3 wire 0-5V AC pressure switch

[marks152]

My dads got a 2007 renault megane 2 and he has just had his air con gas refilled. They checked the system for leaks and assured him that it was leak free. However his AC is not working. Ive checked the compressor clutch and thats not engaging. Ive had diagnostics on it and its showing DF232 Pressure sensor abnormal voltage - PRESENT and will not clear. Ive had a quick look in live data and it looks like there is no voltage being seen by the ECU as its giving a pressure reading of 0.

Im going to have another look at it tomorrow and first plan to check for 5v and ground to the sensor while its plugged in. Assuming these are correct and I still have no voltage at the signal wire on the sensor - what will be the best way to fool the system into thinking the sensor is there for test purposes. I dont want to fit a sensor and then find the system has other faults and have to pay for refrigerant a 3rd time. The pressure at idle with AC off should be 0.8v and 2.2v with the AC on according to Autodata.

Any suggestions?

[Cornbinder89]

Fooling 5 volt sensors to the ECM is a dangerous game, easy to fry expensive parts and often in doesn't work anyway because the ECM looks for a changing signal when the compressor starts, not a steady signal that is "ok". If you have good 5 volt and ground feeds and you know the line is good back to the ECM, than I would change the sensor.

[marks152]

Your right. If power/grounds OK and I have no signal at the sensor then it's confirmation the sensors dead.

I just done want to get it degassed, fit a new switch, get it regassed and find there's further issues.

I suppose it's cheaper than a ecu replacement if it goes wrong.

[Cornbinder89]

I don't know about the Renault, as we don't have them over here, but many switches are on fittings with a valve core in them, so you can change the switch/ sensor without de gassing.

[marks152]

I hope that is the case. It will certainly save him 50ukp/67usd.

[Tom Greenleaf]

Just a note if that self seals or not?: Look hard at new one if it has a pin that would depress a valve much like a tire valve you are "probably" good to go. Warning though.

Best to know for sure as much as possible be ready if not and you do it. It would hiss for a second or two strongly suggest you wear protective gloves, eye protection as well. If slow to undo it then it would just leak away so be fast as possible. There's the risk if it just let's all gas out fast that will cause instant freeze burns same problem as a heat burns - totally dangerous. Just beware.

If you can search out a picture of the part that shows all angles of it would help.

Site is US based, all welcome however vehicles not sold here like Renault just looking for info and parts is not dependable. Always use total caution handling anything to do with refrigerants!

[marks152]

Ive just had a look at this car.This is what ive confirmed.

Ive tested the 5v ref signal and this is 5.00v. Ive load tested it too
Ive tested the ground signal and got 0.01v. I tested it also to battery + and got battery voltage. I load tested the ground and alls good
Ive continuity and load tested the signal wire from the sensor to the ECU and this showed 0.2ohms and passed load test
Ive tested the voltage on the signal wire at the sensor and got 0.2v with engine running an AC on.
Ive tested the voltage on the signal wire at the ECU and got 0.2v with engine running and AC on.

I think the problem lies in the low voltage coming out the sensor. Autodata suggests 0.8v+. So ive sent him back to the place that regassed the AC to have it pressure checked. Maybe they filled it and didnt check for leaks. Maybe they just told him they did. So this is where im upto now.

Any suggestions or comments on the 0.2v

[Cornbinder89]

To be honest, the newer electronic controls are not my strong suite. I know a little and enough to be careful around them, but not much more than you have already done. I recently found that some applications use the sensor and engine rpm to vary parameter's like HPCO, no longer a fix set point but one that varies with engine RPM! It can make diagnosis very hard.
You didn't say what gas it uses, I assume 134a but 1234yf runs about $1000 for 10 lbs here and would be very careful about loosing any at that price!

[marks152]

I'm pretty sure it's 134a.

The voltage did not alter on the signal wire with rpm. I checked this.

[Tom Greenleaf]

Marks152: You can't just check pressure and know it's full even if this was operating properly. You also are caught up with credibility of this Autodata site software or book? The best of any of them can be wrong up to "Dealer only" side for code readings which would say which circuit is in trouble or noticed still doesn't tell which item or why.

Defaults are system shut down and may take a properly programmed scanner with ability to reset it or then know it can't. We aren't there yet at all. If built and sold to a NATO cooperating country this motor vehicle would use 134a in 2007 and say so on decals. Country where made (assembled) and maybe country it was compliant for listed usually on driver's door jam including the VIN# (vehicle identification number.) That's on parts all thru a vehicle if not original to the vehicle a good body repair would duplicate the factory stickers marked duplicates. Not my trade - auto body repair if a factor here. If a collision vehicle all bets off,

[marks152]

I don't have the right connector to check pressure so it's gone to the garage that filled it yesterday for a pressure check.

I have taken autodata values as a guide only. These have lead me down the garden path a few times but I think a full system should be giving more than 0.2v on the signal wire.

[Tom Greenleaf]

For all I know it defaulted to that normally if any failure was seen. This is computer controlled and may take one to reset it? Can't know that either for this they don't even sell them all over the world. If this isn't working meaning no way compressing gas all you are getting is the "static pressure" says almost nothing but it has some gas in it nothing more. It's not able to tell you it's full but the lies about that are all over everywhere or are here anyway. With equipment know accurate for pressure and temperature exactly this chart below shows what to expect. That's all a plain check tells you unless below those suggests system is essentially empty is worth knowing.

Chart below is also listed here at the forums page down a few sections if it doesn't transfer?

[marks152]

The ecu may default to 0 with the code bring present but the signal from the pressure switch should still be present at the sensor as I'm manually checking with a multimeter. For all I know 0.2 might be fine and it needs dealer diagnostics to reset the code.

I'm just trying to find the specification for the sensor and from this I can work out what voltage I should be seeing once I know the pressure of the gas inside the static system. Still waiting to hear back from the place that filled it. They currently have the car.


Cheers for the chart

[Cornbinder89]

It would be nice to have pressure gauges on when taking sensor output readings. Without that, the best I can recommend is a SWAG (scientific wild ass guess) and replace the sensor in question. It is a guess and may not be right or cheap to do.
Since many things in automotive (and elsewhere for that matter) simple systems that were controlled for years by simple switches are being replaced with "dynamic" controls that can be programed and switches replaced with sensors.
I'm not sure why we need to go down that road, but we have, and too far to go back. I think that engineers have "fallen in love" with the fact that they can dynamically control systems while in operation with software and sensors, that were formally left alone and allowed to operate with just a few "limit switches" to protect the system when things go wrong.
Air conditioning is a prime example. From the operating standpoint, the system hasn't changed from the 50's when 1st fitted to automobiles. So why the old ways of protecting the system will still work. Ok, I can see with smaller engines needing a "throttle boost" at idle to keep from stalling and raise the compressor speed at idle, more and more throttles are "drive by wire" instead of a cable and pedal. For an engineer it is just too tempting to want to control everything via software. It simplifes building, but not diagnosing and repairing.