No AC
 

Hi all.


I finally got my car running again with the adaptronic. I just installed my idle control valve and was messing with the settings when I noticed that the ac compressor was not turning on. It worked fine before with the power FC. Has anyone noticed this?

My PFC had zero issues as well.

With my Adaptronic select only the 1st and 4th (slowest and fastest) fan settings allow the compressor to trigger. This happens for some PFC folks as well.

I intend to troubleshoot it when I have a moment and then involve Andy (Adaptronic). I'm not sure that he'll be able to fix it. The way the A/C and blower circuit are wired together I imagine that the voltage the ECU sees is strange. This is why some PFC folks have the same issue.

Let's see what we can figure out.

Yeah, I know what you mean. I had no issues with the PFC but I know alot of people did. I also noticed that the tach needles twitches every once in awhile, which it never did with the PFC. I gonna check to see if the AC relay is getting triggered and see if I can resolve it. Everything in my HVAC is working and even the AC light goes on when you press it but the compressor doesnt kick in. I will also check the pressures, but now that I remember I did use the AC about a week ago with the adaptronic and it worked. I did just update the firmware so that may be it. I'm gonna ask Andy. I installed the beta version 10.126 which has faster comms and ID lag times preloaded. It got rid of the hesitation just off idle. Now I have perfect throttle reponse but no AC. I wouldnt mind if it was October.

I'm on firmware 10.126 and still have the same issue with my AC as when on firmware 10.108. I agree though, the throttle response is greatly improved.

My current thinking is that the ECU requires the digital input (on the ECU) for "AC on" to be grounded when the AC should be on. The ECU then triggers an Aux out (on the ECU) in order to engage the compressor and the cooling fan.

When studying the electrical diagrams in the FSM I noticed that the A/C on / off button is grounded via the fan speed switch. This makes sense as you wouldn't want the A/C to run if the fans were off. Unfortunately it's not a strait forward ground that the fan switch provides when it's on setting 1, 2, 3 and 4. I'll put a volt meter on it when I have a second but it's apparent from the FSM diagrams that it's providing a ground that is elevated above true ground in voltage ... depending on the fan speed that is selected. Mazda has even provided a diode between fan speed 1 and 2 to help with this (I'm assuming that this is what it's for). Diodes have about a 0.6 volt drop though so the ecu is seeing about 0.6v instead of ground (0v) when on fan speed two.

This is all of course theory until I put a volt meter on the ECU input. Once I confirm this I can start thinking about how to solve it. I'm thinking that we need a circuit between the Adaptronic ECU's "digital" input and the car's AC switch output. This circuit needs to take the varying "ground levels" or voltages that are coming from the car and provide the ecu with an input that is either not connected (floating) or grounded. Put another way, on or off without any in-between.

I've seen lots of attempts to solve this on this forum but no real conclusive success. I now have the motivation to solve it though as summer in Southern Cali is starting now. Well, if Andy has a solution I may not have motivation. We'll see. I haven't asked him as I don't feel that I have the problem fully defined.

Kissimmee is in Florida? No wonder you're on this ... heat and humidity!

Hi guys, sorry about the delay, been training up Mark who's going to be helping me with support.

That fan speed voltage drop does sound like the sort of thing that would cause this problem and I think some of the NA Miatas had the same thing.

I will check this in the factory wiring diagram and get back to you; Based on your description I think it will be a hardware change so I'll need to work out a way to make it work for existing guys. It didn't seem to be a problem on the series 8 (Japanese) ones I've tried, and my series 6 was a front cut, my race car doesn't have air conditioning (if you know what I mean, I didn't get the whole car).

I will check this out tomorrow but I could've sworn that the AC worked fine when I first connected the ECU. Last Monday I took it to work and went to get lunch and the ac was working fine. I didn't try it again until I had I stalled the new firmware and then it didn't work. Tomorrow I will see if I can activate the clutch on the compressor manually and I will also check the freon with some gauges. I wonder if I can downgrade the firmware to see if that is the problem.

I checked the F11 screen and it turns the AC does work, but only on fan speed 1. I guess its the problem that James is having as well as many people with the PFC. Whats funny is it worked on every speed before the firmware update and it always worked with my PFC. Oh well, I will dig into this further and let you guys know if I find a fix.

Hi Andy, jmm8904 and all,

I did a bit of testing today and I think that we have two issues on the series 6 cars, well at least I do on my 1993 r1. Here are the issues and possible solutions:

1)
Observation:
The AC button triggers the condenser / radiator cooling fan directly, bypassing the ECU. The fan runs 100% of the time that the AC button is on and the fan speed is set to 1, 2, 3 or 4.

Problem:
There is no electrical load compensation. RPM drops to about 680 or so. When the clutch does engage the ECU adds extra effort per the A/C setting, the RPM drops more, the ECU adds additional effort per the 1st recovery RPM setting, the speed jumps up to 1500 RPM, the extra effort for recovery RPM is removed and the RPM settles down to 900. This of course repeats every time the compressor cycles.

Possible Solution:
Add a setting for low A/C effort and high A/C effort. Trigger the 1st "low" effort whenever the A/C button is pushed. This will support the condenser / radiator cooling fan load. Trigger the 2nd "high" effort when the compressor engages.

2)
Observation:
The voltage presented to the CPU on pin 1E (ECU Digital 4) isn't really on / off, ground / 12v. Here's what I measured with the car at 1500 RPM, DVM between pin 1E and chassis ground.

fan switch setting - voltage
off - 13.71
1 - 1.35
2 - 1.98
3 - 2.41
4 - 1.90

Problem:
It seems that the ECU interprets any voltage greater then about 1.90 as A/C off while the car's electrical system presents up to 2.41 volts when the A/C is on.

Possible Solution:
Place a circuit in-between the ECU and the car's electrical system that changes voltage over 3 volts to Vcc and anything less that 3 volts to ground.

Hope that this helps you Andy!

best,

James

Hi James, thanks so much for that info! I thought it would be something like that but nothing like actually having the data.

If you're feeling adventurous, you could modify the ECU. On the circuit board, there's a resistor marked R11 - it's a surface mount, and it's zero ohms. If you hold the circuit board so you can read the writing, it's on the lower left.

I would change it to 10k. It's an 0603 size. If you're up for it and you have a set of tweezers I can send you some resistors. This will change the threshold up to about 5V.

Otherwise, we should be able to get the same result by cutting the A/C wire into the ECU input, connecting a 3.3k Ohm resistor in series (1/2 W is fine), and connect another 1k resistor from the ECU input to ECU ground (as in this diagram).

Thank you!
Andy

Thanks Andy. I'll give it a shot myself. I'm very good with a soldering iron but have never worked with surface mount devices. It doesn't look hard as long as I don't drink any caffeine that day ...

That's for the offer to send a few resistors to me. At 10 for $1 shipped, it's not a big deal. I just ordered a few. I went ahead and picked up a temperature controlled soldering iron as well. I've been meaning to move beyond my cheap 15w pencil iron for a while now.

I should have everything between Monday and Friday of next week. I'll report back once I've swapped the resistor.

best,

James

Hey all,

I did some research last night on the PFC forum and I'm going to try to install a relay to ground out the 1E pin to the ECU. That way the ECU sees a true ground. I already have a relay and a harness and I'm gonna try it tonight. I'll post results.

Yes, a relay would do it perfectly as well - for the benefit of the audience:

Cut wire 1E
Pin 85 = 1E input on car / loom side
Pin 86 = +12V ignition power
Pin 30 = ground
Pin 87 = 1E input on ECU.

Just finished the 10k resistor install. I had just enough time to start the car and cycle it through all four ac blower speeds. Works like a charm! I'll drive it a bit tomorrow and post again. I'll also post some pics for anyone who wants to know what's involved.

It works / some pics
 

Hi all,

My A/C is good to go. All four fan speeds blow cold! Here's some pics of the R11 resistor to give everyone a feel for what's involved in replacing it. You basically need some experience soldering, a magnifying glass, tweezers and a steady hand.

Thanks Andy.

James

JMM8904 any luck with the relay?

Andy,

Any thoughts on my other (post 8, above) issue? That is where the fan runs when the compressor is off?

It's pretty annoying at a stop light. The car nearly stalls and then revs to 1500 RPM each time the compressor engages. There's details in my post above but basicaly when the compressor isn't on the ECU isn't providing any load compensation for the electric radiator fan.

I'm thinking that an output in the software for "A/C button depressed" or something similar would be useful. This way I could add effort for the cooling fan and a separate effort for the compressor.

thanks!

James

All the ECU circuit boards we have here have now been modified with the R11 to 10k. Thanks very much for the feedback.

As for the turning the idle up whenever we have the A/C button on, I'll check the wiring diagram but I don't know if we can tell (from the ECU) if the A/C button is on but the thermo/pressure switch means the compressor is off... I will check. Perhaps we can use the radiator fan output as an input as well like on the NA Miatas...

This probably why I couldn't get my ac to work last time I put it back together.

Might be ... different cars might show different voltages to he ECU. The voltages at fan speed 1, 2, 3, and 4 on my car are so close that I can see how another car might not work at all. That is, all four settings might be just out of the range needed to trigger the A/C compressor.

Maybe this is why the PFC operated the A/C perfectly in some cars and not so perfectly in others.

Welcome!

Good point about the thermo switch. These cars have three fan speeds that are (I think that I remember correctly) triggered by four different relays. One for the ECU fan output, one for the A/C, one for the thermo-switch (mechanical switch in the thermostat housing), and one other that escapes me at the moment. The idea is that the fan speed increases with each additional input. Point is that the ECU's radiator fan output is electrically separated from the A/C radiator fan output so one can't provide feedback for the other. I need to study the factory diagram a bit to be sure but I'm 99% sure as I type this.

Two ideas from my side:

1) A "timed A/C fan effort" setting. This would have two parameters. One that specified the additional effort and another that specified for how long to apply it after the A/C input trigger was released.

So the idea would be that when the A/C thermo switch interrupted the A/C on signal the additional effort timer would compensate for the fan load until the thermo switch caused the compressor to engage again ... sort of like how ECU operates the fuel pump.

2) Wire a relay to the A/C fan relay in the engine bay. When the A/C relay powers the radiator fan it will power the new relay as well. Use the new relay's contacts to ground the power steering's engine load sensor. This should let me set up the power steering additional effort input to compensate for the A/C fan load.

OK, there's my $0.10 ... I'll let you point us in the right direction Andy!:)

Hrm

According to this diagram, the A/C switch connects through the thermoswitch before both the ECU input AND the cooling fan relay (refer diagram).

So - if the thermoswitch is closed (ie, it's hot):

1. You turn the A/C switch on
2. This turns on the condenser fan via connection 49 at the top of the diagram
3. This also turns on the A/C input on the ECU, and the ECU adds on the A/C additional idle effort immediately (should help with the cooling fan control)
4. The ECU turns on the A/C compressor output

And if the thermoswitch is open (ie, the evaporator is too cold), and you turn the switch on, it's as though you haven't turned the switch on at all - the idle speed remains the same, the condenser fan doesn't come on.

If the A/C is on, and the evaporator gets too cold, then it switches off the input to the ECU and the condenser fan at the same time. The ECU drops the idle back and disables the magnetic clutch at the same time.

Are you saying that the condenser fan actually still comes on when you press the A/C button and the thermo switch is open circuit?

Yes, my fans run if the A/C button is on and shut off when it is released. The fans don't cycle with the compressor clutch, rather they run continuously. Both of my Rx7s worked this way ... the current 1993 and my former 1994.

Anyway, I spent some time looking at the diagram that you attached. It matches my 93 service manual exactly. After sleeping on it and doing a bit of searching around this forum, I think that the issue has to do with a Mazda cooling system recall.

Here's the background info from Mazda.

This diagram explains my car's behavior. I believe that it's from a 95 service manual. From my reading this morning, Mazda installed the recall from the factory starting with 95 model year cars. Also, since Mazda sent the letter (first link, above) to all owners, the recall electronics seem to be common place in the US 93 and 94 model year cars.

Here's a diagram with explanation for my fan's behavior highlighted in red. Basically Mazda moved the cooling fan relay connection in front of the thermoswitch so that it grounds directly to the A/C switch.

Aha! Thank you, I'll check this out... hopefully we can get that signal into the ECU somehow...

Thanks Andy!

If anyone needs the R11 resistor replaced, PM me. I don't want to do shipping but I'm happy to replace yours if you bring the ECU by. I have about 6 of the 10k resistors.