AC system design - orifice tybe, expansion valve, and HP draw?

[mikedc]

The orifice tube design is said to draw less HP/TQ off the motor than the expansion valve design. Is it a big difference in practical terms? I assume it's just a matter of the OT design not activating the compressor as much.

Are there any experts on vehicle HVAC out there?

[Tim C]

Been a mechanic/technician whatever you want to call us for 20 years. In my opinion, orifice tube systems were a compromise from a cost or weight standpoint. O tube systems came about in the late 70s, had less bulky parts and less expensive parts than the comparable systems of the day. A 50 cent plastic o tube was way cheaper than the POA or STV systems had before. And cycling the clutch to prevent evap freeze up had the added benefit of reducing hp load and helping fuel mileage on the huge monster ac compressors of the day.

Modern expansion valves and compressors are much smaller and more efficient than the old ways.

I personally believe an expansion valve (txv) system cools better because the valve will react to the evaporator inlet and outlet pressure/temperature and adjust orifice size based on that in relation to system flow (such as high or low rpm, higher mph for better condenser air flow, etc). The result is more consistent and usually colder vent discharge temp and with the compressor usually running 100% duty cycle no hunting idle from the ac kicking on or off. The txv will open or close to prevent evap freeze up on a property sized and designed system. There's usually a backup evap temp sensor that will cycle the clutch if the txv can't keep up such as on a cooler night but it doesn't cycle as often as an o tube system.

Many manufacturers are going back to txv systems these days. My DD 2016 f250 has a txv where my 03 f250 had an o tube. Same cab basically, different hvac boxes obviously but the 03 truck barely would output 55* even new, the 16 puts out 40* at the vent.

[mikedc]

Interesting.

Have the actual expansion valves changed/improved over time? It seems like a pretty simple device.

The longer story is that I'm trying to do a custom-built HVAC setup using a lot of modern OEM parts. The factories have sold a lot of vehicles with OT systems in recent decades but the TXV design seems generally superior to me. And now they are doing variable-displacement compressors too. I'm trying to make sense of it all.

[Tim C]

Basically I think yes, the tvx valves are much smaller in general. Google 60 and 70 gm and ford POA, stv, and vir ac system pics. They were bulky, had capillary and various pressure tubes snaking around. They didn't look good and had more leak points or places to get damaged. The h block expansion valve that pretty much everyone uses replaced all that. It greatly simplified the whole thing and just bolts between both evaporator lines at the evaporator bulkhead through the firewall usually, though some are in the dash.

The o tube was just cheap to manufacture basically. Like I said above it did cycle more so the old compressors of the day didn't spin as much helping the gas mileage some at the time, but modern compressors, let alone variable are more efficient at pumping without as much hp loss. O tubes work, I've got a window ac in my shop that uses one, and I've seen home ac units that had them too, I think that's where they do great actually, in a fixed system with pretty well known unchanging parameters. The condenser and evaporator air flow are known based on fan speeds, the compressor is usually a single or just two speeds. Your car is different, it can go from 600 rpm to 6000 with the compressor along for the ride and both core air flows are variable. That's where the tvx shines as it can change orifice size to compensate. I said before my old 03 f250 o tube system barely hit 55* vent temp. Well after a line leaked and while I had it open I changed the fixed o tube to a variable one. Basically it's a standard tube with a spring loaded check ball that can seat against the outlet of the orifice. The ball seat has a small groove so it doesn't seal completely. At low flow such as idle the spring keeps it seated so the small groove is the orifice, at high rpm high flow the pressure overcomes the spring and the standard orifice size gets used. The result was a 10* drop at idle and a happy driver in NC heat and humidity. I was glad to see Ford went to a tvx when I bought my new truck and it's been great.

I have an old f150 I use rock crawling, built as a truggy. It had no ac factory so I pieced one together with junkyard parts. I used the rear unit from a lincoln navigator for the evap, fan and tvx. It's mounted behind the passenger seat and air gets ducted up along the console with shop vac hoses. This is a function over form truck mind you! The stock heat box in the dash handles heat and defrost still. I mounted a jeep cherokee compressor on a custom welded bracket and bought the biggest universal condenser and drier I could use from vintage air. I specifically chose this setup because the tvx, it made life easier with mismatch parts since it adjusts. I've thought of building my own system for my Fairlane but I'll probably just go vintage air for simplicity.

Variable compressors are cool tech too, but you need a computer to operate them as far as I know. Lots of modern stuff runs them. No clutch, they always turn, the computer varies the stroke of the piston from 0 to whatever to dial in the refrigerant flow as needed for variable conditions. At 0 stroke they pull basically no hp or they would still have a clutch because mpgs matter for manufacturers.

[mikedc]

I actually looked at the rear of a Lincoln Navigator for my project. I came away thinking it wasn't big enough for a muscle car cabin. And it doesn't look like something they intended you to service in general.

I hadn't thought of the size & inefficiency of older compressors being a motive for the switch to orifice tubes. That makes perfect sense.

I'm trying to put together a muscle car setup that is easy to deal with. The issues with modern systems . . . compressors/clutches dying early, freon levels being finnicky, snap-tight connectors on the hard lines that should have been threaded nuts, the under-dashboard stuff being a PITA to access . . . I'm trying to minimize all that.

On paper I'm sure it's not worth all the effort. But it's a matter of outsmarting my own laziness. HVAC problems are not like engine or brake problems; you can't procrastinate on failing brakes. You can procrastinate on failing AC. If the repair job is big & annoying then I might spend 2 months in a hot car.

[Tim C]

The Lincoln setup was fully serviceable. You can split the case to replace evap, heater, tvx, etc. The evap was the same width and thickness as an f150 front evaporator of the same years, but 2" shorter. But the f150 front evap is the same wether it's a single, extended, or crew cab. So I figured the 2" difference wouldn't hurt on my single cab 78 f150, and it's fine. I don't know if the thing is big enough for a muscle car cabin, that's probably more cubic feet to cool with more glass greenhouse area too.

As far as compressors go if I piece a system together I'll use a sanden 508 probably. It's pretty much an industry standard. It comes with several different OEM mounting configurations but also in universal mounts too. That's what vintage and the other aftermarket guys use and they have tons of brackets for all the common engine families. They are used in everything from semis, school buses, farm tractors and OEM cars and SUVs.

[mikedc]

Do you know what year of Lincoln Navigator that HVAC box was out of?

[Tim C]

No it was back around 2012 or so. I want to say it was the 05 to 08 ish. The first gen with independent rear suspension. But I can't say 100%. I didn't use any of the electric door actuators in my application. I cut the box in half just after the evap and make a new end that I tapped for shop vac hose ducts.

[mikedc]

The Ford explorers (probably a smaller platform) went IRS by about 2002. I'm not sure when the Navs did it.

I've investigated the idea of building a whole custom HVAC dash box. It seems like the best result but it's probably more work than I will tackle.

I don't think it would be THAT difficult. Not compared to the design & effort that guys put into other aspects of muscle car projects. But it's mostly uncharted territory. There aren't exactly online tutorials for it. It's rare to see a custom-built dashboard box for the heater core alone, never mind a whole HVAC box. I haven't found a single case of anyone else even attempting it.

[Tim C]

If you do, take plenty of pics. That'd be an awesome build to follow in my opinion. I love this stuff, homemade ac, junkyard electric power steering.

[mikedc]

I see good reasons to consider a custom one.

Pros:
You could fit some BIG cores inside a 50yo car.
Use modern parts that won't become rare for a while yet.
Have the core tubes pass through the firewall where you want them to.
Make the cores & blower motor all individually removable, without taking the whole box out of the car.
Do a blend-door design rather than a heater control valve.
Put a cabin air filter inside.

Cons:
Trying to make a box that holds all that stuff correctly. It's playing Tetris in 3 dimensions.
It needs flaps/doors and those might need some kind of power assist.
My inclination is to build the box out of steel sheetmetal but that would end up being pretty heavy.
Airflow efficiency would be an unknown in general. Sorta like trying to port cylinder heads by eyeballing it and hoping for the best.

[CamaroAJ]

Anyone figure out if TXV location is really important? I'm planning on running a C7 condenser and a C6 A/C compressor on my car and need to use a TXV for a long system life. So far all the systems I've looked at has it close to the evaporator with the exception of the Chevy Volt Hybrid battery pack. I had a TXV for battery cooling and it has about 15" of line between it and the evaporator section.

[Tim C]

Tvx or orifice location isn't that critical. But the closer the better, the coldest point in 5he system is immediately after the restriction, after that the refrigerant is picking up heat. If the valve isn't right on the evaporator or if any lines between the valve and evaporator are not enclosed in a box and exposed to the in car air flow they should be insulated so they don't pick up heat. Otherwise you reduce the cooling capacity.

Most tvx are right on the evaporator for packaging. Since it needs to sense both lines may as well sandwich it between the hoses and evap for smaller packaging and convenience. I have seen o tubes mounted right at the condenser exit. Dodge and gm trucks did that but then you have to insulate the hose from the condenser to the evaporator or it'll be pulling under hood heat into the refrigerant and won't be as cold once it gets to the evaporator.

[mikedc]

TXV location - yeah, that makes sense. The closer to the evaporator the better. Any pipe distance between there is heat-soaking room to reduce the accuracy of the valve's reading.


Generally speaking - I've always heard & read that the condenser is the limiting factor in most AC systems (more than evap core, the blower motor, or the compressor). Would you agree with that? As in, always try to run the biggest condenser you can fit?

[Tim C]

Definitely. You want the biggest and most efficient (more fins, serpintene vs tube style, etc) as you can fit. OEM guys are even using what's called a suction line heat exchanger now, but it's part of a matched system, the tvx has to be calibrated for it. I've seen it on ram trucks and Hyundai and Kia cars most obviously. It basically lilooks like the liquid line (from condenser to tvx) is teed into a fat aluminum section of the suction line (from evap back to compressor) but actually the liquid line is inside the suction line for the length of the fat section of line, it uses the still cold evaporator discharge to keep the refrigerant in the liquid line cold, and therefore still a liquid before it gets to the tvx. Basically it adds to the condensing action especially since it's actually colder than ambient air temp which the condenser uses. It also prevents the refrigerant from getting warm enough to flash boil from underhood heat as you want a liquid going to the tvx, the pressure drop after the tvx and change from a liquid to a gas is what makes the ac work in the first place. Any refrigerant that flashes to a gas before the valve is not cooling the car, reducing capacity.

[CamaroAJ]

Definitely. You want the biggest and most efficient (more fins, serpintene vs tube style, etc) as you can fit. OEM guys are even using what's called a suction line heat exchanger now, but it's part of a matched system, the tvx has to be calibrated for it. I've seen it on ram trucks and Hyundai and Kia cars most obviously. It basically lilooks like the liquid line (from condenser to tvx) is teed into a fat aluminum section of the suction line (from evap back to compressor) but actually the liquid line is inside the suction line for the length of the fat section of line, it uses the still cold evaporator discharge to keep the refrigerant in the liquid line cold, and therefore still a liquid before it gets to the tvx. Basically it adds to the condensing action especially since it's actually colder than ambient air temp which the condenser uses. It also prevents the refrigerant from getting warm enough to flash boil from underhood heat as you want a liquid going to the tvx, the pressure drop after the tvx and change from a liquid to a gas is what makes the ac work in the first place. Any refrigerant that flashes to a gas before the valve is not cooling the car, reducing capacity.

GM has been doing that for a few years now and refer to it as a super cooler line. I've been keeping an eye out on how the bends are in a few different cars to try and work that into my system.

[mikedc]

That sounds like quite a bit of added complexity for a minor gain in the cooling efficiency.

Is it because of the physically large engine bays in modern trucks? Their underhood lines have inevitably got a lot of distance to cover.


Could you reap some of that benefit by just insulating the liquid line more thoroughly? Or maybe pairing those two lines side-by-side inside the same insulation layer?

[Tim C]

IDK if it's a minor gain or better. I have seen a paper published that says it can reduce pumping losses by lowering the head pressure. That means less HP wasted turning the compressor. I'm sure that's a big reason the oe guys use it. Got to eek out that last .01 mpg somewhere. As far as complexity it's not really from a service standpoint. I'm sure it's more complex byo make but to service it's no different than changing any hose. Just there's 2 connections to loosen at each end.

[mikedc]

Yeah, googling up some images, the line-inside-a-line thing is not as complicated as I was initially thinking. Seems like a decent idea if there's a practical gain from it.

[Twentyover]

The main driver for tube in tube A/C line heat exchangers was CO2 credits. Prior to about 2018, when the EPA lost control of carbon emissions, they had a table of carbon emission credits vs design changes. Tube in tube, PWM blower motors, 1234yf refrigerant, among other design features would be granted reductions in CO2 emissions from the measured amount in the Federal Emission Test Procedure. The current drive by OEM's for 1234yf is EU regulation. Want to sell in the EU? Not doing that with 134a.

Regarding designing a mix plenum as suggested in posts 9-11- I think most would be surprised how much effort is put into this at an OEM level. Getting Temperature Control Curve slope positive so they don't invert (temp lever goes higher, actual outlet temperature drops), outlet airflow and temperature stratification (near equal, or specified, airflow from each outlet, and equal or specified temperature difference from outlet to outlet) is where months of development work take place. Start moving components around after the heater core, and these kind of go away. It was a problem that gave me grey hair in my 20's.

[mikedc]

Regarding designing a mix plenum as suggested in posts 9-11- I think most would be surprised how much effort is put into this at an OEM level. Getting Temperature Control Curve slope positive so they don't invert (temp lever goes higher, actual outlet temperature drops), outlet airflow and temperature stratification (near equal, or specified, airflow from each outlet, and equal or specified temperature difference from outlet to outlet) is where months of development work take place. Start moving components around after the heater core, and these kind of go away. It was a problem that gave me grey hair in my 20's.

Yeah, I had thought about inconsistent temp distribution issues that could crop up. As in, once the air is past the heat/evap cores & blend door and into the manifolding for the vents.

But I hadn't heard of inverted temp control effects. Care to elaborate on that? I had thought it was a pretty straightforward matter of using a blend door.

In the big picture, I was thinking about the airflow sort of like airflow in an engine. Try to keep the port size consistent throughout the trip. Tight-radius bends & sharp edges are bad. Avoid tall/narrow airflow areas (at least after the cores & blend door) to help keep the temp consistent throughout the air mass.

I'm aware that there would be other issues. Colder air is denser, and denser air carries more inertia when it moves. But I'm less familiar with tackling that.

[Twentyover]

......

But I hadn't heard of inverted temp control effects. Care to elaborate on that? I had thought it was a pretty straightforward matter of using a blend door.

...



You'll never see it in a production car because it's tuned out. It's a mistake. The temperature curve should look like a lazy S, the slope of the curve varying, but always upward to the right on an outlet temp vs control position graph.

Air mixing in the plenum does funny things. Being an HVAC engineer became more about aerodynamics in the last years I was working, trying to get air around corners we were playing with Coanda (Romanian aerodynamacist) effect, and as cabins became progressively quieter, moving air silently became a moving target.

Most people aren't as critical as the development specification for a vehicle. On test drives I'd frequently as the platform A/C guy about features I detected, and we'd discuss if I'd ask for warranty service if it was my car. We were tuned to be very critical of performance




......

In the big picture, I was thinking about the airflow sort of like airflow in an engine. Try to keep the port size consistent throughout the trip. Tight-radius bends & sharp edges are bad. Avoid tall/narrow airflow areas (at least after the cores & blend door) to help keep the temp consistent throughout the air mass.

...

HVAC always sucked hind tit in packaging space. Ducts should be round, until you had a radio in your space, you'd put in a kerdunk (colloquial term for an irregularity), then the instruments woulod squeeze you, and the stylists had control over the outlets, a MOST IMPORTNT FEATURE. And they usually f****d them up. (Note here- mid 80's GM outlets ike the ones on my G body el Camino are the one's I like best- low restriction and good directional control)

I was constantly amazed how little priority was put on HVAC, last time I looked (a good number of years back) about 30% of warranty dollars were spent on HVAC. And so important to customer satisfaction




......

I'm aware that there would be other issues. Colder air is denser, and denser air carries more inertia when it moves. But I'm less familiar with tackling that.

Density difference between cold air and hot air are no so great, so not so relevent to planning airflow

[mikedc]

You'll never see it in a production car because it's tuned out. It's a mistake.

Yeah, I get that. I'm just struggling to figure out how it could happen at all. A blend door is a simple device.



Being an HVAC engineer became more about aerodynamics in the last years I was working, trying to get air around corners we were playing with Coanda (Romanian aerodynamacist) effect, and as cabins became progressively quieter, moving air silently became a moving target.

Most people aren't as critical as the development specification for a vehicle. On test drives I'd frequently as the platform A/C guy about features I detected, and we'd discuss if I'd ask for warranty service if it was my car. We were tuned to be very critical of performance

I've noticed modern HVACs running quieter even when the fan is blowing hard.

Behind a modern dashboard you find big molded-plastic ductwork. I assume the quietness is mainly accomplished with larger duct sizes + smoother internal flow.

I suppose if the passages are too large then the airflow won't exit the vents fast enough to reach back into the cabin. But with modern dashboard packaging demands, that issue may not come up very much.


HVAC always sucked hind tit in packaging space. Ducts should be round, until you had a radio in your space, you'd put in a kerdunk (colloquial term for an irregularity), then the instruments woulod squeeze you, and the stylists had control over the outlets, a MOST IMPORTNT FEATURE. And they usually f****d them up. (Note here- mid 80's GM outlets ike the ones on my G body el Camino are the one's I like best- low restriction and good directional control)

I was constantly amazed how little priority was put on HVAC, last time I looked (a good number of years back) about 30% of warranty dollars were spent on HVAC. And so important to customer satisfaction

What amazes me, is that there's no decent dashboard access to service the HVAC systems. Imagine building a car without an opening hood on the front end.



Density difference between cold air and hot air are no so great, so not so relevent to planning airflow

That makes sense.

[Twentyover]

You'll never see it in a production car because it's tuned out. It's a mistake.


Yeah, I get that. I'm just struggling to figure out how it could happen at all. A blend door is a simple device.



Until you consider that you are sending a fraction of the air thru the heater core and a fraction bypasses the core, then they re-blend in the distribution plenum. What are the chances that there will be some eddy currents, or regions of high velocity where the entry air 'blows over' the directed air.

The condition is heavily pronounced on side window demisters. You have a channel directing air near the front base of the side window to clear glass for rear view mirror visibility, and defrost air from the windshield will bend that air current mitigating it's effectiveness replacing it with air that has cooled defrosting front glass. You could close off the defrost duct- nah, frosty windshield is a bad tradeoff for clear sideglass.

[TheBandit]

Hi Greg, just want to say thank you for all the information and sharing your experience/background engineering AC. I became fascinated with the original AC system on my 70 Nova from an engineering perspective and got it up and running to my satisfaction, but imagine there is still a lot I could do to improve performance on the air handling side of things. You've given me a few things to think about, so thank you for that!

Are you still working in this space or have you moved on/retired?

[Twentyover]

AIRFLOW AIRFLOW AIRFLOW. CONTROLABILITY CONTROLABILITY CONTROLABILITY

Don't know what the current airflow of your Chevy Twice is, but if it was my car I'd look at a blower motor and wheel from a late 70's early 80's A, B, or G body w/ HVAC and try to adapt it to your scroll housing. Add a relay so you feed the blower full voltage on high. Cars start getting comfortable about 225 cfm total airflow at zero body pressure. I like a car with about 250 cfm @ ZBP.

In the early 80's were were testing an economy Japanese car, and it's HVAC performance sucked. As a less to my young self, a mentor had us get a Cavalier blower mottor and scroll and cobble it in. Marked improvement. This car also had 'rubber vents'- didn't matter how you positioned them, within a couple minutes they were in THEIR happy place not yours. We taped some GM vents in place and the deal became very comfortable. We had some Japanese guys in the back- on the initial run without change they were sweating like pigs- but they were, of course, Very Comfortable in subjective ratings (Alot of ride testing uses subjective comfort ratings. Good engineers turn subjective ratings into objective performance specifications.) American rear seat passengers were screaming about comfort levels. After the change in blower and outlets, everybody was happy

Another thing is- make sure your ducting is sealed. Sucks when you loose 20-30 cfm in duct leaks



Hi Greg, just want to say thank you for all the information and sharing your experience/background engineering AC. I became fascinated with the original AC system on my 70 Nova from an engineering perspective and got it up and running to my satisfaction, but imagine there is still a lot I could do to improve performance on the air handling side of things. You've given me a few things to think about, so thank you for that!

Are you still working in this space or have you moved on/retired?

[TheBandit]

Thank you for the kind response Greg. I'm actually facing a failing blower motor right now. It is drawing too much current and blowing the fuse on high, and it seems the airflow is a little weak. The system has a relay to provide full voltage on high speed. I have a handheld anemometer but I'm not sure how to get to a total system airflow. Would I estimate that based on approximate vent area and air speed at each vent?

I do need to go through the system and replace all the duct seals. The interior stuff needs some refurb.

I'll see how much work it would be to adapt a later model blower. Just depends on the dimensions of course.

[Twentyover]

.... I have a handheld anemometer but I'm not sure how to get to a total system airflow. Would I estimate that based on approximate vent area and air speed at each vent?

It's pretty much an OEM test. The GM variant seals the entire vehicle cabin (including exhausters). A large dual plenum with a large industrial blower on one plenum sits beside the car. The is a venturi plate between the plenums. The other plenum is ducted to the car interior. A crude method is open one window, seal with a plate (we frequently cut a cardboard plate) and a gauge to measure pressure between the cabin and the outside. This is going to sound hokey, but we would use a piece of saran wrap loosely taped to a hole in the plate, It's incredibly sensetive to small (really small) pressure changes.

The car was powered to 13.8V, and the HVAC blower on outside air and full blower speed

The pressure is adjusted on the blower plenum until the saran wrap was fluttering between positive and negative cabin pressure (I told you it would sound hokey.) There is a pressure differential between the car interior plenum and the blower plenum to keep the cabin pressure at zero with respect to the environment when the HVAC blower is running flat out. The venturi plate between the two plenums has a calibrated curve for that venturi, so you read the pressure differential vs airflow to get the airflow at zero body pressure.

That's how we did it in the 80's- now days the machine directly reads the pressure differential across the venturi and goes to a look up table programmed in and sends a number to a digital readout. Alot simpler, but alot less fun