Arduino - Digital AC control system for '80 911

[Discseven]

Harness Cables Feeds

Evap sensor cable = fed through capillary grommet.
Ambient sensor cable = cabin cable is fed through bulkhead grommet into frunk at extreme left side under dash and behind a small air duct. Fender cable from sensor is fed through fender grommet into frunk.
12v Power from fuse panel to Arduino relay switch = fed through same cabin bulkhead grommet as ambient cable.
12v power from fuse panel to Arduino relay coil operation = same.
Display contrast control switch cable = same.
Relay box = on cabin bulkhead just ahead of stereo.

Power to Compressor/Relay

What fuse to take power from for the compressor/relay remains in question. Looking at my fuse panel reveals systems this car does not have but fuse panel has them wired: sun roof / rear wiper, fog lights, headlamp washer. Plan to use fog fuse terminal to power relay's coil. Beefier lines are sunroof and headlamp wash... Anyone see a problem with using either of these to power compressor?

[kuehl]

Karl,

The typical compressor clutch is 3.5 ohms nominal. Assuming, in a perfect world of 12vdc your draw is 3.5 amps, at 13.5vdc your at 4 amps nominal; that's all at the clutch not at where you are supplying the power which is normally at the factory mechanical thermostat.

Assuming the factory circuit wire from the thermostat back to the compressor is a 2,5 mm diameter, say between a 10 and 11 awg, possibly 12 ft max from factory thermostat to compressor area in engine bay, and depending upon the number of strands and turns, you might use a nominal resistance 1.5 ohms per 1000 ft.

On occasion I have see clutches fail to clamp do to resistance issues or voltage drops between the thermostat and clutch.

For S&G (****z and giggles) check your resistance between the factory's solid green wire at the thermostat and the compressor's power wire terminal back by the oil filter.

On "hysteresis", or maybe think of the total system response time in terms of what the user expects. On a system with stock controls and improved condenser and evap fuctions, let's say you have a good system that is pumping out vent temps under 40F on a baking 100F day, and the mechanical thermostat shuts down the compressor because its near or at the freeze point. By the time the compressor starts up again the vent temps can rise up into the 50F's and the recovery time back down under the 40's can take quite a long time. That's using a simple quick acting mechanical thermostat that is in good condition (charge wise).
If you are using a thermistor, and depending upon it's inherent response time plus insulation characteristics around it (encapsulation compound, metal jacket, rubberized coatings, etc.) you might find that 'swing or lag time' at the vents could be lengthy as well which might invite you to reduce the compressor's cycling time. If the on-off-on clutch time is too compact, short duration, the clutch will start to overheat due to friction (clutches 'drag' when they engage). Just something to ponder and test.

[Discseven]

Originally Posted by kuehl

For S&G (****z and giggles) check your resistance between the factory's solid green wire at the thermostat and the compressor's power wire terminal back by the oil filter.

CG... Appreciate input and S&G. I'll do this by measuring resistance of "extension wire" (reaching from cabin to compressor power wire) and resistance between multimeter contacts... then connecting it all and measure total resistance... then subtracting the connecting wire and contact resistance. Am saying this because am not an electrical wizard---perhaps I'm doing it wrong. Will post outcome later today.

Originally Posted by kuehl

On "hysteresis", or maybe think of the total system response time in terms of what the user expects. On a system with stock controls and improved condenser and evap fuctions, let's say you have a good system that is pumping out vent temps under 40F on a baking 100F day, and the mechanical thermostat shuts down the compressor because its near or at the freeze point. By the time the compressor starts up again the vent temps can rise up into the 50F's and the recovery time back down under the 40's can take quite a long time. That's using a simple quick acting mechanical thermostat that is in good condition (charge wise). If you are using a thermistor, and depending upon it's inherent response time plus insulation characteristics around it (encapsulation compound, metal jacket, rubberized coatings, etc.) you might find that 'swing or lag time' at the vents could be lengthy as well which might invite you to reduce the compressor's cycling time. If the on-off-on clutch time is too compact, short duration, the clutch will start to overheat due to friction (clutches 'drag' when they engage). Just something to ponder and test.

Evap sensor is waterproof'd with a rubberized coating and so a slight delay in sensing temp changes is evident (compared to other sensors being used.) When system shuts comp OFF at 32.1 dF, way program is currently set... comp turns back on when evap sensor measures a 1.1 dF rise in temp. I see what you're saying in terms of lag---definitely a point of interest. Perhaps you can sort this lag out in your head given your AC experience... I cannot. So will have to experience system operating to determine if & how to adjust program for better performance.

On short city drives (15 minutes or less) as is most common in my case, it's full bore AC all the way. On a Miami > Key West run (with no stops / no opening doors/windows,) enough cabin heat would be flushed out for system to go into a consistent cycling mode. That seems a better "condition" to see how the system cycles.

[kuehl]

Originally Posted by Discseven

I see what you're saying in terms of lag---definitely a point of interest. Perhaps you can sort this lag out in your head given your AC experience... I cannot.

Let's put mechanical thermostat vs. electronic on the side for the moment.
Lag, not like 'turbo lag'.

The higher the heat load on the evap core, ie. 100F ambients out there, humidity, mid summer sun cooking through the roof and 2500 nominal square inches of glass, 2 occupants pumping out 250+ btu's each... when that compressor clutch disengages and the TEV dribbles down, the evap core quickly rises from, say 32F up to 50F+ cause all that "stuff" is heating up the air flow through the coil. Then, clutch re-engages, pressures build up, TEV starts spewing liquid, liquid changes state to gas, hot air mass gets sucked up by the coils' fins and things start to get cooler at the vent again. But it ain't gonna happen quickly. That's why you are having fun here, right ?

Commonly, thermo sensors alike what you are using, are usually rated in response time based on a thermal bath in oil, time rated in seconds.

Unless your algorithm (program) can compensate, you need a different evap core sensor, the term is "thin film" aka TFTC.
Then, you want to account for 'wetting'. Ideally you are measuring a surface or air temp in the evap core fins.
The issue is moisture wetting on the sensor. The moisture affects the surface or air temp reading (evaporation of the water on the side for the moment).

[Discseven]

Originally Posted by kuehl

...Then, clutch re-engages, pressures build up, TEV starts spewing liquid, liquid changes state to gas, hot air mass gets sucked up by the coils' fins and things start to get cooler at the vent again. But it ain't gonna happen quickly. That's why you are having fun here, right ?

Yes. Am curious to see inter-relation between cabin and evap environments. How long will a near-freeze evap take to turn back on given the conditions this one is operating under?---We're gona find out.

Originally Posted by kuehl

...Unless your algorithm (program) can compensate, you need a different evap core sensor, the term is "thin film" aka TFTC. Then, you want to account for 'wetting'. Ideally you are measuring a surface or air temp in the evap core fins. The issue is moisture wetting on the sensor. The moisture affects the surface or air temp reading (evaporation of the water on the side for the moment).

Interesting stuff... TY. Right now there's no algorithm compensation to deal with what you're referring to. This is high end thinking on your part that you know I admire. It's intricate stuff that will take some learning on my part and subsequent evolution of the system. John could surly have programmed greater operating complexity but it fell on my shoulders to make the decision on how the system would initially operate. Basically the idea was to get a fundamental baseline running---then see what might be done in regards to refinement. ...It's all experimental.

Resistance Check- Set everything up to check green cabin wire to compressor power and discovered my peepsqueekmultimeter (GB GMT-312) was not up to resistance checking. Volts and continuity is as far as it goes. Will buy into upgrade after TG.

[kuehl]

Well, you got evap inlet temp, evap metal core temp, evap box outlet temp, and vent outlet temp.
Then you got ambient temp, cockpit temp (top and bottom), cfm rate, humidity,
the sun (E36 Y Factor).

[Discseven]

^^^ Upgrade arrived early... Just need to read instructions, then will join Kuehlemon resistance movement.


Progress notes...

Anticipated vibration doing slow disconnect on serial ports...






12-5v Step Down arrived from China Wednesday. Took 9 days via tracked postal service.





Would not have ordered this StepDown had I known it was coming from China---cloaking China part orders seems to be on the increase. No cultural prejudice here. Just want stuff sooner than later.

Side road: For China interested… wonderful country with warm people (as I experienced them.) On the street, no half-English-help is available to get around. And forget about reading signage---it’s all abstract art. Recommendation for fellow Pelicangringos… do China as a tourgrouper.


Beijing sidewalk. Looking for tall blonde China babe is easy. Finding one is not.


On tributary off Yangtze.





House of warriors in Xi'an.


Was told ambient-buffet is for local guts… gringo’s can go down-for-the-count quick & hard on this fare. Whether true or not… there was too much China to see to risk spending my days bunked.

[Discseven]

Step Down’s 12v-side wires. They’re similar to 26 gauge am using for 5v. Not what I expected to see.




Connecting skinnies like this to heavier gauge (fan switch) wires causes me some concern---I know narrowing path for electrons = not a good idea. Opened Step-Down to see if internal connections offer ability to swap heavier gauge into unit. Not looking good---might screw things up trying a wire swap. Will do evil and connect mismatched wires. Must remember to check wire temps during operation. If even slightly warm… will deal with it.




Stock thermo switch with capillary… no longer part of AC system.




Harness wires as they stand at moment. Everything is connected but for relay’s ground (and USB power to Arduino.) Will hook AC control box to this tomorrow...




Power through relay switch for compressor is now connected to “Sun Roof / Rear Wiper” fuse terminal. (There’s no power serving this fuse unless ignition is ON. There were two stock wires connected to that fuse’s service side---both are now disconnected.) Relay’s coil power is connected to “Cig Lighter” terminal. (There’s also no power serving this fuse unless ignition is ON. Cig lighter wire is disconnected.)




Disconnected Sun Roof / Rear Wiper and Cig Lighter wires. Appear to be stock given “blocked” ends. (Two black disconnects were in Sun Roof / Rear Wiper fuse. Blue is Cig Lighter.)




Loc-Line installed on bow-tie outlet. No simple matter to neatly adjust these vent's positions given them being on 12" boom. Since I rarely touch “bow-tie”… doesn’t matter in my case.


Light inside vents are reflections oddly exaggerated by camera.


Wasn’t going further with LL plumbing than install on bow-tie outlet, but am liking this stuff so moving ahead with side vents---will do whatever is necessary to adjust air flow when time comes. Starting with driver’s side, have explored a few different paths. Preferred solution is to use flasher’s hole. Hole happens to be close (but not perfect) fit for ¾” LL plumbing. May seem an odd vent location but positioning it here is simple, clean looking, and there’s no chopping dash up. Fit of LL in hole is tight enough so no need to secure plumbing to dash. Nozzle can be directed with relative ease. How to connect this plumbing to car has yet to be sorted.




Applied mild heat to dash to minimize those two punctures next to hole.


Formally, the proboscis vent. Rono, know you're thinking in an alternative vein.

[Discseven]

When flasher’s out… turn signals don’t work. Might be other things not working as well. Reconnected it and stuffed it up under dash.




Module fits female one way only.

[Discseven]

In-Car Test – Found stock junction in smugglers not connected well---terminals were not in plastic housing. Had to push terminals into housing and so into each other... one by one. (Think age is getting to all this stuff. Is ’80. Admire you guys keeping even earlier cars in shape.)

Different fan speeds, sensors, Set Temp, and contrast adjustment… all working. Only item not doing so is relay---is not switching 12v to compressor wire. Checked current input across relay board and all's good---both 5v and 12v. Spoke with John. He asked if relay clicked. It doesn’t. Plan is to pull relay and move signal wire to operate second switching module. Maybe first module is bad.



Relay.




Air volume from LL vents (off bow-tie outlet) is less than expected. Clearly it's going to take some time to tune flow among all the vents.




Scuff marks in console’s leather are touched up with Sharpie. Leather is treated with Dr. Jackson’s.






Stock holes that join AC control box to console are usable. Speed nuts used to be here. Going with 10-24 screws & rubber washers instead. Console is tapped... screw will drive through console but sit loosely in box. Idea is to keep box from slowly vibrating out of console. Screw is like deadbolt.


Checking loose fit of screw in box's hole to see how much registration accuracy
there needs to be when inserting box into console.



10-24NC can be driven into console's existing hole by hand.

.

[Discseven]

Relay issue - Changed wires to use 2nd switch and it too is not switching (to provide power to compressor.) Once again checked all wires for required voltage... all there. Have order 2nd relay.

[kuehl]

Karl,
new relay failure.... what are the odds?
what are those funky jump pins for on the relay board?

[Discseven]

^^^ Yeah... odds are way out there CG, but possible. Is also possible I fried it. If so, is small-fry stuff---a minor delay & easy fix. Pins are elec connections. Could put wires into terminals that insert onto those pins. Because I put disconnects for all relay wires outside the relay's box, I soldered all wires to those pins.

While project is in suspended animation awaiting new relay, am into resistance movement today---large gin in hand for ya ... this is new territory for me... TY

Did you see fracking vid? (Seems there are significant ramifications for NE's water table---which is why I shared vidlink with you. Gas industry promotes their product as "clean energy"---In reality it's an environmental disaster unfolding before us. Am not one to focus on bad news however in this case, the natural gas sector has to be recognized for what they're doing. Pardon this heading into rantland but this qualifies as a significant heads-up for anyone that drinks... water.)

[kuehl]

The fracking thing.....

Most people, outside of the gas and oil exploration and production industry, are completely unaware of how their energy is gotten and the true 'cost'. Most have no clue as to the damage to land and the water supply. Things are not just F'd up here in the east, check out this Fracking America

Thank you to the Bushes, Cheney and Halliburton.

I grew up in a town with a water supply completely contaminated from 120 yeas of DuPont. There were many cancers and they continue today. The land is still polluted and will be forever; just like Love Canal. Because of politics/lobbying it will remain so. Just follow the money.

And, I thought our generation would make a difference.

[Discseven]

Originally Posted by kuehl

And, I thought our generation would make a difference.

You're not alone in this.

What to do? Can't keep head in sand---talking to myself here.

[GH85Carrera]

Originally Posted by kuehl

The fracking thing.....

Most people, outside of the gas and oil exploration and production industry, are completely unaware of how their energy is gotten and the true 'cost'. Most have no clue as to the damage to land and the water supply. Things are not just F'd up here in the east, check out this Fracking America

Thank you to the Bushes, Cheney and Halliburton.

I grew up in a town with a water supply completely contaminated from 120 yeas of DuPont. There were many cancers and they continue today. The land is still polluted and will be forever; just like Love Canal. Because of politics/lobbying it will remain so. Just follow the money.

And, I thought our generation would make a difference.

You are getting into PARF Charlie. Fracking has been going since on before Bush Jr. was even born. It is not a perfect thing but it done way deeper than ground water. It is a complex issue and has nothing to do with the a 911 technical site.

[kuehl]

Oh.
Gee!
Really?

Lions and tigers and bears. Oh mY!


Karl. I apologize for fin'g up your thread by answering your question you asked... on your thread.

[Discseven]

Originally Posted by kuehl

For S&G (****z and giggles) check your resistance between the factory's solid green wire at the thermostat and the compressor's power wire terminal back by the oil filter.

CG... resistance is 0.3 based on 200 setting on multi. If you would explain what this means... much appreciated.

1 = Probes touching
2 = Probes + extension wire
3 = Probes + extension + green wire from T-switch back to power wire at compressor.

[kuehl]

Karl,

I don't know squat about electrons, negative runs to positive.

I'm guessing 200 might mean up to 200k ohms and 2000 might mean 2000 ohms ?

Most meters, read 0 (zero) when you have the unit set to continuity, ohms or resistance.
and you are only touching its 2 probes together, (keep your friggin fingers off the contacts).


Anyway... Let's get out the User's Manual.

[Discseven]

Originally Posted by GH85Carrera

You are getting into PARF Charlie. Fracking has been going since on before Bush Jr. was even born. It is not a perfect thing but it done way deeper than ground water. It is a complex issue and has nothing to do with the a 911 technical site.

Glen... I know you to be inclined to staying on point. With respect, will ask that you tolerate the few side roads taken. (Knowing Charlie's local I thought he---and perhaps others---would be interested in fracking vid. China pics that related to China order I placed... posted them for those who might be interested seeing a glimpse of the country.) I'm with you on this being a tech forum. To back that up, would like to think I make sound deposits in our Porsche-brain-bank. At same time am inclined to believe we should offer the brotherhood here a little bit more than pure tech---my reason being to keep things from being overly one dimensional.

If there's widespread discontent with my wanderings, let me know.