Resurrecting a dead A/C - Where to start? - Page 2 - PeachParts Mercedes-Benz Forum

Diseasel300 · #16 03-30-2017, 12:53 PM

Now might be a good time to clarify my intentions: No gas is going in the system until I'm satisfied that it is reasonably gas tight. Throwing gas in a system that fails a pressure test is ridiculous, that is not my intention.

My point is that replacing all the rubber hoses "just because" is much more time consuming and expensive than making sure the system is leak-free as-is and topping up every few years. Obviously if the gas leaks out quickly, it isn't being refilled until the leak is found and rectified.

I'm not stupid, cheap, or lazy. A lack of tools and lack of nearby shops to perform the hose change and crimping is the major limiting factor. Spending a small fortune on hoses that may or may not be any better than what's on there now is foolish at this point. Let's see how these play out first.

Diseasel300 · #17 04-01-2017, 04:58 PM

Nothing quite prepares you for the stupidity of things the PO did. Apparently the compressor was changed at some point in the 90s, as it has a 90s identification tag and shows it is suitable for HFC-134a refrigerant. No idea what refrigerant the system was running, it does not have conversion caps on the Schraeder ports, and the original 1986 R12 expansion valve is still installed. I'm assuming they were running R12.

Unfortunately a butcher got a hold of it and I'm at somewhat of a loss as to what they were trying to do. Clearly the system was failing to work due to leaks at the O-rings (rock hard, compressed, brittle, etc). Someone decided that the best way to deal with this was to remove the RPM sensor (completely removed, wiring harness included!) and simply attach a hot wire to the remaining clutch terminal. Wire was still present, but no longer connected to anything.

You can guess the result. Compressor had a very unhappy end. Seized solid, and the swash-plate inside is heavily galled and shows signs of melted metal. The filter-drier seems to have done its job, everything past it was clean, however the filter-drier itself appears to be from the factory as well. Big no-no not to change it when changing the compressor!

Spent most of today taking lines apart, replacing O-rings, and flushing the system. The condenser was a horror show. Evaporator was clean, but full of oil. I mean FULL of oil. No way that could have happened without the pressure switches being bypassed.

Regarding the expansion valve: Take whatever your expectations of how bad that job is going to be and multiply them by 1000. The fittings had clearly never been off of mine before, and what an epic pain in the ass. I'm debating saying it was worse than the injection-pump fiasco from last year. Drain line is rotted out and will be replaced with tubing. What a stupid design....

Tomorrow will get the compressor mounted, Filter/drier installed, pressure-test, pull vacuum, and hopefully charge with refrigerant. Stay tuned...

funola · #18 04-01-2017, 05:11 PM

I came up with a "trick" to remove the expansion valve on my 85 300D, don't know if this helps on your SDL. I left the glove box in place. I loosened the 2 evaporator(upper) flare nuts first with a big adjustable wrench on the body of the expansion valve as counter torque. The expansion valve + hoses can then be lowered to give you more working room to undo the remaining hose flare nuts without worry of torque-ing on the evaporator. Do the reverse on install, i.e. install the hose flare nuts onto the expansion valve first, then install onto the evaporator. Makes the job much easier.

Diseasel300 · #19 04-01-2017, 05:29 PM

The expansion valve on a 126 is on the other side. Right above the accelerator pedal and tucked up under the dash. 0 work room would be preferable, in fact I'm pretty sure whoever thought up the system in that car went into hiding from death-threats from frustrated A/C techs. The 2 lower flares came right off on mine. The two on top were a whole 'nother story. Had to remove the side panel off the center console, pull out the carpets, loosen and move the "crash bar" aside. What a *****!

I did get it apart and back together, but man, what a massive PITA.

Junkman · #20 04-01-2017, 06:41 PM

I may go to the yard and get the crack free dash and let that be an excuse to deal with anything AC or climate related while it is out.

Diseasel300 · #21 04-02-2017, 09:08 PM

Well the mechanical end of things are working beautifully. New O-rings in the entire system, everything flushed, fresh oil, new compressor. Pressure test was perfect. Held 110PSI of argon for 1 hour. No bubbles with the soapy-water test. Drew a deep vacuum for 30 mins and locked off the gauges. Held vacuum for half an hour.

First sign of problems was upon engine startup with the compressor plugged in. With a vacuum still on the system, the compressor was pulled in and running! ACC switched off, but compressor running. Quickly dumped refrigerant in and proceeded to charge the system. Pressures are good and it blows nice and cold (52˚ out of the center vents at idle and high fan).

I'm suspecting the Klima relay is toast. When I got the car, one of the wires to the pressure pressure cutoff switch had broken and shorted to the chassis, which kept blowing Fuse #5 in the fusebox. NOTHING shuts off the compressor, even unplugging the pressure switches. As long as the key is on, the clutch is pulled in. Pulling the Klima relay shuts it off. I haven't done much testing yet (tired) but a cursory glance at the FSM seems to suggest that there's not anything else that could cause this issue other than a Klima relay that's stuck on. Gee, I wonder what burnt up the other compressor!

Still, it seems the big battle is out of the way - the mechanical side is done.

Diseasel300 · #22 04-02-2017, 09:57 PM

Pretty sure I have enough to condemn the Klima relay. After consulting the FSM, there's not really anything else to it. If the engine is shut down, the relay for the clutch should not be enabled, and even if it were, there should be a 10 second delay after receiving an RPM signal of 600 RPM or above.

When my key is in "run" with engine stopped and A/C switched off and Klima relay pulled out, the compressor is disengaged.

Plugging in the Klima relay causes an immediate internal "click" in the Klima relay and the compressor clutch pulls in.

I've confirmed that with the A/C turned "off" there is no signal from the ACC to pull the relay in either. With the engine stopped, there is also no signal from the ACC.

Yet the Klima pulls the clutch in. Pretty clear to me that the electronics in the relay have a fault of some sort, pulling in the relay when it shouldn't be. Failed transistor or something? Who knows.

Diseasel300 · #23 04-02-2017, 10:02 PM

For those without a copy of the FSM handy, the clutch control diagram is attached as a PDF to this post.

Tests I've done referencing the diagram below:

Pins 5-7 read open on the Klima relay unpowered

Pin 7 is ungrounded and reads coil resistance to ground

Pin 5 is hot with key switched on and dead with key switched off

Pin 10 reads 0V to ground and to Pin 1 with engine stopped and with ACC turned "off"

Low Pressure Switch shows 0V referenced on both pins to ground and reads "closed" since system has adequate pressure

Diseasel300 · #24 04-02-2017, 11:16 PM

Found and ordered a new Klima relay, so I decided to take this one apart and do an autopsy and possibly attempt a repair.

I've attached 2 pictures below with colored references. One is of the "foil" side of the PCB, the other is of the "component" side of the PCB.

The foil side showed 3 spots on a single trace that had been quite hot, and one spot that had burned through completely. Note the red arrows. The BLUE arrow is Pin 1 (ground) and the YELLOW arrow is Pin 11 (compressor RPM sensor ground). The top left red arrow points to the burnt open point in the trace. Everything to the right of that point (including the EPROM) no longer had a ground reference.

My theory as to how this happened is a wiring short at the compressor on the RPM sensor or clutch wires. There is no short or fault present currently, but that doesn't mean there wasn't in the past. The RPM sensor was completely removed from the compressor when I took it out and the wiring harness on the compressor body butchered! Why? Who knows...

Long story short, I traced the wiring out and soldered in jumpers to complete the circuitry. (Brown wires and red arrows on component side). Took the relay out to the car to test it, and voila - no more compressor engaged with the key on! I'll take the car out tomorrow and test it, will report back with the result.

BillGrissom · #25 04-03-2017, 03:41 PM

Quote:

Originally Posted by Diseasel300

... Spending a small fortune on hoses that may or may not be any better than what's on there now is foolish at this point. ...

I spent ~$30 (not a fortune) replacing the AC hoses in both of my 300D's. I have a Master-Cool crimper for a professional look, but as I said anyone can use Oeticker clamps, and many factory products use them. One always need to pick their battles wisely, but I always try rebuilding wear items when I have them off the car, so one less thing to bite you later. If the OP had installed R-134A, then you may have internal corrosion since it sat long without pressure and PAG oil absorbs moisture to form acids.

Diseasel300 · #26 04-06-2017, 09:50 PM

New Klima relay showed up yesterday and got installed today. Compressor does not engage. For whatever reason I do not seem to have any sort of signal from the PBU/ACC to ground Pin 10 on the Klima relay. Not sure if I have a broken wire under the dash or some other fault. The ACC is regulating temperature normally.

I did go for a test drive with a temporary jumper grounding the back side of the pressure switch. New Klima relay is doing what it is supposed to and all the systems seem to work.

When I charged the system last weekend, I deliberately undercharged it slightly to check final pressures and operation on a hot day (which I haven't had yet). Today was the same temperature as when I charged it, and the pressures were still exactly the same. Air blows cold at idle and is downright FROSTY when the car is in motion. I forgot how cold R12 systems blow! The last vehicle I drove with a functional R12 system was in 1998. Pretty sure you can use one of these cars as a refrigerator in a pinch.

All I have left to do at this point is to find out why I'm not getting a ground at X26-12.

Diseasel300 · #27 04-07-2017, 12:03 AM

Found it!

Had nothing better to do this evening (what an empty life eh?) so I pulled the ACC out of the dash to do some troubleshooting. Confirmed that the wiring is good all the way back to the ACC unit which is great news as far as I'm concerned. I was NOT excited about trying to fish a wire all the way back to that thing!

The problem is in the ACC itself. It would seem the transistor that switches the signal to the Klima relay has self-barbequed. You can see in the photos below. The first clue was the overheated circuit trace. Used a multimeter to determine that Pin 7 (compressor enable) goes straight to it. Took the ACC apart to find the carnage and charred remains of the transistor.

Surprisingly I was able to find replacement parts to attempt a repair on the board (stay tuned) and that is the route I will try first. Everything else appears to be ok. My plan is to replace the main power transistor and the smaller trigger transistor together as a set. No telling what damage got done to the trigger transistor. The trigger transistor is still manufactured by a variety of companies and is a BC33725. The original configuration of the power transistor is no longer made (obsolete package), but is made in a modern package. The original part number was BD435. The modern part number is BD435S. The difference between the two being a reversal of Pins 1 and 3 (meaning the new part will go in "backwards"). Otherwise, they are functionally and electrically identical.

As an excuse to get the parts order up to a value that the parts house will ship, I threw in some solder wick and capacitors to recap the ACC board "while I'm in there". Assuming the repair works, the ACC repair will be less than $1.50 in parts. Certainly worth a try IMHO

Diseasel300 · #28 04-10-2017, 11:38 PM

As promised, dead compressor pictures.

Examining this thing begs more questions than it answers. Where is the speed sensor? Why is the wiring harness all torn up? They didn't really try to run it like that did they?

This thing got hot - REALLY HOT before it died. Check out the black char in the discharge ports. Looking in the suction ports you can see the swash plate and the heavy scoring and discoloration. This compressor didn't die a pleasant death.

With the malfunction in the Klima relay holding the compressor on all the time, I'd imagine the system ran out of gas and just kept right on running, eventually trashing the compressor. A truly remarkable amount of aluminum dust came out of the condenser, but post-filter/drier everything was remarkably clean - clearly it did its job. The refrigerant was R-12 and the evaporator was FULL of mineral oil, easily 5-6 ounces. Probably the only thing that saved it.

The system works and is holding gas - been over a week now and the pressures are still the same as when I charged it. Now to get that ACC unit fixed...

vwnate1 · #29 04-10-2017, 11:58 PM

Subscribed .

Sooner or later I have to bite the bullet and re learn all the AC repair stuff I learned in 1984.....

THREE W123's and no AC in any of them right now

.

Father Of Giants · #30 04-11-2017, 09:45 AM

I too have a 300sdl with a dead AC but I'm scared it's going to cost a fortune to repair.