FredR

I posted about how my pressure transmitter on the ac system failed a pressure test at 100 psig leaking from the underside but no pick up on that to date.

My question is about the logic- when something is referred to as a "transmitter" the inference is that it is sending out an analogue signal for some other logic element to read and interpret what action to take. The transmitter I have shows continuity between the two terminals with no pressure applied so I am assuming it may in fact be nothing more than a high pressure switch.

The mount is interesting in that within the condenser discharge manifold it sits on there is a schrader valve and when the transmitter is mounted there is a pin that presumably opens this valve suggesting the valve can be changed whilst the system is live. My assumption is that the transmitter facilitates a condition wherein if the discharge pressure gets too high the compressor will be disengaged. Not sure I understand the need for this if as I understand, the compressor has a high pressure relief valve built into the case.

Thoughts?

worf928

It is an analogue sender. See WSM troubleshooting procedure 19-23 step 2.

I always replace its o-ring and the Schrader valve on the ‘up-tube’ when refreshing an A/C system.

FredR

Thank you for the reminder- I keep on forgetting about the cooling section and how the system reacts to fridgen pressure to modulate the fan voltage.

I changed the O ring as a matter of course. I could not find any sizing data for that O ring no matter how hard I looked. In the end it seems as though it is the same size as the one for the pressure switch- Porsche logic says this is not so given they have different Porsche part numbers in PET. A number of the O rings have been superceded [conversion to HNBR?].

When I found the body leak on the transmitter I thought I had solved the "losing test pressure" aspect but it turned out it was still leaking a bit of pressure overnight and I found the schrader valve itself was a tad less than fully tight- pinched it up and it held 100 psig test pressure no problem. Changing it out is a good idea- no idea if there are special variants for ac systems though.

Given the current lock down situation and that I run the ac more or less all the time for the hot season [next 7 months] - any suggestions for a temporary workaround to get me back on the road given I have no viable transmitter? If I were to leave the circuit open presumably it will see that as infinite resistance and drive to full fan voltage- correct? On the other hand my cooling system generally keeps the coolant return temperature at about 90C. At this temperature the fan voltage is the same as for normal HP pressure of about 250 psig.

dr bob

Fred --

The Schraeder valves are in fact refrigeration-specific, evn theough they physically interchange with a tire valve. Difference is the material of the little button seal on the wet/inner end. HNBR for the latest AC-rated valves, rubber for tire valves.

worf928

O-rings for the two senders are different. I have a cross-ref for one, not the other. I’ll post # presently.

It’s a resistance circuit. If less than X ohms the fans won’t run unless commanded by rad temp. What the exact value of X is, is unknown to me. I know for a fact that if the sender works and the system is empty of refrigerant the fans will not run due to only A/C on. Thus X must be >22 + N where N is unknown to me. I don’t know open circuit behavior. Very easy to test though: push button.


Yup —^ you can buy HNBR valves by the handful.

worf928

What I have for the pressure sender o-ring is: Santech 0242

FredR

Thanks for that- appears to be 8.9 x 1.8 which as I recall is the actual size I used after some deliberation sifting through my 3 boxes of HNBR O rings.Checking my notes it is the pressure switch O ring I could not find data on. The PET quotes part No 900 174 011 40 but could not find any data. The "best fit" was a 9 x 2- does that sound reasonable?

After pressure testing for 3 hours at 106 psig and no pressure drop, I then put vacuum on the system and my little vac pump quickly got down to the max it can deliver- unfortunately to date I have not acquired a micron gauge so cannot give a reasonably accurate indication as I am relying on the bourdon gauge and that is not reliable enough to quote a meaningful number. However it can easily resolve 0.2 of a bar and the workshop manual says if the vacuum drops by that amount in 10 minutes one has a leak. I find that very strange as I expect my system to hold vacuum for ages without dropping. Unfortunately my vacuum is dropping slightly- hardly noticeable over a 1 hour period but over 4 hours it is something like the 0.2 bar mentioned in the WSM. During last years tests it would hold overnight at the deepest vacuum my pump could make. Now I am wondering whether I have an issue [or not].

I also have to fill the compressor with oil. My firendly shop put some into the compressor but were not sure how much so I drained it out and got the correct spec of oil. The system has been well flushed and thus the next issue is where to add the oil. The manua suggests to distribute it if possible, the compressor normally comes with a full charge of oil in the casing, the WSM says when you add oil add it through the HP connection, my friendly ac shop tell me not to do that and to add it to the suction [that makes more sense to me]- any suggestions?

worf928

Every time I have experimented with o-rings I've gotten screwed. So, I will not opine on .1 mm difference.

However the part number I have for the pressure sender is 999 707 261 40 and I have good reason to believe that it cross references to the MT 0242.

Your part number - 900 174 011 50 - is for the low pressure switch which is a slightly difference size.

If my part numbers are reversed then I - Friday - put a receiver/dryer stack together incorrectly.

I have a micron gauge. You'll have to do some conversions to interpret the following:
After vacuuming for 8 hours on Friday I let the system sit overnight with everything closed but still connected.
I vacuumed 12 hours on Saturday with a final system vacuum of 425 microns. I shut off the pump and closed all the lines.
This morning after vacuuming the gauge set to 300 microns, I turned off and closed the pump, opened the valves, turned on the gauge and after a few minutes system pressure showed 1700 microns.
This indicates an overnight drop of <0.006 BAR which I attribute - mainly - to the gauge valves and port connectors.

I consider this good enough although I have seen smaller overnight drops from a similar final system pressure.

I also suspect that I didn't swap o-rings on the r/d stack senders. If I did then the size difference is - it seems - immaterial.

On the other hand, holding vacuum doesn't always mean holding pressure.

FredR

The O ring I have a problem with is the LP switch item- got myself confused- your data is fine and correlates with what I used. If you happen to have the MT part number for the LP switch O ring that would be very helpful.

I like to test both vacuum and pressure tightness. At the moment the pressure test was OK at 106 psig but really needs to be 250 psig- I assume both suction and discharge can handle those numbers even though the suction will never see such in practice. Just did another vac test and it lost 2 psi in 90 minutes which according to the WSM is fine but I think it is lousy. Sadly I am suspecting the compressor rebuild is less than perfect. At the workshop they tested the compressor to 250 psig but they do not test for full vacuum.

When I charge the system I work on the principle that Greenpeace are not looking! I pull a full vacuum then bring the pressure back to close atmospheric with gas. I then evacuate further so that partial pressure of any residual non condensables is negligible. The amount of gas released is next to nothing - they took R12 from us and caused a lot of issues- R134 is no worse than car exhaust ozone wise and the amount is nothing.

dr bob

Fred --

Put the total system charge in the compressor. Before you run the system the first time. rotate the compressor by the end of the clutch BY HAND, so you clear any oil that might have settled in a cylinder. A dozen rotations will do the job. Then start the engine and activate the clutch to run the system.

Putting oil in the low pressure port risks a slug of oil sitting in the compressor suction, waiting to lock the compressor when you run it. Putting some in the HP port would be OK, as it will get pushed into the drier, just not too much. Putting it in the compressor and making sure it doesn't hydraulic lock on you will get the job done handily and easily.

worf928

That's the only one I don't have. Once I do, I will have the magic a/c o-ring decoder ring and will post it.

FredR

Bob,

As always that sounds like sound advice. Originally I planned to inject oil in correct proportions to different parts of the system.

I am thinking of dropping the compressor again to do a vacuum test without the compressor in position to confirm whether it is the compressor that has the "slow" vac loss I reported last night.

FredR

Cheers- good to know I have not missed something obvious.

The entire ring sizing concept started me thinking some last year. Given the ac connections are nominally imperial sized units one might expect the associated ring sizes to be expressed in fractions of an inch rather than [conveniently?] rounded metric units.

When I had doubts about ring sizing first time I went through this exercise last year, I took careful measurements and did trial fits, undid the joint and examined the thing for signs of distortion and/or pinching. I then did a system pressure test best I could and soap tested each ring joint. The joints I found to be most "fickle" were the ones on the rear ac solenoid valve [6.5 x 1.5]. I have no way of torquing them other than "feel". How and why they ended up with a different size to the other HP joints I know not.

Once the system is operating I then do a final soap test on the joints.

worf928

I realize now, that we need to start over maybe.

Which sender on the ‘r/d stack’ is suspect? Were you originally suspicious of the low pressure switch or the pressure sender?

FredR

The pressure sender is the unit that has failed. I then discovered the schrader valve was not fully tightened so that was also passing when the sender was off the car. The O ring sizing was a secondary issue and I errantly quoted the ring on the sender when in fact it was the O ring on the LP pressure switch that I could not define size wise.

Made a bit more progress today-

First of all I removed the compressor and then using a couple of bungs from my Mityvac kit I pulled vacuum on the system less the compressor and over two hours it did not lose any vacuum whatsoever so the slight vacuum loss I am not happy about is on the compressor [or one of its nozzles]. I then checked the allen bolts holding the distribution manifold onto the compressor and one of the bolts was almost loose- urrgh! Pinched it up and went round all the bolts to ensure they are evenly tightened now.

I drained whatever oil was left in the compressor. I then risked life and limb dodging the corona virus and got a bottle of Liqui Moly PAG46 with dye [300 cm3]. I then added 170 cm3 into the inlet port after bolting a nozzle blank onto the discharge port so that I could decant oil into the compressor without worrying about oil coming out of the discharge port. I then fathomed out the best way to fit the compressor and this involved mounting the suction line onto the compressor with the HP port blocked off with a blank. Wrestled the compressor into position and initially loosely fastened the compressor to the rear tie rod to take the dead weight. This time I managed to hang the compressor relatively easy, connected the suction line at the union next to the LP schrader valve. The bit I was worried about was whether when removing the blank from the discharge port it would pee oil all over me. I opened the blank carefully and with a catch cup ready only a few drops emerged after removing the blank. tightened up the belt tension adjusters and put the thing on vacuum. Now I will have to see whether it loses vacuum or whether i have cracked that one.