM does it, they just measure the a arm movement from the rotation around the inner mounting point of the a arm. Although I've heard that they only look at magnitude and direction, maybe the lack of positional accuracy is why.

      I've not done any real programming, which is why the nature of the Arduino programming is enticing (Though I see that PIC programming uses only 35 instructions, so I'd imagine it wouldn't be too hard). Though I'd like to get into PIC encoding, if I use an arduino it'll be easier for people to copy the setup. I mean I don't know how many other people out there would be into etching boards for car stuff.

    13. #13
      Join Date
      Dec 2013
      Posts
      179
      Country Flag: United States
      ..you seem to be doing something similar except.. you can easily get ahold of the suspension controllers for the system. the hard part, that ive run into, is tracking down a BCM with the correct operating system in it. GM used an intuitive tuning program for the MRC during testing. you even get a glimpse of it in their promotional video on youtube. there is a board that can be used to tap into the GMLAN but the manufacturer hasnt exactly been talkative.

      i know.. its a bit of a necro post and a first post.. however.. this thread should NOT die as there is a HUGE potential for teh ability to retrofit these systems.

    14. #14
      Join Date
      Dec 2009
      Posts
      261
      Country Flag: Canada
      I'm not sure that the stock controller would get me more than an custom one would. Don't worry, I'll be getting the set up running sooner or later, I'm just a little ways from my tools and parts at the moment.

    15. #15
      Join Date
      Sep 2013
      Posts
      448
      Country Flag: United States
      you may want to consider optical isolation just to be safe.
      https://www.sparkfun.com/products/784
      or
      https://www.sparkfun.com/products/9118

      also, definately overkill but you could use these if you're ordering from adafruit (btw...that is my favorite vendor, especially if you're new to electronics and programming)
      http://www.adafruit.com/products/355
      regardless of which you use, you will need an external 5v power source though

      if you're really new to all this, i'd suggest the uno r3 for the sole reason that it has a socketed chip. if you wire something incorrectly, it could be an easy $5 part instead of buying a new board:
      http://www.adafruit.com/products/50
      I'd also highly recommend the arduino based solely on one factor. the community. The arduino forums are very helpful to those that are new to programming.

      another option that you may want to consider is backing this project:
      http://www.kickstarter.com/projects/...ld-for-arduino
      if you're making a one off piece and want a nice prototyping area, then this would be a nice solution for you. he's also selling a parts kits for $20 that will allow you to populate 3 boards.


      as for potentiometer, you could possibly use a hall sensor ic and a magnet to get rotational value. reading that on one of the analog pins would give you a pretty durable and inexpensive solution... if you can do some machining.


      Quote Originally Posted by ace_xp2 View Post
      I was hoping to actually be able to post doing this to my car this winter. But it seems that I'll be heading out to the hinterlands for work this year. So rather than keep ahold of the idea, I figure I'll put it out there for any whom are curious.

      I've been looking at Magnetorheological shock absorbers for quite a while. They represent a number of interesting possibilities to mess with transients, or hell, even just to make spring changes that much easier.

      The biggest barrier to entry has been the price of the shocks themselves however, as the corvette mag shocks and similar mag vehicles start at the 5-600 range per corner. Now while that was about what I was looking to spend in the non mag range (say afco doubles), it's a little steep just to test the envelope of damping I can get away with in a given shock.

      Enter cadillac, who decided that the escalade EXT really needed to have a magnetorheological shock system, and luckily, that they could sell it much cheaper. ACDelco sells them for a scant 270 a piece for rear shocks:
      https://www.acdelcodirect.com/produc...truts/25871432
      That's a cheap entry point to see whether they'll have the damping range that can suit the critical damping ranges that we'll be after. Plus a large truck is hopefully already higher in terms of force rates than a typical passenger car, given the weights involved.
      Finding out what we can get away with will take a basic shock dyno, something this:
      http://www.circletrack.com/howto/828...ng_shock_dyno/
      Though finding out an accurate initial value could prove trying. Alternatively you could build something like this:
      http://honda-tech.com/showthread.php?t=2325055
      Though that is mainly overkill, since we will be able to vary the velocity response at will, our main concern is seeing just how much resistance the shock can generate between 0-5v.

      That brings us to the next big problem, controlling the shocks themselves. They need a 5v pulse width modulated signal which controls the damp rate of the shock. Increasing the duty cycle increases the damping rate, luckily the shock response is apparently pretty linear, so there shouldn't be any annoying peaks or dips in shock response to program around. As for what to program, I'm unsure on that point.

      I've been thinking of using something along the lines of an Arduino, which has the pwm capabilities, but can only supply about 40ma of current. Now I think that's where you'd use say a tip120 transistor to act as a relay and provide the dampers with the current they need. I believe at that point you'd need an additional 5v source to run thorough the tip120, not really sure though. An arduino leonardo would provide more than enough output and input pins, I'm just a little concerned about the duty cycle and general hardiness of these. They seem to be pretty tough, it's just not something you'd want to fail at an unexpected moment.
      http://www.adafruit.com/products/849#Description
      http://www.digikey.ca/product-detail...120-ND/1052441
      http://www.ebay.com/itm/DC-DC-Conver...-/171002548980

      As for inputs the sky is the limit really, you could pretty much add just about anything you want since an arduino like the leonardo has a dozen input pins. start with the requisite potentiometers on each corner to measure suspension movement (gm stock ones would likely work, but you could put something less robust up and out of the way with pushrods) and add whatever you want to use as a variable for the damping rate.

      Here's a Delphi presentation on the subject:
      https://docs.google.com/viewer?a=v&q...-FRxUrbAXaEb6g
      and an interesting thesis on mag shocks which goes into some of the algorithms and even shock design:
      https://docs.google.com/viewer?a=v&q...h_5LlqNjmsKGGA
      Who knows, maybe this'll help someone else with too much time on their hands
      Scott
      '66 Chevelle

    16. #16
      Join Date
      Dec 2013
      Posts
      179
      Country Flag: United States
      Quote Originally Posted by ace_xp2 View Post
      I'm not sure that the stock controller would get me more than an custom one would. Don't worry, I'll be getting the set up running sooner or later, I'm just a little ways from my tools and parts at the moment.
      except the stock controller, BCM, and brake controller all have the models/programming that you would base your custom controller on. not to mention its ready made.

    17. #17
      Join Date
      Dec 2013
      Posts
      179
      Country Flag: United States

    18. #18
      Join Date
      Nov 2006
      Location
      Mountain Springs, Texas
      Posts
      4,488
      Country Flag: United States
      Cadillac CTS-V rear shocks are around $280 each at Rock Auto for another possibility.
      1969 Camaro - LSA 6L90E AME sub/IRS
      1957 Buick Estate Wagon
      1959 El Camino - Ironworks frame
      1956 Cameo - full C5 suspension/drivetrain
      1959 Apache Fleetside

    19. #19
      Join Date
      Dec 2013
      Posts
      179
      Country Flag: United States
      Quote Originally Posted by Ron Sutton View Post


      Haha ! That's funny because it is way too true!

      Craig, all my race teams write down the click changes in a super detailed run sheet EVERY session ... and we still get off sometimes.
      If a car does something different than what it was supposed to ... the first things I check are the tire durometer and recount the shock settings.

      I can't tell you how many times the problem has been the shock settings are off from human error.

      thats why a computerized system would be so nice. you log it like you would when tuning an engine except instead of MAP, G/Sec, G/Cyl, IPW, IDC, AFR, Timing, etc, it would be suspension travel, yaw, lateral g, steering angle, wheel speed, rate of travel, rate of accel, dive, etc. the physical model is there, its a matter of putting it to work.

      a base tune like that in the C5 and C6 F55 optioned car is just fine for most cars to be drivable while tuning.

    20. #20
      Join Date
      Sep 2010
      Location
      Benicia, Ca.
      Posts
      4,118
      Country Flag: United States
      There is also the new Magride system from Viking about to be released.... They have even gone to a smartphone app that will allow you to control each corner realtime!

      http://www.nmcadigital.com/2014/01/2...ker-asm-shock/

      Most of you that made it to SEMA probably saw this system on the Viking Camaro in their booth, quite the system! I have already started a wait list for these systems, ever since Magride was released by GM it's been a hit, the technology is simply amazing and the ride/handling it provides by being able to adjust so fast and now, being able to be custom tuned by your Smartphone.... What's next lol!

      Matt
      MCB - Matt's Classic Bowties
      5360 Gateway Plaza Dr.
      Benicia, Ca. 94510
      866-628-8746
      TEXT ONLY: 925-989-9091 (Replied 8-4)
      www.mcbparts.com

      Email: [email protected]
      Like us on Facebook for exclusive offers, info and chat: www.facebook.com/mcbparts
      Hours: Monday to Friday 8am - 4pm Pacific
      We ALWAYS enjoy talking to our customers!
      We have a full time staff to answer ALL Questions, Tech Support or Order Status Tracking!





    Advertise on Pro-Touring.com