Autoheat Debugging on an 89 Carrera Coupe
 

Yes another one! I've got a bunch of threads saved off, and just wanted a little advice on some debugging, here's the situation:

1. my fans all run (rear and footwell)
2. my fans vary in speed with the dial (3 speeds)
3. I get heat, and the servo is opening and closing the valves

BUT, the heat doesn't vary it seems. On the first setting it gets quite warm, and doesn't vary heat. This means if i'm driving with heat I have to turn it on and off, mixing with cool are doesn't really work well, or just drive with heat on and windows open (which I do a lot of).

So, I think this is probably:
1. An issue with the interior sensor (or the little fan the pulls are through it)
2. An issue with the rear sensor
3. Possibly an issue with the circuit board between the seats

The bentley manual has testing plans for the thermo resistors.

Would this be the approach you'd suggest? start with these components?

Want to make sure I'm not missing something.

Auto heat.
 

The auto heat in our '86 has nine fan speeds.

Are you sure you have auto heat?

Gerry

I don't see how that's possible. Was the 86 system different? Fan speed is controlled by the resistor that has 3 speed positions.

The dial has 9 positions, sure, but those aren't 9 fan speeds. There are 3 speeds. Positions 1-7 low fan
8-9 - med fan
defrost - high fans

And yes, I'm 100% sure I have auto heat. No levels and it's on the options list/sticker.

I have the same exact issue. Even in the #1 position it gets too hot and I have to turn off the heat after a bit. Any help troubleshooting this would be greatly appreciated!

Happy to post any findings :-)

it has 9 + DEF settings for temp but 3 different speeds.

Best guess is that your interior sensor is bad. Does the little motor in it run?

Auto Heat.
 

I have the factory CKT diagram if anyone needs it.

Gerry

I've had the same problem. Without taking it all apart and testing the potentiometer with a multimeter, I've tinkered with the adjustment screw under the control knob - the result always seems to be that the heat falls way off which works fine on moderately cold days, but is too little heat on really cold days. Probably need to get to the leads and set the thing to the correct setting.

the fans ONLY have 4 speeds. off, low, med and hi.
pull the fans out anyway and clean them up per the link below. not hard to do.

if your fans run with the heat off, the micro switch in the controls is going.
I need to do more testing as I have been informed that the autoheat I suppose to very the heater box flaps in between full open and full closed. this will NOT run the fan unless the flapper is fully closed.

here is a link with more info than you probably ever want to know about auto heat.

Gerry, if you have some diagrams of the actual CKT board in the auto heat I would like them,
tyc@sc.rr.com

http://forums.pelicanparts.com/911-930-turbo-super-charging-forum/941050-auto-heat-foot-well-blowers-part-deux.html

Will be tearing into the car today, picked up some thermo-resistors from the local electronics shop over the weekend 'just in case'.

I did confirm the fan in the interior sensor is working/ running. Along with all 3 heat fans, and the servo. The question is heat variation and calibration. I also have a passenger-side footwell fan that is a little 'squeaky' on startup 'cold' so i appreciate the link on the fans. Will try a lube job before replacement ($325 ugh).

Question: My understanding is the auto-heat ONLY adjusts the valves, not the fan speeds, correct?

no, it does the fan speeds.
no fans until the arm/flaps are full closed. then lo speed based on the switch with the red/blk wires. that also provides power to run the fans on the other 2 speeds.

there are 2 switches that operate from the dial. one is med at settings of 8,9 (last 2 before DEF). then hi for DEF. there is a separate relay for hi. the switch in the controls only provides a ground from the relay BUT the power is from the RD/BLK switch. so that switch HAS to work for any fans at all.

I tested mine without the temp sensors connected so the arm was either fully open or closed. no in between. I do want to go back and test to see if it does regulate the opening of the heater boxes. but as soon as the arm comes off full open the fans go off and it is only air from the engine fan pushing heat in.

my fans were very noisy, cleaning them up fixed it.
also make sure the cable housings for the heater boxes are clean and lubed. it takes a lot of force to close the heat.

I don't mind turning the heat on and off as needed, I did it before but mostly trying to get the fans to run.

ok i got it.

So the sensor will rotate the servo which also rotates the mechanism to trigger the switches.

I actually saw that playing around with it today, where i turn to 0 but fans run as the servo closes the valves and rotates the microswitches.

I'm determined to reverse engineer this pinnacle of 80's electronics engineering :-)

Test the resistance of both sensors at the potentiometer in the autoheat box.

The procedure is in Bentley. You'll know based on these values if one of the sensors is bad.

My caloric sensor was bad (I think it was open circuit). I drilled out the old one and used a $2 thermister from digikey. Works great now.

Here is the link to replacing the caloric sensor:
http://forums.pelicanparts.com/porsche-911-technical-forum/584237-caloric-feeler.html

as soon as the arm comes off the limit switches the fans should turn OFF.
if they don't the switch with the R/B wires is going bad.
if it goes bad either your fans will run all the time and or they may eventually stop working, which means you are pulling it out again.

I utilized the limit switch for both purposes. to run the fans AND limit the arm. you can try to find the switch and just replace it if it is going bad.

Ok confirmed:
--at turning on system flaps open all way at setting 1, fans switch to low
--all 3 fans function and vary in speed (off, low, med, high) (micro switches good)
--confirmed limit switches are good
--confirmed calibration of knob in spec
--servo functional and moves flaps, but at setting 1, opens fully
--confirmed flap movement is smooth and not difficult/binding
--verified cabin sensor reads to spec + fan in that sensor functions
--Confirmed the exhaust sensor/caloric feeler is "open"/bad!

For now, since I don't feel like jacking the car and drilling this thing out, i'm going to trick it to thinking the car has a 'hot' exhaust and wired in a 200 ohm resistor (sounds like spec is about 135 ohm at 100c, but I have some 200s laying around...) at the control unit for now. this will give me some variability in the heat through all the settings on the control knob. The interior sensor will also give some variability. Basically at 1, fans go low and flaps open a little, at 6 flaps are fully open (this is at about 18c in my garage... Socal! :-)

Will test drive in a few.

were these test done with the exhaust temp sensor connected?

what I have been told and suspected the way the system works is if the exhaust is cold, the heat does not come on.

try it without connecting the exhaust temp sensor. only thing that may happen is the heat will come on when the air is still cold.

Tested many variations with a potentiometer to simulate the resistor.
Front resistor tested good.
Keep in mind front resistor was reading about 18c (1.9k ohm)
When rear resistor open or resistance is high the arm opens all the way
When it is low it opens less.
I originally thought it might work same way as you did, when temp is low in exhaust limit the cold air to the cabin.
What I found was opposite. When rear cold it opens more and as resistance drops (heats up) it closes the flaps. This could be logical too as a Porsche heats up very fast writhin a minute or two you have heat.
Also, this resistor will reach min value rather quickly and just stay there due to exhaust heat
The cabin resistor is much more critical I think.
Anyway what I tested is with the cabin resistor constant st 18c the rear jumped at 200 ohm I get 6 distinct steps of variation with the knob calibrated at 950 ohm at step 5.

One other thing I found is that the servo movement triggers fan on and off at or off positionn0
But other the. They the servo does not vary the low med high speeds. That is done with manual movement of the knob.
What the servo will do is open and close the rear valves which has the same effect of diverting less air to the cabin as it dumps out the valves. But the footwell fans are always low med or high based on the knob.
However the servo does trigger the fans to off when returning to the closed position.

It got late and was raining last night so didn't test drive it but will do so today later and report back.

Cool, sounds like you've got it isolated. Check out my link above on replacing the caloric feeler.

You may not have to drill out the rivets. You may find (as I did) that the metal tube that contains the actual sensor is not actually connected to the piece riveted to the flapper. Rather, the tube has a crimp in it about mid way, and that crimp sits in the diamond cut out shape of the riveted piece.

You will probably find that the hardest part of all this is drilling out the old sensor from the metal tube. I have no idea what that black material is, but it sure is good at dulling drill bits.

Ok cool. Well after test driving the car just now (granted i live in SoCal and it's 55 outside, but that's chilly for us) I'm not sure I'm going to bother. With the resistor wired into the console to the caloric feeler plug mimicking a 'hot' engine all the time, the system seems to be working as designed otherwise and the internal sensor is regulating the temp.

Where before on 1 i'd get heated out of the car, with the windows up and 55 outside at 1-2 the car is comfortable. With the windows down 4-5 is comfortable. This is great!

Here's behavior I'm seeing now:
--first start of car on position 1, in a minute or two warm air, with windows closed, within a few minutes it was warm inside the car and the valves nearly shut off the heat (internal sensor regulating)
--turn to 2-3-4-5-6 i get increasing amounts of heat
--windows up with setting on 3, car gets quite warm then the heat will be closed off as cabin warms up (internal sensor regulating)
--windows down, 1-2 the fans/heat are quite low. But 4-6 it is plenty and stays that way as the internal temp is not 'heating up' since windows are down

Anyway, I have a couple thermistors I picked up at the local electronics shop and next time I have car up in the air I'll give a look at the sensor to make it fully 'right'.

its just nice to not be heated out of the cabin any longer. Will take wife for a drive down PCH in a little while and see if it meets her approval :-)

And... it passes wife test.
Mid 50s out. Setting 1-2 with windows up temp stays perfect in cabin
4-5 with windows down
6+ and it's cranking hot.

Able to just set and forget.

awesome.
not sure if you were clear, but the fans DO NOT come on until the arm is in its full position which would be max heat from the engine.

about the flapper boxes.
with my flaps closed that is max heat. full open is no heat. this can get confusing. when my flaps are open the heat just blows to the out under the car.

I would just jumper the rear sensor so it thinks the engine air is hot all the time.

I cant wait to get my car back on the road to see how mine works.
I will test my sensors to see what they measure. I can also use a heat gun to heat up the rear sensor and test it.

I was using term opposite. Open = heat. But I get what you mean.

I'll evaluate the motor switching again today.

It's just nice to have real variability on the heat and the "autoheat" part actually works.

The two temp sensor values and the preset potentiometer (under control knob) value (200 ohms)
is critical for proper operation. If either of the sensors is not connected or bad, the unit operates
open-loop, and the motor arm just moves from one limit point to another as the knob is turned
from off to the first position and then back to off. It's a very simple system that most mis-diagnose.

This isn't entirely accurate. BAD means its either OPEN ( no connection) or CLOSED (0 ohms).

In the testing I did with a variable resistor (potentiometer) wired to the rear sensor plug, at OPEN (which mine tested as, no resistance at all) the control arm DOES move to the full OPEN/full heat on setting 1. However, at any resistance below about 300 ohms, including to full short, the control arm is variable, and does vary based on input from the interior sensor and the top side knob.

So, if its not connected, or fully open: flaps go full heat on setting 1
If shorted or very low resistance, the flaps will be variable.

All I did was wire in a 200 ohm resistor at the sensor for now.

I suspect the majority of failures of the rear sensor result in an open circuit, therefore auto heat has no variation.

Really? You found that definition doing a Google search, right? Keep "stabbing in the dark" as you
attempt to troubleshoot. You might eventually need to buy another unit.

Are you attempting to be an @ss or does this come naturally?

I was describing how I was using the term. If you need further reading comprehension help I'm happy to assist further. SmileWavy

There is no stabbing buddy. This system is not complex in the way it functions.

Great thread.
I am troubleshooting mine.

Can you explain to me how I measure the caloric feeler under the control knob?

The manual says turn knob to 5 remove knob and measure resistance, should be 950 plus minus 50.

But I can not see how I measure ohm on this knob as it is just a screw in a brass nipple surrounded by plastic. what am I missing?

The terminals are inside the autoheat box. So you need to remove the box from the car (which means you need to remove the passenger seat to disconnect the servo rod) to get to them.

this is what I did, measured it at the wires, you can seem them pretty clearly.

In my case, I just jumped the circuit as I didn't want to bother replacing the sensor in the exhaust.

It worked fine and regulated the heat well.

Thanks,

I got the factory specs when measuring except for the thermistor in the heatexchanger.
I got 350 ohm instead of 1500 at 20 degrees celcius.

Should that be replaced?

Michel

What symptom are you trying to fix? You really need to check all components. The rear caloric feeder simply allows the system to pass more heat until the car warms up then you have more valve regulation. If your rear sensor is at 350 ohms then you will likely get full control of auto heat. At least thst is what i found at 200 ohms the system worked like a warmed up csr and regulated temperature.

This was a few years ago and i don’t own the car any longer....