Arduino - Digital AC control system for '80 911

[schwarz633]

I think you need this 5V relay module instead of the 12V.

[schwarz633]

Then you can wire the relays directly to the Arduino.

[Discseven]

Originally Posted by Jonny H

^ I'll have a stab and say that the digital output that is connected to the relay board opto coupler needs its configuration set to 'open drain'.

What happens when you ground IN2?

Jonny... John also suggested I ground IN2 and see what happens. (John, if you're following along, pardon me that I neglected this.) Will do this today.

Originally Posted by Rawknees'Turbo

Karl, was this ^^^ and example of ---->

HA! Not in this instance Rono. Dinner was genuinely dinner.

Originally Posted by fitchn

Karl,

I agree that, if everything is wired/powered correctly, just grounding an input should activate the indicator LED and switch the relay. If that works but the uC doesn't, then you probably have a code and or output driver problem. Your setup should work with either an open-drain or push-pull output topology on your Arduino. However, you mentioned that your control signal was switching correctly so I don't think that is your problem.

Your link only makes me more confident that your current setup can't work. The info you provided was using a 5V relay (SRD-05VDC-SL-C). You don't have a 5V relay, you have a 12V one (SRD-12VDC-SL-C). You mentioned your first effort powered the relay off of 12V and that it still didn't work but I can't help troubleshoot your previous setup without more information. I just don't think your current setup can work, and even if it does it will most likely be unreliable as components will be operating well outside their specified parameters.

I looked at John's code related to driving the relay, there is nothing that appears to indicate pin "output"... or "open drain" as Jonny noted. Before messing with the code, I want to make sure all wiring and voltage is correct. I looked at voltage from the new step down to the coil side of the relay yesterday and it was "around" 5v. Will note exactly what it is today.

In post 76 there's a parts list. As far as I know I've been consistent when referring to the relay as SRD-12VDC-SL-C.

Originally Posted by schwarz633

I think you need this 5V relay module instead of the 12V.

Schwarzy... This relay is to ON/OFF ac's compressor. I believe 12v is obligatory.


After breakfast am diving back into car. Will collect as much data as possible for comprehensive status post later today. Thank you guys for your participation

[fitchn]

Originally Posted by Discseven

Schwarzy... This relay is to ON/OFF ac's compressor. I believe 12v is obligatory.

Karl, I believe the 05 vs 12 in the part number refers to the CONTROL side voltage and not the load voltage. The load side rating is written right on the relay body; 10A @ 30V for both the '05' and '12' part versions. If you want to run your relays you will need to provide 12V to the control side of the relay. The NPN transistor in the schematic I showed (Q1) must be able to pull enough current through the relay to active it. For your relays, 5V just isn't enough head voltage to provide adequate current through the ~400 ohm control-side coil. You have two options: (1) switch to the 5V version; SRD-5V-SL-C or (2) wire what you have correctly with 12V available for activating the coil.

Best of luck with your testing, Noah.

[Discseven]

Originally Posted by fitchn

Karl, I believe the 05 vs 12 in the part number refers to the CONTROL side voltage and not the load voltage. The load side rating is written right on the relay body; 10A @ 30V for both the '05' and '12' part versions. If you want to run your relays you will need to provide 12V to the control side of the relay. The NPN transistor in the schematic I showed (Q1) must be able to pull enough current through the relay to active it. For your relays, 5V just isn't enough head voltage to provide adequate current through the ~400 ohm control-side coil. You have two options: (1) switch to the 5V version; SRD-5V-SL-C or (2) wire what you have correctly with 12V available for activating the coil.

Best of luck with your testing, Noah.

School'd! I am reading that incorrectly. Thanks Noah and pardon me Schwarzy. System was first car benched with 12v feeding relay. I stepped it down to 5v to see what would happen.

Continuity:

Between car and relay, everything has continuity. Display is working and all readouts appear appropriate for the day.

Voltage:

12v input side of new step down for 5v service to relay’s coil = 12v (I know now this is meaningless for this application but here it is... for the record)
Output side of step down = 5v
JDVCC / relay’s coil side = 5v via step down connection

Relay has 2 Ground Pins = common ground
Signal light ON, Signal Pin = 0v
Signal light OFF, Signal Pin = 5v
Power Pin on Signal Side = 5v
1st and 2nd relay have matching voltages.

...Sensor Failure:

Yesterday, to be able to “Set Temp” above and below cabin temp (in order to turn system ON/OFF), cabin sensor was cooled with bag of ice. Condensate soaked sensor and it disconnected. System will not work without all sensors connected/working. Figured overnight it would dry and work today... it didn't. (System shows a sensor disconnect/failure by freezing on Porsche logo.) Replacement sensor is ordered.

This does not bode well for the ambient sensor located upside down in left front fender / door jam---straight behind antenna. Question is: will ambient sensor survive driving in rain? I believe failed sensor got soaked. This soaking may be different condition than ambient sensor will ever experience.

Grounding Test: Did not complete due to sending hardware to John.*

On coil (left side) jumper is connected and there’s common ground between left and right sides. On right, power is supplied to 5v pin. Now grounding IN2…
5v to coil side:
Click = ______
12v to coil side:
Click: _____

* John returned from out of town yesterday and got caught up with relay situation. While he benched the system successfully, I recall he checked the signal to the relay---which worked---but not the relay itself. No blame on him---I'd have done same. If signal is there, relay should switch (provided relay is good and wired correctly.)

Bottom line… John requested hardware sent to him and he’ll figure out what's going on. Is on the way. In the meantime, am going to look further into the sensors used for ambient & cabin data.

[Rawknees'Turbo]

Uncle Karl, there are some seriously smart, electronics MFs posting here - am vicariously learning a lot from your thread (and having some good laughs, too, buttofcourse )!

[kuehl]

Other than cockpit sensors, anything outside the vehicle need to be weather proof (sealed).

However, if your occupants have high metabolism rates (they sweat a lot), if you ever intend to wash out the cockpit with a garden hose (like some used car dealerships do),
you spill your coffee often during down shifts, the neighbor's dog sneaks inside the door you left open, you believe the climate guru's are PARF or.... if you live in an area that see's humidity for long periods of time.... 'sealed' units my friend.

[Discseven]

Originally Posted by Rawknees'Turbo

Uncle Karl, there are some seriously smart, electronics MFs posting here - am vicariously learning a lot from your thread (and having some good laughs, too, buttofcourse )!

^^^ I know you Rono... padded cell with uberairsystem and multicolored dancin-teeners... you're no technoslouch in your own right. (Respect ya.) And CircusCircus too boot! (Your over-da-phone diagnosis lives in infamy Bro. Will consider putting you in my will... when I get to it. My Alodine collection is definitely yours )

Originally Posted by kuehl

Other than cockpit sensors, anything outside the vehicle need to be weather proof (sealed).

However, if your occupants have high metabolism rates (they sweat a lot), if you ever intend to wash out the cockpit with a garden hose (like some used car dealerships do), you spill your coffee often during down shifts, the neighbor's dog sneaks inside the door you left open, you believe the climate guru's are PARF or.... if you live in an area that see's humidity for long periods of time.... 'sealed' units my friend.

Stuff I read on DHT21 AM2301 Digital Temp & Humidity Module seemed promising. Knew it would not hold up to direct downpour and so the sheltered location. Given what has been experienced... am looking into DHT sensors further. Also looking---as you suggest---into sealed units. Problem with "sealed" is ability to read humidity. We'll see how sensing pans out.

On completely different topic... believe I heard---possibly from you---that charging AC system is best done on hothot days. Is this true? If so, what's the technical purpose/advantage?

[Rawknees'Turbo]

Thanks for the props, Karlicous, and in celebraion of your a/c pimpin project, I present the following (since it's kind of like you are on the a/c stage! )! Brand spankin' new material from one of my fav bands.

[schwarz633]

I'd be looking at a solid state relay (SSR) myself:

1 Channel SSR Solid State Relay High-low Trigger For Arduino Uno R3 5A 3-32V | eBay

You just have to make sure that the output is DC rated.

[Discseven]

Originally Posted by Rawknees'Turbo

Thanks for the props, Karlicous, and in celebraion of your a/c pimpin project, I present the following (since it's kind of like you are on the a/c stage! )! Brand spankin' new material from one of my fav bands.

On top of the skull&dagger muzak we're so fond of... add puppet profundity. Way to pick em Rono!

Originally Posted by schwarz633

I'd be looking at a solid state relay (SSR) myself:

You just have to make sure that the output is DC rated.

Like the idea Schwarzy... thanks for recommend. This will be the way to go if the 2 Songle relays prove faulty. First relay I may have messed up---with heat. Second was treated with kid gloves. John will receive both of them tomorrow. Basis of this project/thread is R&D. Goal is to find problems... solve them... and share the learning curves. Before abandoning Songles, John has op to uncover this Songle mystery and resolve whatever the problem is.

[Discseven]

Just got word from John. Neither of the two relay boards are functional. Exactly why requires an autopsy (which is John's call to do or not.) Is enough for me to know they're not working. Priority is to get one that does.

Ordered 5v as backup (after learning of my interpretation error---TY guys) but before Schwarzy's input on solid state. Should arrive in day or so. That unit will go to John for his testing (before I do anything with it.)

[Discseven]

Update = John's test new 5v relay... it works with addition of 1 ground wire added to signal side of relay. I screwed up ordering replacement temp/humidity sensor---failed to place order days ago. Just discovered this fact yesterday when getting all parts together. So... will be a few days waiting on missing sensor to arrive.

[Discseven]

Since last post, new parts acquired include: 5v relay (replacing 12v), DHT temp & humidity sensor (replacing one I ruined during testing.) Together with new parts, John and I have been sorting through a few issue…

Refinements made:

STEP DOWN & RELAY – 12v relay has been changed to5v. Dedicated power for relay has now been run from fuse panel (cig lighter fuse) to dedicated 12v > 5v step down.

5v relay has same appearance as 12v but for text printed on modules. Compared to 12v relay install, 5v relay now includes a ground added to the VCC side. (Could use single module relay here. Extra module serves as spare / back up.)




TEMP & HUMIDITY SENSORS - Two DHT temp & humidity sensors exist in system---an inside for cabin temp reporting (toArduino and data display screen), serves to compare target Set Temp to Cabin Temp with system cutting compressor OFF when both temps match. System turns compressor back ON after hystersis buffer is achieved. Outside sensor monitors ambient temp and humidity for entertainment only.

Having tested DHTs under various circumstances, there’s doubt as to whether the exterior sensor will continue to operate successfully after a 100% humidity experience. Reviews say these DHTs work through 100% humidity. My experience suggests differently---but then I forced cold & humidity on the DHTs for testing purposes. Perhaps a DHT responds more favorably to “real world” temp & humidity change. Because the exterior unit is wired and Arduino is programmed specifically for the two DHTs… will leave exterior DHT to the test of time. Cabin DHT should experience no issues.

With the first DHT’s failure (due my wetting it,) a new unit has replaced the wetted one. New unit was also subjected to 100% humidity. Humidity data froze at 99% (according to info displayed on the system’s data screen). Sensor was disconnected at housing terminal… humidity sensor wiped off… AC powered OFF… sensor reconnected... then system powered back ON. Sensor then resumed working properly.



DHT sensor’s plastic housing is clamshell with friction fit. It splits in half with force applied at its seam. Wires coming with unit are terminal-attached to a circuit board that’s single-screwed into housing.



Sensor board's bottom:




DATA CABLE FAILURE – During system tests, the Set Temp control knob began working both the Set Temp and Screen Contrast adjustment. John advised there being a wiring issue. Indeed, wires at the data cable’s junction at the AC Control Box were snapping off from the junction. One is snapped here. Other wires were almost to this point.




Stress on wire connections at the junction---due to constant handling of AC Control Box---and so the cable’s repeated pivoting at the junction is to blame. My bad (for not securing cable when it was first installed.)




Old junction's been cut off and new installed (keeping all previous wire positions intact.) Cable’s now secured to AC Box with zip tie (passed through two new holes put in Box.)




Refined system was just bench-tested in car. All data collection & on-screen display is correct, screen contrast control works correctly, new 5v relay is switching compressor power feed ON/OFF correctly.

John---your persistence in dealing with the info-tech / Arduino side of this journey…especially lately given my noted cable engineering shortcoming… much respect & admiration your way!

REMAINING DIGITAL TEST – As mentioned above, is questionable whether or not exterior DHT temp & humidity sensor will hold up to rain of any degree. If DHT sensor fails, will see about modifying it to prevent 100% moisture from reaching sensor.

NEXT STEPS - AC Control Box gets installed in center console. Console gets mounted in cabin. AC system gets evac’d & charged (brought back into operation from Black Death.)

Complexity Renote

Why bother complicating an ‘80 911’s AC system with digital control?

This is a collaborative build-journey to explore new ideas & techniques, resolve issues that arise, and arrive at a working system as a conclusion to the creative challenge accepted at the outset... The main purpose being to gain---and share---new insights. (Complexity is a necessary byproduct here, not a goal.)

[Discseven]

Schwarzie... thanks for taking the time to comment/suggest 5v relay and for putting together this wiring diagram. (I'm not IT guy in this build so would be up to John to comment on your wiring plan.)

Originally Posted by schwarz633

Then you can wire the relays directly to the Arduino.

[Dr J]

The reason we are not powering the relay from the Arduino as per Schwarz663 is due to a conversation Karl and I had, earlier in the project where he wanted to isolate the relay from the Arduino. Thus there is a separate 5V power supply to energize the relay, whereas the Arduino activates the control circuit on the relay board. In that fashion both components are separated by an optoisolator in the relay board.

As to the earlier discussions on using 12V vs 5V relays, it is irrelevant. which is used The 12 V and the 5V are the voltages necessary to energize the relay coil. The control circuit on the board used the Arduino signal in either case. The problem Karl had was that both 12V relay board were defective or damaged at some point; thus, nonfunctional.

Good going Karl!

[Discseven]

Am back in the saddle after being pulled in various directions since last post.

Recent progress...

Pin rack fault - Pin racks were added to Arduino to allow female plug-in terminal connections. Each pin can be pushed through rack if desired, also if not desired. Got system running and power began fluttering then... OFF completely. Found power pin rack being at fault. Changed rack to more solidly fixed pins.



Faulty pin rack as just removed. Longer pins go into Arduino.




Bench testing system in static car. Ice is used to chill cabin sensor for testing compressor relay ON/OFF. Tilt of display in mount is intentional---is so driver sees black headliner reflected in screen.



Got system going... and relay issues persisted---albeit with relay at least switching now. Were also anomalies with data coming consistently from sensors. Minor mod to wiring and refinement of Arduino sketch/program eliminated problems. System then began operating as planned. Much respect to John for relentless participation. (BTW, John's developing Arduino-based EFI system for his supercharged engine. Will add link here later. LINK: Another AEM EFI Conversion)



Been driving with system installed in console with console attached solidly to car and AC ON for a few days. Purpose being to see if vibration & road shock will screw things up. Also to confirm consistent data collection & accuracy over various temp ranges. So far so good. Only issue found is "one wire" sensor (is in evap's core) getting lost above 98dF---it rotates between two different temps. Is no concern (since evap won't operate at that temp except to start with at times. Once below 99 there's no prob. Believe this anomaly is sensor limitation but not sure.) The highly reflective bezel does not work optimally. Depending on ambient lighting, it suffers severe reflection at times. Most of time it's good. (Toasty lately... highest cabin temp while driving, both windows open = 114dF. More typically it's in high 90s.)

AC is absent refrigerant at moment. Will evac & charge next week provided digital control requires no more attention.

For newcomers to thread, is relevant to recognize this project was taken on to see if and how an Arduino-based AC controller could be worked into an '80 SC.

[Discseven]

Originally Posted by Discseven

Hysteresis Test Notes:

Goal is to determine what the actual working hys is and whether or not it’s stable (across various temps and settings.) For an ’80 911, this fine detailing can be considered irrelevant… particularly where short drives are concerned. Others have said it...“All I want is cold.” Agreed. Purpose to this fine-fiddling is to understand what the program is doing in reality---got to have this baseline intel in order to make sense now of whatever programming refinements should come. (Dimmer Pot was reconnected to system for this test.*) Test results:

Evap Freeze Prevention:
AC's power consistently turns OFF at 32.1 dF regardless of other settings/temps. This is a priority setting---over-rules all else.

Recovery Hys = 1.1 dF
After freeze prevention shut OFF, power consistently turns ON at 33.2 dF.

When "Cabin" Temp Matches “Set Temp”:
Power consistently turns OFF when “Cabin” temp matches “Set Temp” regardless of "Set Temp" level. (This, provided evap signal has not already turned compressor OFF.)

Recovery Hys = 2 dF
Power consistently turns ON when “Cabin” temp reaches 2 dF above “Set Temp.”

When Adjusting “Set Temp:”
When raising “Set Temp” starting from a level below “Cabin” temp… power consistently turns OFF when “Set Temp” and “Cabin” temp match (regardless of "Cabin" Temp level.)
When lowering “Set Temp” starting from a level above “Cabin” temp… power consistently turns ON 2 dF below “Cabin” temp (regardless of "Cabin" temp level.)

Wiring Note:

Stock fan and Temp control switches. Green/white stripe wire going from stock fan-to-temp-switch is disconnected from stock temp switch and will connect to power side of 12-5v step-down---this powers digital system. Outgoing terminal at digital relay gets green wire that disconnected from other terminal at stock temp switch---this powers AC system/compressor. (Incoming power to relay switch has yet to be sorted.) Stock temp switch with capillary gets deleted.

Evap Core Temp Minimum & Hysteresis Mod:

AC system's ON and OFF automated compressor control settings (32.1 dF noted above) is now thought to be too conservative. This conclusion is made prior to charging and running AC. Automated compressor ON/OFF has been changed (7.26.18) to:

Minimum evap temp = 29 dF
Hysteresis = 3 dF

Theoretically, intent is to push slightly into evap ice-land and cut compressor OFF before ice builds too much. Hysteresis will then "bleed" off ice accumulation... allowing compressor to come ON @ 32 dF. This should provide lower vent temps than prior settings offered. Curiosity looming is how far below 32 dF can evap cope with---in Miami's humidity---before reaching an ice-block level that renders system useless? With Arduino, can reprogram evap's core temp until optimal functional temp is found.

Been waiting for 90+ dF day---without rain or prior commitments---to vac & charge AC. Almost got whole enchilada a few times. More patience needed for ALL conditions to align.

.

[kuehl]

Been waiting for 90+ dF day

Put the car in a garage with a heater.
We have many shops in the west where their
shops are quite warm inside. I guess their landlords
won't allow them to have AC.

[Discseven]

^ Can't tell if you're in leg pulling mode or serious. IAC---Point of 90 is to charge and test ambient-to-vent temp delta all in one shot. Couple of times I've got vac going and ambient's at 90+... am thinking AC's gonna get refrig today!... then afternoon monsoon comes and ambient nose dives... then AC progress shuts down. This pattern can't go on for much longer. (August is historically the hottest month in Miami.)