Climate control system crisis

[michaeld]

I drove from Anaheim to Palm Springs yesterday.  I finished the trip in a tow truck.

As many of you forum faithful know, I have been reading a hot temp gauge for a while.  Three different mechanics have said the car was NOT overheating. 

On Thursday, after about an hour of stop and go driving, my gauge was showing redline.  When I got to my destination, I left the car running and used two temp gauges - one a noncontact unit and one w/ a thermocouple - and came up with the following temp table:

                                          Noncontact            Thermocouple
Upper hose                               176F                         130
Water Pump at 212 switch           200F                         180
Expansion tank                          150F                         136

I neither know nor care why there is such a discrepancy between the readings.  I also took temps of my radiator using my thermocouple, which ranged from 172 to 103 degrees.
The top left at the inlet was 128F, bottom 172.  The top right was 112, bottom 103.

In any event, not too bad, I thought.  I had just finally become totally convinced that my "hot gauge" problem was nothing more than a bad gauge, sending unit, or the wiring between.

Yesterday, I drove back to Palm Springs (I had made the 100 mile drive w/o incident a few days before, stopping after 60 miles for a scheduled stop), and had gone about 60 miles when DISASTER struck.  I started seeing smoke/steam coming from the right side of the hood at the cowl.  Something inside of me cringed and wimpered, and I got off the freeway right quick and stopped the car.

When I opened the hood, there was a lot of steam coming from the climate control module, and the unit was spitting and hissing.  A puddle of coolant began to form.  I checked the expansion tank, and saw that there WAS coolant present - as it was about a third full (which disappeared as the system cooled).
I could find no bad hoses, or obvious signs of damage to the climate control module (no cracks).  I wasn't in to taking chances, so I arranged for a flatbed to tow me the final 53 miles home.

Today, I checked more for signs of damaged hoses or cracked components, and again found nothing.  I added 2 quarts of coolant to the system to fill the 'cold' level of the expansion tank (which implies I got off the road quickly enough to avoid losing a catastrophic level).  Then I had a neighbor start the car while I watched the expansion tank w/ the cap removed to check for bubbles that would indicate a cracked head gasket.  No bubbles.  I also quickly looked for any signs of coolant leakage, and found none.  I did not wait for the thermostat to open, but quickly pulled the car into the garage and shut her down.

Now, there are two possibilities.  The first is that the gauge was right, and me and my three mechanics are idiots.  The water got to redline temp, and the climate control system or heater system couldn't handle the temperature and SOMETHING ruptured  [I was driving with the system turned to "auto low" and the selector dial at 70F].

The second possibility is that the climate control or heater system failed and the "hot gauge" thingy was just a coincidence.  Months back, whenever I drove the car, I would hear the climate control "buzz" like a fuel pump all the time.  Someone said that indicated the unit was near failure.  Well, the buzzing stopped, and the unit worked great for several months.  Just in the last week or so, the "buzz" came back. 

All I know at this point is that 1) I'm not leaking until the car is fully warmed up (and maybe not until it gets VERY warmed up); 2) that the leaking coolant came from the area of the climate control (i.e. passenger side of the engine compartment at the firewall); and 3) the only thing I'd like less than to have to tear the dashboard apart to get to the heater blower/core would be to pay someone to do it.

Has anyone had a leak in this area before?  Where did it turn out to come from?  What advice can you give me?  I'd like to at least TRY to deal with it myself before paying a mechanic the big bucks.

My plan as of now is to:
1) check the water temp sender/gauge to see if the system is working right or not.  If not, replace whatever needs replacing so I can have an accurate and reliable temp reading.
2) Start the car and let it get hot.
3) Try to locate where I see coolant first.
4) Fix the broken thing.

Any wisdom would be appreciated. 
Mike

[jjccp]

Hi Michael,

It's not too often that I get to give info on this site, I usually get it from all the good people here. In this case I think I can help a bit. Many of our friends on this site that are not in the US don't get to experience the joys of our climate control system (nightmare).

Of course, I've been reading your posts about the over-heating problem hoping that some day we would finally know exactly what your problem actually is. Maybe is has something to do with the servo. Here are a couple of suggestions.

I've just had my servo rebuild by a guy in Montreal, Canada and just got it back a couple of days ago. I haven't had a chance to install it yet, but it sure looks nice. He told me that the first test to see if the servo is bad is to run the car, bring everything up to temp, run the heater a bit, then go outside the car and open the hood and have a friend shut off the engine. With the engine off and you standing near the servo, you should be able to hear the servo continue to work (buzz) as it re-parks itself. It might buzz for 20 or 30 seconds and then shut off on it's own. That's pretty much the first test. No buzz = bad servo. In my new 6.9 I was experiencing no heat and no fan as my first symptoms, then the no buzz told me the unit was cooked. It's also a good chance that if the servo is cooked it will also kill the Climate Control Amplifier, but that comes later.

So I yanked the servo out along with the little auxiliary electric water pump. It all comes off as a nice little unit. The servo actually controls vacuum, electrical and coolant. I stripped off the connecting hoses and pump and sent the servo off to Canada. Once I realized how long it would take to go through customs inspections and the US and Canadian post offices, I realized that I would be unable to run the car for maybe a month or more. I took a good look at some diagrams and how the water flowed and bypassed the servo. There's actually enough hoses and clamps left over to splice everything together and the only extra parts you will need are 2 small pieces of pipe to splice in to complete the circuit. I just used copper pipe. Without the servo you get either 100% heat by pushing Defrost or you get 0% heat by pressing Off. The A/C will not work at all.

I'm telling you this because you can simply take off the servo, by pass it and then continue with testing of the engine cooling system and you have eliminated the servo as a potential problem. You might want to consider flushing out the heater core while you have everything apart, just in case. If you eliminate the engine-overheating problem, get the servo replaced or rebuilt. I paid US$277 to have it rebuilt, plus some additional postage costs. Since I haven't installed it yet, I can't tell you if it was a good deal or not, but I might get around to putting it in today. I see used servos on eBay all the time for US$300, but I would be leery of buying used. Mine was already the type rebuild with the aluminum case and the PO told me it was the second replacement he knew of. If mine ends up working I will be glad to refer you the guy that rebuild mine if you like.

I hope that helps.

Jim

[michaeld]

JCCP,
Thanks for your response and insights into "The unique American tragedy" re: climate control.  The units work fantastic when they work; when they fail, they are expensive and complicated to replace.  My understanding is that they were made by Chrysler ('nough said there!).

If in fact my problem is w/ the climate control system, I have to choose between bypassing and replacing.  If I remember correctly, a bypass kit is sold for under $50 that includes instructions and all fittings.  I understood that a bypassed system would lose the A/C and possibly even the heater in addition to the ACC.  It's good to know that you at least keep the heater.  Also available is an electronic module that keeps all the functionality; it sells for about $500 but is guaranteed for five years.  I have seriously considered buying the more expensive unit and preserving my functions, but would want to know more about how well it works before I jump on one.

As I said, I couldn't see any overt signs of damage to any of the visible components or hoses.  But I had a significant leak coming from somewhere!  As I said, when I opened my trunk, the ACC servo was steaming and hissing, and coolant appeared to be spitting from it.  I'd like to know what happened and why it happened.

I'd very much like to know is where the leak I experienced is most likely coming from (i.e. the heater blower, the core, or some specific component in the ACC system).  I've never torn apart the dash to access the heater blower/core, and how to's on that would be nice as well (since the fact that I can't see the damage implies that the damage is in a place that I can't see!).  I have never had to tear a dash apart; I've lived w/ a few issues rather than do so, but I've never done in and never wanted to do it!  I may have no choice now.  Links on DIY for this stuff you've come across would be marvellous.

It seems to me that the first order of business is to verify the gauge and sender.  And next to find and fix - or bypass - the leak.  I have a dye-based leakfinder kit that may come in very handy about now.

In the past week or so, I've occassionally caught a whiff of what smelled like coolant in the passenger compartment when I was idling w/ the windows up.  Then it was gone.  I never found any sign of moisture in the footwells, and my expansion tank level remained fine, so I just filed it away under "weird."

In terms of the "buzz" we are talking about, my issue wasn't that I wasn't getting a buzz, it was that the buzz wouldn't quit!  It was running constantly!

[jjccp]

Mike, you're in luck. I put my rebuilt servo in today and have pics for you.

Servo being disassembled


Servo and pump out of car


Servo bypassed


Servo bypassed


If I was you I would just bypass the servo, drive the car for a month and make sure the engine is cooling correctly, if everything is ok, get the servo rebuilt and reinstall it. You don't need to do anything behind the glove box at this time. You may end up needing an amp back there. For now I would eliminate questionable items and focus on the engine temp.

You don't need a kit to bypass the servo. There's enough clamps and hose leftover to do the job. The only additional thing you need it 2 short pieces of pipe, like the copper pipe I used. You can only see the upper splice in the pics. There is another one that is blocked in the last pic by an electrical connector. See the lower hose? There is another piece of pipe behind that connector with the green wires.

I've considered the Unwired Tools unit too. I talked with a local garage that had installed 3 of them. They charge $1200 for everything. I looked at the on-line instructions for the unit to see if I wanted to try it myself. That's when you need to tear out the glove box. And then make quite a few modifications back there. I opted for rebuilding the servo.

[michaeld]

JCCP,
Dude, you so totally and completely rule!!!

Your "Idiot's Guide to Servo Bypass Surgery" is very helpful - particularly if the servo is behind my problem.  I changed the subject in my post to the thread to help future searchers do their own homemade servo bypass.  A good picture really is worth a thousand words.  Thank you for drawing my attention to the other hose in the lower pic.  I see it.

I would have been afraid to bypass it myself (i.e. on my own w/o guidance) lest I screw something up and make things worse.  Your ingenuity does you credit, my friend.

Maybe you should apply for the title role in the future remake of "MacGyver."

Thank you very much.
Mike

[jjccp]

No problem Mike, your timing is good since I just did mine today. This diagram will help too. Even though this isn't your engine or mine, the coolant flow direction is correct for your car. Just eliminate the servo from the diagram.



When disconnecting the servo, there are a bunch of things to take off. Start with the big vacuum connector on top, take off the little screw first. Then take off 2 electrical connectors. Be careful with theses, they snap/lock in place. Then start loosening up the coolant hoses. There are 2 individual vac hoses directly on the bottom of the servo too. They are each different sizes so you won't mix them up when reinstalling. You also have an electrical connector going to the little aux pump to take apart. Then take out the pump and servo as a single unit and disassemble everything else on your bench. Cut 2 lengths of pipe for splices, you'll figure out the sizes and patch everything back together using the hose clamps and leftover hoses. You can just leave all the connectors sitting there out of the way and still run the engine and heat. You will spill some coolant when doing the job, so have some on hand to top off the system when you have the servo bypassed. 

That should keep you busy for a while. I hope you've found the real problem. I wouldn't worry about what went wrong with the servo, if it's leaking anywhere, it's crap and needs to go away.

Jim

[robertd]

Hi Michaeld,
I recently had the water pump replaced in my 380SL, probably failed due to lack of use.  However after it was replaced the car ran at a higher temperature than before, close to the 100 oC. I checked everything, flushed out the radiator and engine block replaced with fresh coolant. The thermstat checked out ( 84 oC ) . There were no apparent problems, just the car running close to boiling point. In the end I replaced the
(84 oC) thermostat for a(75 oC) thermostat. Now the 380 sits comfortably between 80 and 90 on the temp gauge.
I haven't read all your post so I'm not sure if you have already changed your thermostat or not.
Good luck with it.
Regards Robert

[Papalangi]

When Ben bypassed the servo before I picked my 450SEL up, he used a ball valve per my request and placed it where jjccp's plain copper line is.  That way, I can control the heat by modulating the valve.  I have no problem with cooling when the vavle is full closed. The servo is still in place and the vacuum and electrical connections are still made but the motor was goobered up by some fool (OK, me) who thought he could fix it.  If the AC still held a charge, it would work fine with the way mine is bypassed.

The www.mercedessource.com "kit" is a old fashoned choke knob working a heater valve in place of my ball valve.

The Unwired Tools digital unit completely replaces the servo and amplifier.  I'll be going this route when I can scrape up the funds.

I think you have a leak that opens with heat and pressure.  The casting is prone to crack and leak, thus leading to over heated and damaged V8's

Michael

Michael

[michaeld]

Hi guys.
I appreciate that diagram, JCCP.  It will help me visualize what's going on.

Robertd, the first thing I did when my gauge showed hot was replace the thermostat.  It is my understanding that they are the most common cause of overheating, and they are also one of the cheapest components.  No change.  I used a Behr 75C unit, which was same-same as the unit in my car.

Now, Papalangi, your "ball valve" addendum is interesting.  If I go the bypass route, I would want to maximize the functions I have available.  I would also seriously think about the electronic unit, but would not want to do it w/ the gun I have over my head now.  I can visualize how you would install your ball valve, but not how you would control it's modulation - particularly from within the driver's compartment.  In the famous words of Homer Simpson, "Lord help me, but I'm just not that bright."  I need pictures, diagrams, cartoons, "connect-the-dot" coloring books, that sort of thing.

Now you also said, "I think you have a leak that opens w/ heat and pressure.  The casting is prone to crack and leak, thus leading to overheated and damaged V8s."  Let me make absolutely certain: Are you talking about the servo unit?

It seems possible that the leak is coming from the heater core or one of the hose attachments.  I hope that is not the case (in terms of getting to it) but it is possible.  In favor of the servo theory, when I opened the hood, THAT'S where I saw the lions share of steam, and the unit appeared to be spitting coolant.  Against the servo theory, the unit appears to be completely undamaged from outside the unit.  It is possible that the crack is inside the aluminum cover.

I have not started the car since I got it going just enough to drive it into the garage.  I just haven't mustered up the courage.  I know only that it didn't leak in the two minutes I had it running to get it into the garage.  In that time, the thermostat wouldn't have opened, and coolant wouldn't probably have began to run through the servo system and heater.

[Papalangi]

Hi guys.
I appreciate that diagram, JCCP.  It will help me visualize what's going on.

Robertd, the first thing I did when my gauge showed hot was replace the thermostat.  It is my understanding that they are the most common cause of overheating, and they are also one of the cheapest components.  No change.  I used a Behr 75C unit, which was same-same as the unit in my car.

Now, Papalangi, your "ball valve" addendum is interesting.  If I go the bypass route, I would want to maximize the functions I have available.  I would also seriously think about the electronic unit, but would not want to do it w/ the gun I have over my head now.  I can visualize how you would install your ball valve, but not how you would control it's modulation - particularly from within the driver's compartment.  In the famous words of Homer Simpson, "Lord help me, but I'm just not that bright."  I need pictures, diagrams, cartoons, "connect-the-dot" coloring books, that sort of thing.

Since I don't have incredibly long arms, I just stand out in the rain a tweak the valve by hand.

Now you also said, "I think you have a leak that opens w/ heat and pressure.  The casting is prone to crack and leak, thus leading to overheated and damaged V8s."  Let me make absolutely certain: Are you talking about the servo unit?

Yes, I'm refering to the servo.  The valve section is know to crack and leak.
http://www.duricy.com/~imperialist/AutoTemp/autotemp.html

It seems possible that the leak is coming from the heater core or one of the hose attachments.  I hope that is not the case (in terms of getting to it) but it is possible.  In favor of the servo theory, when I opened the hood, THAT'S where I saw the lions share of steam, and the unit appeared to be spitting coolant.  Against the servo theory, the unit appears to be completely undamaged from outside the unit.  It is possible that the crack is inside the aluminum cover.

If it is the servo and you choose to retain the original system, you may want to find out if George Murphy is still doing servos as he came up with an aluminum valve body to replace the cast part.
http://www.imperialclub.com/Repair/Air/servos.htm , scan down for info about Mr. Murphy.

I have not started the car since I got it going just enough to drive it into the garage.  I just haven't mustered up the courage.  I know only that it didn't leak in the two minutes I had it running to get it into the garage.  In that time, the thermostat wouldn't have opened, and coolant wouldn't probably have began to run through the servo system and heater.

As for your gauge problem, somebody did post a resistance vs temp table for the sender but I don't remember which thread it's in.  It occured to me that you may want to run a wire from the body of the sender to the negative terminal of the battery just to verify that you don't have a bad ground between the sender and the battery.  If there is a bad ground, I would think the jumper would help.  Be aware that if the ground is real bad, when you start the engine your added wire will get hot!  This would be a good indicator the you need to address the ground issue before you do anything else.

[kenny]

servos always crack on the side near the firewall.
I had my third one give up last Christmas while I was on a road trip.  The leak made it look like a servo hose leaking.  Got home and discovered the true location. Bought an aluminum replacement which should avoid the housing crack problems, but there are many things going on in servoland.

[michaeld]

Kenny and Papalangi,
It sounds like you guys think the leak is probably coming from the servo unit, stupid, miserable thing that it is.

I looked around with a mirror, but couldn't see any issues.  However, today I found a smaller, better mirror, and will look around some more.  It would be nice to SEE the leaking component and be able to say, "Ah yes, THAT'S where it's coming from!"

I don't have the original owner's records, but it seems to me he replaced the servo at least four times while he owned the car.  As I was looking at the records - and the amounts he paid for ACC maintenenance - I kind of decided beforehand that I wouldn't be paying THAT particular dancing monkey.  I'll have to revisit all my options now.  I'll have to consider this aluminum-housing unit (which was actually what I thought I had, given the aluminum cover, but I noticed Kenny's old unit had that as well).  I want something that is going to go the long haul.

Now, Papalangi, you said something very interesting (and no, it wasn't the part about your having incredibly long arms!) about grounding.  As you said, there was a post that incluced an ohm-temp relationship that I will use.  If I understand right, the sender unit can be removed from the car, placed in boiling water, and checked on a volt meter, but I'm a little vague on the fact that this is a one pole unit, and you need somewhere to place your negative probe for the test.  (If someone understands this procedure, let me know).  It also appears that I have enough wire to unscrew the sender, leave the wiring connected, and place it in boiling water to check the sender-gauge as a system by comparing the temp of the boiling water to the reading on the gauge.  But now for the part that was interesting: you said

"It occured to me that you may want to run a wire from the body of the sender to the negative terminal of the battery just to verify that you don't have a bad ground between the sender and the battery.  If there is a bad ground, I would think the jumper would help.  Be aware that if the ground is real bad, when you start the engine your added wire will get hot!  This would be a good indicator the you need to address the ground issue before you do anything else."

I've been aware that I have had some kind of grounding issue w/ my chassis for some time now (e.g. my radio and antenna has quit working, started working again, and then quit working.  The servo buzzed incessantly, quit buzzing, and started buzzing again).  I have installed an additional negative battery cable to the alternator, ran a wire from the engine to the firewall, and added a grounding strap from the block to the chassis, but experienced no change.  That was months ago.  Now, if I find that I have a grounding issue re: your jumper wire test, what would I do to to correct it?  How do I appease these whimsical electron fairies?  Electron fairies appear to be girls who don't care for either chocolate, flowers, or jewelry.

And JCCP, I'm still genuflecting in your direction!
Mike

[Papalangi]

The temperature gauge is actually an ohm meter.  The sender is a resistor whose value changes predictably with temperature.  When you turn the key on, +12 is applied to one side of the gauge.  It flows through the gauge and a bunch of probably iffy connectors to the sender and then through the engine and back to the battery.

As the temperature changes and thus the resistance of the sender, more or less current flows through the gauge and the reading changes.  Poor connections or a bad sender will make the gauge read incorrectly.

There are several ways to test the system.

Before you begin, if you don't have a set of leads like these you'll want to get some as they will make life much easier.

http://www.radioshack.com/product/index.jsp?productId=2062660&cp=2032058.2032235.2032304&parentPage=family

You can use a known value fixed resistance in place of the sender.  That way you test every thing from the connector back to the gauge in one fell swoop.  I think it was Arthur Dalton who uses this method of testing.  Ground one end of the resistor to the engine by sticking one lead under a bolt head or crimping on a length of wire etc…  Connect the other end to the wire you pulled off the sender.  You could also use your fancy new test leads.  Using the temperature to resistance table, compare the resistance you used to the gauge reading.  They should be the same within a reasonable tolerance.

To test the sender, you can either leave it installed or remove it.  If you leave it installed, you'll have to deal with a hot engine and awkward access but not have to deal with loosing any coolant.

With the sender installed, pull the connector from it and connect your ohm meter using your fancy new leads or fight having to hold both meter leads in place while trying to read the meter and write down the result.  If you still have access to the thermocouple temperature meter, use one of the alligator clips to hold it to the body of the sender.  The brass body of the sender is the ground side and the pin is the gauge side.  It is not polarity sensitive, i.e. you can hook either the red or black lead to either part of the sender.

With the sender removed, use your fancy new leads to suspend it in water heated to something close to Arthur's temperature VS resistance table.

It sounds like you have done a good job of grounding things.  I don't think you have a problem there anymore.  If you have an assistant that you trust, you can use your voltmeter to check it out pretty quickly.  Set the meter to a VDC range of 10 or so volts, the exact amount isn't important. With one lead on the negative battery post and the other on a safe to reach part of the engine, have your trusted assistant start the engine.  If you see more than a few tenths of a volt, you still need to address your grounding issue.

Have you checked your fuses for corrosion?  That's the number one cause of things like the radio quitting and working again.  At a minimum, spin the fuses in their holders to clean the contacts and if you want to really go at it, disconnect the battery and go in there with a brass wire brush and contact cleaner.

If I get a chance, I'll mock something up to take a picture of.  I can also take a resistance reading on my car as I trust the gauge to be accurate.

Michael,
AKA 250 Coupe and Paplangi

[michaeld]

Interesting event tonight...
I had a ladyfriend over and decided to try to perform Papalangi's test for engine ground.  The lady got in to start the car whilest I held my probes.  She turned the key... and it wouldn't start for her. 

I rolled my eyes, had her get out and showed her what to do and how to read the multimeter, and turned the key.  She fired right up.

I guess my old girl prefers someone to start her like a man rather than have a woman touch her.

In any event, the two (other) results worth reporting are: 1) while the lady was unsuccessfully turning the key beyond the alternator mode, I clearly heard the ACC servo unit gurgling in an unhappy manner; 2) when I started the car, the DC volts went up to .04 and then back down to zero.  So apparently I don't have a grounding issue, and I have much better reason to suspect that servo as being my problem.

All I have on my multimeter are the probe-style electrodes; I'll have to see if they sell the clamping-types in the radio shack store, Papalangi.  You're right; they do look like they'd be good to have.

I'll be playing w/ my sending unit soonest.
Mike

[benzlover]

Does your car have the original all plastic climate control servo?  If so it probably cracked as they do with ago...I purchased a metal base one for my car and it never happened again.

Michael
San Diego
1978 280SE