You may have 13V but no current. Check compressor clutch resistance. Check voltage at the splice with black meter lead connected to ground.

 

You have a switching transistor inside the controller, and at one end the transistor is connected to the battery voltage and at the other, the compressor. When you break the compressor connection, you will see close to battery voltage, because the internal resistance of the meter (mega-ohms), is much higher than the off resistance of the transistor. If the controller had a relay contact to control the compressor, you would see 0V, as the open contact resistance is close to infinity.
So if you connect the compressor to the controller and measure the voltage, across the compressor input and ground, you would read very close to 0V.

Coming back to your problem, I think something is telling your controller not to engage the compressor, because possibly a sensor fault ?

 

Keep in mind that there are sensors that prevent the compressor from engaging in the event of a low refrigerant scenario. Your ML is nearly 20 years old. All semi-hermetic AC systems leak . If you haven't topped up the refrigerant in 19 years, it would be worthwhile to top it up with some r134a and dye to identify the possible leak. You can turn on the AC system, charge from the low pressure side (near firewall) and it should engage when there is sufficient r134a in the system.

I really like this product:

Quest R-134a refrigerant - Not for Hybrid vehicles with electrical driven compressors 334 - Read Reviews on Quest #334

If your compressor does engage, give it the whole can, it might take a while. Shake the can to make sure the dye goes with the r134a but don't tip the can upside down.

Buy a "scorpion light" from walmart camping section. Get a big one. Then run the AC as you normally would, and check for leaks with your scorpion light.

 

Sorry, guys, it's taken me a while to get back to this and respond to you.

Witek_M - - - The resistance of the compressor clutch measures approx 132 ohms.
The voltage at the male side of the splice (splice separated) = 13VDC. The voltage at the splice with the splice connected = 0VDC.

mrboca - - - you may be on to something, but I'm not sure how to proceed in troubleshooting my AC.


Clearchris - - - I'm pretty sure there's plenty of refrigerant in the system. The pressure gauges read a static pressure of 70psi at 72 degrees ambient air temp with the system not operating. Near as I can tell, that's what it should read. With the engine running and AC system operating, the high pressure climbs to about 270 and the low pressure drops to about 22. I think that is pretty much "spot on".


I'm at a loss of what to check next.
The mechanical side of the AC system seems to be working as it should.
The electrical side seems to be working - but I'm befuddled about why the compressor clutch will engage when I provide 12VDC from the battery with a jumper wire but does not engage when the "signal wire" provides 13VDC.

What do y'all recommend?
What should I test next?

I'll respond much quicker this time.
Thanks
Bill (Runamuk)

 

The resistance of the compressor clutch measures approx 132 ohms.
The voltage at the male side of the splice (splice separated) = 13VDC. The voltage at the splice with the splice connected = 0VDC.

Compressor clutch coil resistance should be 10-20 ohms. Connect your amp meter straight from battery to compressor clutch coil at separation point. Measure current consumption.

 

Witech_M - - - 4 amps
With multimeter set on DC Amps. Red lead on positive battery post. Black lead on wire to compressor clutch coil. I can hear the clutch engage and the meter reads 4.00 amps.

 

Something does not add up here. According to Ohm's Law I=V/R.

 

Witech_M
I went back and retested a few things. The resistance is 3.3 ohms (not 132, sorry about that).
That makes the Ohms Law formula work better - - 13v/3.3 ohms = 3.93 amps.

 

Witek posted the correct value's in post #6. Your coil resistance is too low.

 

You might find things easier by accessing connector X22 in the fuse box.

Locate X22 and find B4 pin, PINK on one side and Brn/Violet on the other. With this the battery will be closer.

 

Ya, art_arev, I get it that the resistance of the compressor clutch coil is low. And I thank Witech for sharing all his knowledge and input.
What I don't understand is why the clutch coil engages and works fine when it gets 12V from the battery via a jumper wire, but does not work when it gets 13V from the AC system.

 

It is because AAM is not able to provide sufficient current. Or there is voltage drop between AAM and compressor coil. If you scan your vehicle with pro grade scanner I bet there is fault stored in AAM for compressor clutch.

In post #6 I have stated that compressor clutch coil should have 10-20 ohms resistance. That is valid for compressors with internal pressure regulating valve. For early models it should be somewhere between 3-5 ohms.

 

Thanks Witech_M. It's an "early model" ML320. A 1998. Sounds like the coil resistance is within spec's.
So .... how do I fix this thing? What do you recommend?

 

Contact this place and see if they will repair it for you.
https://www.bluelinkdiag.com/contact/

 

Thanks Witech. I appreciate your attention to this and the sharing of your knowledge.

When you say "if they will repair it for you", Are you talking about the compressor clutch coil (which seems to be within specs), or the AAM? If the AAM - what is it? Where is it? How do I get it out of the car?

Thanks again.

 

We are talking about repairing AAM.

 

Run,

Did you ever try adding refrigerant? It sounds like you have the necessary tools to do so, since you mention high and low side pressures. It might be worth the $4 to try adding a can of refrigerant before you go pulling your AAM and sending it somewhere.

I feel like I am stating the obvious in response to your question about why the clutch will engage with a jumper, but not with the system's wiring - perhaps a sensor in the system is preventing the application of voltage/current necessary to actuate the clutch. This is quite often the result of there not being enough refrigerant in the system....

For the love of all things Mercedes man; add some refrigerant!

If you have already done so, then I must have missed it in your posts, and I humbly apologize and offer you the best of luck!

M

 

Thanks for the response Mike Kilo.
You may be right. It's been my (very limited) experience that low refrigerant is often the problem. I am certainly not very knowledgeable nor experienced about auto AC systems.

There are 4 symptoms that led me away from adding refrigerant:
1) When I provided 12V from the battery to the compressor, the system came to life and worked perfectly, blowing 42 degree air.
2) The high and low pressure gauges showed readings that were right on spec's.
3) The action of the gauges with the system running were exactly what you'd expect.
4) The wire from the AAM to the compressor would show 13V for a minute or so, then drop to 0V for a minute or so, then 13V for a minute or so. Exactly what you'd expect as the AAM tells the compressor to cycle on and off. However - thanks to Witek_M, he nailed it in Posts #2 & #13 - when the wire showed 13V, it also showed only 176 microAmps. That's right, 176 millionths of an amp. Pretty close to no current at all. Rubbing your hands together would probably create more current. Certainly a long ways from the 4amps needed by the compressor coil.

So ... the AC is still down as I spend the winter pondering what to do.

PS - I did contact BlueLink (as recommended in Post #15 by Witek-M). They will fix the AAM for $250 plus shipping. That is probably the route I'll go.

 

I too, have limited A/C repair experience, but I have learned that pressure is not the same as volume. It may be that, while your pressure gauges indicate normal operation with the car's wiring bypassed, there might not be enough VOLUME of refrigerant to kick on whatever switching circuitry MB used in this system.

I had the dreaded ticking blend door actuator, replaced it following 43sqd's phenomenal tutorial, and then my A/C still didn't work. Having the hoses and gauges myself, I put about 1.5 cans of refrigerant into the system, and viola! It has been blowing cold air and working perfectly going on 4 months now...

Can't hurt to try a simple, and cheap, effort before getting super crazy into the system components...remember the KISS principle!

Good luck man, and let us know how you fare!

M

 

I too, have limited A/C repair experience, but I have learned that pressure is not the same as volume.

M

Wrong. At given temperature pressure of any compressed gas is the same regardless of volume. 12 oz can of refrigerant will have same pressure as 30 pound jug.

 

Good & valid points, Mike Kilo.

How do I test the volume of refrigerant?

I've seen the spec (but don't remember what it is) for the number of ounces of refrigerant the system holds. Would I have to drain the system, then reload keeping accurate count of ounces of refrigerant? If so, how do I drain the system in a safe and environmentally responsible manner?

PS - I have the dreaded ticking blend door actuator, also. Mine works ok about 90% of time. I picked up a good actuator at a junk yard, just haven't torn my car apart to replace that stupid thing.

 

With 250 psi (17 bars) of pressure at high side, you have PLENTY of refrigerant in your system. Instead of going around circles, have your system scanned for sensor / equipment faults. It could be a bad pressure sensor, outside temp sensor etc etc.

 

Um, perhaps you need to reread my statement, Witek.

I said NOTHING about the relationship between volume and pressure. Merely that they are not the same. Voltage and amperage are not the same. Apples and oranges are not the same. Height and distance are not the same. They may or may not have relationships with each other, but, the fact remains they are not the same.

I did misspeak, however. I should have said amount (or mass) and not volume. The volume of a closed system (such as air conditioning system plumbing) is fixed. The things that vary in an air conditioning system are amount of refrigerant, temperature, and pressure.

In point of fact, pressure and volume have an inverse relationship when temperature and AMOUNT or moles of a gas are held constant:

Boyle's law or the pressure-volume law states that the volume of a given amount of gas held at constant temperature varies inversely with the applied pressure when the temperature and mass are constant.

V is proportional to the inverse of P
Another way to describing it is saying that their products are constant.
PV = C

When pressure goes up, volume goes down. When volume goes up, pressure goes down.

 

I think the formula is PV = nRT if I remember correctly. And this applies only to the gas portion of the refrigerant. T is not constant, in both static and dynamic conditions. When T (in Kelvins) is equal to the ambient temperature (ie, when the car stays overnight , all temperatures are more or less equalized to outside temp, the pressures at high and low sides are more or less the same, and equal to the temperature in Fahrenheit. So, when the OP indicated 75 psi in 72 degrees F (if I remember correctly), this pressure is equals to almost 5 bars, and this typically determines for the a/c controller to start the compressor or not. The controller checks the refrigerant pressure, and if it does not meet the minimum, the compressor will not run. External reading of 5 bars is typically sufficient to start the compressor, but if the refrigerant pressure sensor is bad, the controller will read some other value (too high or too low). I am not saying this is the problem, but if it is, there should be a a fault code B1232 to indicate the fault.

Force-feeding the a/c system with refrigerant can cause over pressurized system which will severely degrade the life of the variable displacement compressor and its clutch.

 

To close the loop on this thread .....

I took the ML into a local auto air conditioning shop where they hooked it up to their fancy-pants computer. They said there was an error code stored in the All Activity Module. They could not tell what the code was (which I thought was pretty weird). They cleared the code and the air conditioning started working great! The whole thing took about 15 minutes and cost = $50.

Problem solved ..... sorta, I wish I knew what the error code was and what caused it.

Thanks to all of you that offered up your knowledge and time. I appreciate it.

Bill