Automatic heater control unit repaired -- 70 cents!

[logician]

Hi,

I thought that the group might be interested to learn how I successfully repaired the automatic heater control unit in my '86 Turbo for 70 cents (plus 10 hours debugging time to trace the circuit...). New replacement units cost $1100 from Porsche, and rebuilt ones are apparently not available.

So the heat in my '86 turbo never worked. The symptom was that the servo in the control unit was stuck in the off position and wouldn't ever open the heating vents. I checked the two temperature sensors for proper resistance as indicated in the Bentley and Factory manuals, and they were within spec. I eventually came to the conclusion that the control unit was somehow fried.

I took the unit out, and disassembled it to extract the circuit board that is plugged into the servo, i.e. the motorized piece that drives the lever to open the vent. My circuit board has a ceramic assembly in the middle that contains one 8 pin dual op amp IC and four miniature SOT89 transistors, all surface mount. The four transistors drive the motor on the servo. I checked the resistance on all of the transistors, and found that one was different and had shorted. So I replaced it, and now works!

The unit in my Turbo was apparently a more modern "hi-tech" unit that uses miniature surface mount transistors. The transistors are just 4mm wide, and the leads are only 1mm appart, so you need to work to about .25mm tolerance -- very tight. Reworking the board was difficult, and likely impossible unless you are experienced in reworking circuit boards. One complication was that I had to grind down the transistor package so that I could remove the leads one at a time, as it was impossible to heat all three leads hot enough to remove the transistor as one unit. I had an assistant help position the replacement transistor as I soldered it down.

Fortunately, the earlier control units seem to be built with older through-hole transistors that have much larger lead spacing, and that have leads that can be clipped individually to facilate removal. Reworking these older units is probably much easier, so if that's what you have, you might be in luck!

There were two types of transistors on the ceramic board. The ones labeled "AA" are Infineon BCX51 transistors, while the ones labeled "BA" are BCX54. I ordered replacement transistors from Digikey (www.digikey.com). I used ZETEX BCX5216 and BCX5516 as replacements for the AA and BA transistors, respectively. Just for reference, the OP amp was a TAA2765.

I suspect that power transistor failure is a common cause of failure in these control units. The way that the circuit is designed, I think that the motor won't really overheat. However the transistors have to dissipate 1W of power without having good heat sink. This is especially true of the older units, as the power transistors don't have any heat sink at all, and this is probably why the later unit was redesigned to mount the transistors on a larger ceramic substrate.

The total parts cost was around $10 -- 70 cents for the transistor, and about $9 for shipping and extra handling charge for order under $25. If you repair your control unit similarly, please post your experience. If you have an older unit, please post the transistor types for those units and where you were able to get the replacement part.

So now I have heat. Next thing is to fix all the oil leaks...

-Juan

[speeder]

Sounds like a potential Pelican tech article to me. Nice work.

[APKhaos]

Achieving succesful diagnostics, appropriate component level replacement [and appropriate substitution, and sourcing] is teh mark of a bona fide zen master of the 911 auto heat controller.

Some pics and a tech article would immortalize your name in the Pelican world, and would probably be a lead pipe cinch for the next edition of 202 Projects.

Nice work, Juan.

[logician]

Hi,

As requested, here are pictures from the repair of my heater control unit. First, a picture of the servo inside of the control unit. The servo is held in place by four screws and is easily removed from the control unit. Note the circuit board.



This is a picture of the circuit board after unplugging it and removing the black plastic shield that covers the ceramic subcircuit.



The ceramic subcircuit has a dual op amp IC in the middle and four power transistors around it. You can see that two of the transistors are labeled BA, one is labeled AA, and one is labeled AM. The one labeled AM is the AA one that I replaced.

Hope to hear of other successful repairs! Actually, I would really like to see some pictures of other servos, as I have one unanswered question about my unit that might be answered by seeing another servo. If you have a disassembled automatic control unit, please post pictures of the servo and circuit board with the same views as my pictures. Thanks!

-Juan

[Kurt peterson]

My 87 Turbo has the same problem (no heat) and my printed circuit board looks exactly the same as yours. I also have had my system apart and have verified that my temp. sensors are within spec. , my rotary switch is OK and the micro switches are working properly. I did have one bad micro switch in the control unit, which I replaced, but it only provides power to the aux. blower fan relay when you are in position 10. I have verified that if I manually move the flapper motor then power up the system, the motor moves to the off position and stays there so I know the motor is OK. I checked the resistance of my transistors and they are all the same but if I switch my leads around then I get an open circuit on one set (AA and AB) of transistors (each set appear to be wired in parallel).

What is the resistance of the transistors supposed to be and should it make a difference which lead of the tester is on the base and which is on the emitter? Also, is there a way to test the IC? The circuit board is available through Porsche for about $300 but if I can solve the problem with a few transistors than I'll be a happy camper. Any help would be appreciated.

[Rot 911]

Juan, excellent posting. This is what this list is all about, inventive solutions to problems. Wayne, this really should be a tech article!

[APKhaos]

Kurt P,
Testing the transistors while they are in circuit can be misleading because the other paths in the circuit can cause false readings. This circuit looks simple enough for that effect not to be a problem. Yes, it makes a difference which tester leads are used for the base/emitter & base/collector test. But it also makes a difference depending on whether it is PNP or an NPN transistor, so to keep it simple:

VOM set for mid range resistance test.
Test base to emitter, then reverse the leads and test again.
Test base to collector, then reverse your leads and test again.

You should see around 100 ohms one way, and around 100K ohms the other way. Same for both tests.

If the transistor tests bad, then de-solder carefully, then test again out of circuit to be sure. There is no simple way to test the op amp - replacement is the simplest test.

[Kurt peterson]

Well, I did the checks and here's what I got:

NPN PNP
Base to Emitter one = 4,000,000 both = 9,000,000
one = 9,000,000

Base to Collector both = 300 both = 4,700

It didn't matter which way I had the leads, the readings were the same either way.

Yes that's 9 million ohms. I only get 2 million through my fingers!!
I would say they are bad if I didn't get the same reading on all 4 transistors (with the exception of the one reading on the B to E reading on one of the NPN transistors but 4 or 9 million is still a bunch and at that sensitivity, the meter moves a bit).

Any suggestions?

[logician]

Hi Kurt,

> My 87 Turbo has the same problem (no heat) and
> my printed circuit board looks exactly the same
> as yours. I also have had my system apart and>
> have verified that my temp. sensors are within
> spec. , my rotary switch is OK and the micro
> switches are working properly. I did have one
> bad micro switch in the control unit, which I
> replaced, but it only provides power to the aux.
> blower fan relay when you are in position 10. I
> have verified that if I manually move the
> flapper motor then power up the system, the
> motor moves to the off position and stays there
> so I know the motor is OK. I checked the
> resistance of my transistors and they are all
> the same but if I switch my leads around then I
> get an open circuit on one set (AA and AB) of
> transistors (each set appear to be wired in
> parallel). What is the resistance of the
> transistors supposed to be and should it make a
> difference which lead of the tester is on the
> base and which is on the emitter?

Checking resistances on the transistors is a little ambiguous because there is other circuitry in the path. So while an unconnected transistor should show diode behavior (open one way and low resistance the other way) between the base an emitter, it won't necessarily show this when it is in a circuit. Had I measured the resistances and recorded them, then those measurements might be useful. But I didn't.

One other complication is that the resistances measured on transistors are dependent on the voltage that the ohm meter uses to measure resistance. If the ohm meter uses a voltage higher than .6v, then the transistor can become forward biased and will show lower resistance in one direction. I was fortunate, in that the way that my circuit failed resulted in a short circuit between the base and collector, and this is an unambigious indication of a failure.

Fortunately for you, the circuit is symetric, so you can compare the left and right sides. The way that the circuit works is that the two transistors on the left side of the board (using my picture as reference) drive one side of the motor, and the transistors on the right side drive the other. The difference in function between the upper transistors, labeled AA, and the lower BA transistors, is that the lower transistors pull down (ground) their side of the motor, while the upper ones pull up (connect to +12V) their side. The idea is that when the upper transistor is turned on in the left side, while the lower transistor is turned on at the right side, the motor is driven one direction. To reverse direction, the upper transistor on the right side is on, while the lower transistor on the left side is on. BTW, the circuit is in and emitter follower configuration, so the emitters of the transistors are connected to the drive side of the circuit, and the collectors to power and ground.

So, you should compare the resistances of the transistors on the left side to the similar transistors on the right side. Notice that the transistors on the left side are upside down, so the positions of their leads are reversed. Measure the resistance between all three leads and make a note of this. Try to get a good digital ohm meter so that the resistance does not change when you reverse the leads. If your Ohm meter has a "diode" setting, don't use it. What we are looking for is one transistor that has resistances that are lower (by more than a factor of 10) than the others. I think the resistances should be roughly the same between the left side and right side circuits.

One other thing to measure is the resistance between the driving side of the circuits and ground and power. The driving side of the circuit is the the left most pin of the upper left transistor, and the right most pin of the upper right transistor. Power is the center pin of the upper transistors, and ground is the center pin of the lower transistors. What I am looking for is whether the driving side of the circuit is shorted to power or ground.

Please post your results.

> Also, is there a way to test the IC? The
> circuit board is available through Porsche for
> about $300 but if I can solve the problem with
> a few transistors than I'll be a happy camper.
> Any help would be appreciated.

Actually, $300 for just the circuit board sounds like a reasonable deal for the effort. My Porsche dealer told me that the only option was to replace the whole control unit at a price of $1100. I would be curious if you would be able to purchase the circuit board.

Here are links to data sheets for the transistors at the infineon (Seimans) site:

> http://www.infineon.com/cgi/ecrm.dll/ecrm/scripts/public_download.jsp?oid=9127

> http://www.infineon.com/cgi/ecrm.dll/ecrm/scripts/public_download.jsp?oid=9110

One other question for you that would help me. Can you look at the picture of my servo and compare to your servo? Specifically, on mine, you will see that there are three microswitches. Two are on the lower part at 6 O'clock, and one is at 9 O'clock. The two at 6 O'clock are stacked on top of each other. Are your micro switches configured the same way, or do you have two microswitches at 9 O'clock, and one at six? Actually, if you could post a picture of your servo unit taken from the same perspective, that would be very helpful to me.

Thanks!
-Juan

[Siena911]

Juan,

Fantastic article, you have probably saved me 10 hours work

I can't believe it, I have just pulled out my heater control yesterday, to check it out and have it in front of me as I type. It is from a 1981 SC and it is slightly different, I will attach pictures of it here (if it works :-))

I will be measuring the transistors later on this afternoon.

Thanks for posting a very informative article.

Cheers Jakes

[Kurt peterson]

Attached are photos of my control unit. It's the same as yours except I had to replace one of the micro switches (the top one of the stacked pair) and the terminals are a little different. The switch I had to replace sends power to the aux. blower fan relay when you switch to #10 position.

In my last post I included most of the values for my transistors. The only value I didn't include was the collector to emitter. The reading is 9 mil. ohms on all 4, the same as the reading between my base to emitter. I checked for a short to ground and power but I don't have one. All 4 transistors give the same reading so I'm thinking that they are not the problem. I measured all the diodes and resistors and all seem fine. That only leaves the IC. Correct me if I'm wrong but I don't think you can check them can you? I'm using a new Snap-on DVOM so I have to assume the meter is fine.

I verified with two dealers that the board is available separately for about $300.

My micro switches are OK, my rotary switch is OK, my flapper valve position sensor (the round 3 wire sensor on the control unit with the gear on one end) is OK and my temp. sensors are OK.

I'm running out of ideas other than the IC.

Just tried to upload the picture but it's too big, I'll need to take a lower res. picture tomorrow if you still want it.

[logician]

Hi Kurt,

> Attached are photos of my control unit. It's the
> same as yours except I had to replace one of the
> micro switches (the top one of the stacked pair)
> and the terminals are a little different. The
> switch I had to replace sends power to the aux.
> blower fan relay when you switch to #10
> position.

So the microswitch that you replaced was on the servo unit, or was it on the manually operated rotary switch? The #10 position on the rotary switch doesn't directly affect the microswitch on the servo. I would still like to see a picture of your servo unit, so please post photograph taken with same vantage point as my picture, so that I can see the postition and stacking of the microswitches.

> In my last post I included most of the values
> for my transistors. The only value I didn't
> include was the collector to emitter. The
> reading is 9 mil. ohms on all 4, the same as
> the reading between my base to emitter. I
> checked for a short to ground and power but I
> don't have one. All 4 transistors give the same
> reading so I'm thinking that they are not the
> problem.

Right, so the resistances sound reasonable. The NPN 300 Ohm resistance is due to a 300 Ohm resistor between the base and collector. I think the 4700 Ohm resistance between base to collector on the PNP is probably due to current runing through two resistors to get to the power rail, and then back to ground.

> I measured all the diodes and resistors
> and all seem fine. That only leaves the IC. Correct me if I'm wrong but I don't think you
> can check them can you?

I don't think you can check the IC simply by using an Ohm meter. So the other choice now is to apply power to the servo and see how it behaves. You can check the bias voltages on all the transistors, and try tweeking the sensor input to see what happens.

You know that the control unit functions in one direction, to close the heating vent, so it is not completely burnt out. So I think these are the choices for failures:

1) The IC or other circuitry on the board is broken.
2) Your assessment that the temp sensors were OK was flawed, and in fact the control unit is being told to shut down.
3) The wiring in your rotary switch is broken (I haven't traced that out yet).
4) One of the power transistor failed by open-circuit instead of short-circuit.

I would do the following:

1) Recheck your evaluation of the temperature sensors.

2) Check rotary switch and make sure that it is working correctly. The Bently manual describes how to adjust the potentiometer in the rotary switch -- setting it to 950 ohms. Have you gone through the Bently manual debugging procedure?

3) If that fails, then I think your best bet is to try powering up your servo and measuring voltages. It is pretty easy, as two pins on the 4 pin wiring harness supply power, and the two pin harness that has the yellow and gray wires (at least on my unit) is the combined input from the temp sensors. When I supplied power to my servo with the two pin sensor harness disconnected, the servo retracted to the shut position. I think that pulling up or pulling down the yellow lead is what will cause the servo to move to the open position. But I didn't try this. If you do that, then use a resistor -- try 1K first, and then 100 ohms if you get no movement from the 1K -- rather than simply wiring it to +12V or ground. You want to make sure you don't burn out any circiutry.

> I verified with two dealers that the board is
> available separately for about $300.

Interesting. Your location indicates Tokyo Japan. Is that correct? I wonder if that might be the explanation for the part being available. Can you tell me what the Porsche part number is for your circuit board? Also, can you look at the bottom of your servo, and tell me what is the Porsche part number for the entire assembly?

> Just tried to upload the picture but it's too
> big, I'll need to take a lower res. picture
> tomorrow if you still want it.

Yes, please send picture.

-Juan

[Kurt peterson]

What I meant by the micro switch comment is that the top switch in your stack of 2 is the power to the relay that powers the aux. blower. When you go to 10 on the switch, I assume the logic in the system tells the servo to open the flapper fully. When the flapper is fully open, the arm on the servo closes the micro switch and it sends power to the relay. I think they made the circuit overly complicated but........

I checked the sensors several times and yes I did go through the Bentley procedure to adjust the base resistance of the rotary switch.

One thing I haven't done is check voltages etc. on the unit when the car is running. I think I'll do that this weekend. Is the motor a 12V motor? I see a large resister in the power circuit and the last thing I want to do is burn out the motor.

Yes I do live in Tokyo but I don't buy any of my parts her. The 930 was never officially imported into Japan so I doubt the dealers would have any parts. Even if they did I imagine the prices would be outrageous. Even at Y130/$1 I still pay $12.30 for a quart of Mobil 1 - and that's at a discount auto parts store!!

I checked the board at Hendricks Porsche, Stoddards Porsche in Ohio and some dealer in Houston. Hendricks was the cheapest but they won't ship international.

Not sure what I'm doing wrong but at the lowest resolution I can shoot at with my camera, the file is 81K and it won't send. I'll let you know how my checks go this weekend.

[logician]

Hi Kurt,

> What I meant by the micro switch comment is that
> the top switch in your stack of 2 is the power
> to the relay that powers the aux. blower. When
> you go to 10 on the switch, I assume the logic
> in the system tells the servo to open the
> flapper fully. When the flapper is fully open,
> the arm on the servo closes the micro switch
> and it sends power to the relay.

Did the heat ever work for you? Can you describe it's operation? Do the fans operate at multiple speeds, or only when you set the switch to the 10 position? I believe the behavior should be:

0 fans off
1-7 fans slow speed
8,9 fans medium speed
10 fans high speed

Actually, my heat never worked right, and I can't see how it could have ever worked right with my servo unit, which btw, seems to be wired slightly different than yours as you describe...

> I think they made the circuit overly
> complicated but...

You should trace out the schmatics for the circuit board! Alternately, check out the factory wiring diagrams for some real fun...

> I checked the sensors several times and yes I
> did go through the Bentley procedure to adjust
> the base resistance of the rotary switch.
>
> One thing I haven't done is check voltages etc.
> on the unit when the car is running. I think
> I'll do that this weekend. Is the motor a 12V
> motor? I see a large resister in the power
> circuit and the last thing I want to do is burn
> out the motor.

The motor has a resistance of 33 ohms, so 12V would drive about 400ma, which I'm pretty sure would be too much. The transistors are rated for 1W, so you proably don't want to put much more than 200ma through them, assuming that each one drops about 4V, and the motor drops 6V.

When you check the voltages, should see that with the heat turned off, the base voltages on one side should be within a volt or two of the other side. The circuit is actually linear, so I think it always drives a little bit of current through the motor, but not enough to make it move. When you turn the heat on to full blast (10), the base voltages should switch the other way and differ by several (6?) volts. Actually, I think that turning the heat to 10 is a useful test because it bypasses the temperature sensors, so the servo should operate even if the sensors are broken in some way.

Now that I'm thinking about it, you should check two other things. First, there are two diodes on the board. Check that those work, i.e. low resistance one way, and high the other. The current driving the motor one direction goes through one of those diodes, and going the other direction goes through the other. If one diode was bad, you would get the behavior that you are seeing -- that the motor operates only in one direction. The second thing is that you should check all three microswitches on the servo. Two of the switches interrupt power when the servo reaches its maximum position. If one is broken, then I think it would prevent it from going one direction.

> I checked the board at Hendricks Porsche,
> Stoddards Porsche in Ohio and some dealer in
> Houston. Hendricks was the cheapest but they
> won't ship international.

Can you send a pointer to the Hendricks web site? I couldn't find it.

-Juan

[Kurt peterson]

My heat has never worked since I bought the car. I had other things to worry about and I actually bought the car when I lived in the Middle East so heat was the last of my worries.

My blowers only work in the 10 position so yours must be wired differently. The owners manual says the turbo and non-turbo work differently. I think it said that on the turbo the fans only run on 10. The way mine is wired is that the power side of the coil in the relay is fed from the micro switch that closes when the flapper is full open and the arm closes the switch. However, a micro switch attached to the rotary switch operates the ground circuit for the coil in the relay and that switch only closes in the 10 position. So even though I have the fan resistors in my car, they don't do anything. I assume it was easier to leave the resistor block in all the cars than to have a different harness and build procedure for the turbo and non-turbo. I hot wired the relay through the connertor at control unit and the fans worked. I have checked all the micro switches and with the exception of the one that I replaced, they all work fine. I checked the diodes on the circuit board and they are fine too. The #%^&** thing should work. You know what they say, when nothing else works, hit it with a hammer!! Time to break out the ball and peen. Thanks for the advise on checking the system, I'll try on Sat.

Hendrick Porsche is www.hendrickporsche.com. The guy I e-mailed was very helpful and the price he quoted me for the board was $180. At his point I think I'd pay $180 if I knew for sure it would fix my car. Hopefully I'll know more this weekend.

[subdewd]

I've done some board work on surface mounted components
and found you can use a heat gun to remove components.
Just make sure to localize the area being heated
and go slow. Also some chip heads told me Digikey
will send samples........

[logician]

Hi Kurt,

> My blowers only work in the 10 position so
> yours must be wired differently. The owners
> manual says the turbo and non-turbo work
> differently. I think it said that on the turbo
> the fans only run on 10.

Very interesting. So in my case the original problem that I was having before my circuit board fried was that the fan would turn off when the servo reached its maximum position. The thing that I have not been able to figure out is how the circuit could have ever worked, as the microswitch actually turns power off to the fans and relay when it reaches the max position.

So one possibility is that it was always broken and that the microswitch was miswired at the factory. Do you recall whether the microswitch is on or off when the maximum position is reached. In my case, the max position turns the switch off -- no current flows.

The other possibility, although I can't completely explain the behavior, is that the wrong fan relay is installed, and perhaps it should be off when the solenoid is powered, instead of on. I'll check the relay part number later today.

> The way mine is wired is that the power side
> of the coil in the relay is fed from the micro
> switch that closes when the flapper is full
> open and the arm closes the switch. However, a
> micro switch attached to the rotary switch
> operates the ground circuit for the coil in
> the relay and that switch only closes in the
> 10 position. So even though I have the fan
> resistors in my car, they don't do anything.
> I assume it was easier to leave the resistor
> block in all the cars than to have a different
> harness and build procedure for the turbo and
> non-turbo. I hot wired the relay through the
> connertor at control unit and the fans worked.

Sounds like you've done as much tracing out of the screwey wiring as I have. Do you have turbo schematics, or are you simply figuring it out? I've been using schematics, as the previous owner had purchased the Turbo shop manuals. Actually, they aren't that expensive.

> I have checked all the micro switches and with
> the exception of the one that I replaced, they
> all work fine. I checked the diodes on the
> circuit board and they are fine too. The #%^&**
> thing should work. You know what they say, when
> nothing else works, hit it with a hammer!! Time
> to break out the ball and peen. Thanks for the
> advise on checking the system, I'll try on Sat.

You know, it doesn't really sound like the circuit board, although for $180 it sounds like it would be worth the try at this point. The behavior that you describe is consistant with the servo signal wires -- the two pin connector -- being open. That forces the servo to go to full off.

One other thing. Do you know that the mechanical linkage is not binding and holding it closed? That seems likely if the heat was not used for many years. Actually, the heater flaps have really strong springs on them, so it takes a surprising amount of force to move the mechanical lever that the servo pushes on, and it may be difficult to distinguish between the spring force and the flaps binding.

-Juan

[Kurt peterson]

Juan,
I stand corrected. I hooked everything up today and the aux. fan does run at very low speed in pos. 0-7, 8&9 is a little faster and 10 is full blast. It only works that way, however, if the motor is driving the flapper full open and the arm closes the microswitch which energizes the aux. fan relay (and also sends power in a parallel circuit to the fan resistor block to give you low and med.). If I turn the rotary switch to off, the motor starts moving toward the closed position. The owners manual says only the Carrera works this way and the turbo aux. fan only works in "max heat" which I take to mean position 10 on the rotary switch. Having re-looked at the wiring diagram, I now see how you get low speed fan in 0-7. The problem is that since my heater has never worked since I bought the car, I don't know how it's supposed to work and I don't know if the logic of the system drives the flapper to full open and leaves it there or if it modulates. I assume since it's supposed to be auto temp. control than it modulates and probably only drives the flapper to full open when the car is very cold inside, hence the owners manual statement about max heat. Not sure.

The bad news is that it acts the same as before. If I manually move the lever to the full open position, my aux fans run and the controller will leave the flapper at full open until I turn the rotary switch to off, then it closes the flapper. If I turn back to 10 on the rotary switch, the motor just sits there, it doesn't try to drive the flapper open. When the motor is closing the flapper I have voltage on the red wire going to the motor but when I go to 10, I don't have voltage on the black wire which tells me that either my one set of transistors is bad or the IC is bad. If I had voltage on the black wire then I would think the motor is bad but I don't.

My micro switch sends power to the relay when it's closed, that's the only way to get max fan speed. The micro switch also sends power to the resistor block, through both resistors, which is how you get the lowest fan speed in 0-7. A micro switch on the rotary switch sends power to the resistor block, but only through one resistor, in positions 8&9 which is how you get med. fan speed.

My linkage is fine and the flappers move freely, but you're right, there is a lot of spring pressure to overcome to open the flapper. That explains the elaborate gear reduction set up on the controller.

Looks like it's time to spend $180 to get a new board, unless I want to try 2 transistors and an IC. I don't have a heat gun here nor do I have a small soldering iron or experience working on printed circuit boards.

[logician]

Hi Kurt,

> I stand corrected. I hooked everything up today
> and the aux. fan does run at very low speed in
> pos. 0-7, 8&9 is a little faster and 10 is full
> blast. It only works that way, however, if the
> motor is driving the flapper full open and the
> arm closes the microswitch which energizes the
> aux. fan relay (and also sends power in a
> parallel circuit to the fan resistor block to
> give you low and med.).

Interesting. So all this makes sense with my unit, except that in my servo, the fan microswitch turns power off when the servo is at max on. Very strange. The only thing that I can figure out is that it was somehow miswired at the factory. I don't think the servo has ever been hacked, as you can see the red ink dots are undisturbed (see photograph). Two possibilities: either they used the wrong micro switch, or the soldered the wires to the wrong terminals on the microswitch. Question, to which terminals on the microswitch are your fan power wires soldered. If you look at mine, they are soldered on the two outside terminals. [Note that I have added a jumper wire to short the two terminals together. This is my workaround to supply power to the fans constantly.] To get the behavior that you describe, I believe that the wires would need to be soldered to the two inside terminals. Can you check your unit, and let me know how it compares? Actually, if you could try again to post your pictures, it would be very helpful also.

> The bad news is that it acts the same as before.
> If I manually move the lever to the full open
> position, my aux fans run and the controller
> will leave the flapper at full open until I turn
> the rotary switch to off, then it closes the
> flapper. If I turn back to 10 on the rotary
> switch, the motor just sits there, it doesn't
> try to drive the flapper open. When the motor
> is closing the flapper I have voltage on the red
> wire going to the motor but when I go to 10, I
> don't have voltage on the black wire which tells
> me that either my one set of transistors is bad
> or the IC is bad. If I had voltage on the black
> wire then I would think the motor is bad but I
> don't.
>
> Looks like it's time to spend $180 to get a new
> board, unless I want to try 2 transistors and
> an IC. I don't have a heat gun here nor do I
> have a small soldering iron or experience
> working on printed circuit boards.

So I have one other suggestion. If you look at the upper left hand side of the circuit board, you will see two diodes. These diodes are wired back to back and connected in series with the motor. So normally, you would expect them to prevent current flowing through the motor. However, the microswitches on the servo are wired in parallel to the diodes, one across each diode. When the servo is not at either end of its throw, the microswitches are closed, therefore bypassing the diodes. When the servo reaches one end of its throw, one of the microswitches opens, and so one diode is in series with the motor, allowing the motor to go only in one direction. These diodes and microswitches serve to stop the servo motion in only one direction when it reaches either end of its throw. Trace out the circuit to see how it works.

So my suggestion is that perhaps one of these diodes is bad, or one of the microswitches is open. Check these components out.

One other thing. You keep saying that you manually move the lever on the servo. How do you do that? It is pretty tedious to rotate the motor by hand. I'm wondering if you somehow have disconnected the arm on the servo from its shaft, and are rotating the arm that way. If so, when you put it back together, you need to make sure that the arm fits back in the original position on the shaft. This is because there is a potentiometer -- that round thing near the microswitches -- that feeds back the position of the servo arm. If the relationship between the arm and the potentiometer is changed, then the circuitry will get confused. Actually, there is a possibility that the potentiometer could be bad. The resistance should be 1K and the tap should be somewhere in the middle. I think my tap was around 400 ohms when the servo was in the closed position, but I can't say for sure.

Let me know how it goes. Surely this debugging is not cost effective, but it's a matter of principle! I hate to be outsmarted by a lame piece of electronics...

-Juan

[Kurt peterson]

I threw the old micro switch away so I'm not sure how it was wired but I think it was to the two outside terminals. Since the two stacked together are the same switch and the bottom one is wired to the two inside terminals (normally closed) then if the the top one is wired to the two outside terminals it should be normally open which doesn't make sense in your case. Having said that, I looked at the diagram and assumed the switch should power the relay when it's closed which is how I wired the new switch. Again, my system never worked since I bought the car so I'm not sure how it's supposed to work. Makes sense since the fans turn on when the flapper is wide open. If you have your switch hot wired, then your fan is running at low speed even when your rotary switch is turned off - may be OK now but won't help when it's time for A/C.

I checked the micro switches and diodes and they are fine. If you pull back the black plastic cover around the gears, you can turn one of the larger gears with your finger to move the lever pretty quickly. I didn't remove the lever. My potentiometer has about the same readings so I assume it's OK.

I seem to be back at square one.