Cruise control repair and thoughts
#1
Rennlist Member
Thread Starter
Cruise control repair and thoughts
So my 86.5 928S has not had cruise control for awhile. Annoying, so I bought a cruise control brain on Ebay and popped it in. Worked for about 2 days (fairly well actually) then failed with the symptom that when you push accel on the stalk the CC speeds up the engine (throttle pot works) but it can't hold speed.
Bought a second CC brain that said "tested", put it in. It will set a speed, but if you set it at say 70mph it will set, allow the car to slow down about 3mph, then will shift the engine down to bring it back to 70. Then it coasts down, floors it, repeat. Works, but no real stability.
So a question: I have 3 CC brains with 3 levels of problems. Anyone have schematics or thoughts on what tends to blow up? I tried reflowing the solder on the plugs on the first one, no difference, and the solder joints look good under a 4x loupe.
If it's simply a specific drive transistor or one of the comparators I can swap that out without much difficulty. Or is it maybe a speed sensor on the engine/transaxle?
Didn't see many threads on CC repair so thought I would do my own.
Bought a second CC brain that said "tested", put it in. It will set a speed, but if you set it at say 70mph it will set, allow the car to slow down about 3mph, then will shift the engine down to bring it back to 70. Then it coasts down, floors it, repeat. Works, but no real stability.
So a question: I have 3 CC brains with 3 levels of problems. Anyone have schematics or thoughts on what tends to blow up? I tried reflowing the solder on the plugs on the first one, no difference, and the solder joints look good under a 4x loupe.
If it's simply a specific drive transistor or one of the comparators I can swap that out without much difficulty. Or is it maybe a speed sensor on the engine/transaxle?
Didn't see many threads on CC repair so thought I would do my own.
#2
Rennlist Member
I have had the same issues with my '86.5 cruise. Had it repaired but it worked for a while and looses speed about 3 mph. now. I need to pull the computer and send it back to the repair facility since it had a lifetime guarantee. It's definitely a brain box issue with capacitors being old. My later S4 cars don't have the problem. A different computer I believe.
#3
Rennlist Member
Thread Starter
Interesting. I see the guy on Ebay, and my guess is he has a pinout and a test jig set up to simulate the inputs and such. I might build one myself for the heck of it, but curious if anyone else has gotten anywhere with this.
C
C
#5
Nordschleife Master
Have you followed the testing protocol in the WSM?
There are a variety of places the cruise can have problems.
The brain box is by far the most common, but you can eliminate all the others, or find out that it is one of them, by simple testing.
There are a variety of places the cruise can have problems.
The brain box is by far the most common, but you can eliminate all the others, or find out that it is one of them, by simple testing.
#6
Rennlist Member
Thread Starter
Yes I have. It seems to test the stalk, the resistance on the valves to the dashpot, and the brake switch. All seem to work properly. The test for the sensor was more difficult, but it did read low/infinite which means it's a hall sender of some sort. What I haven't done is put a portable oscilliscope into the car to see if the waveform is good, but if it works on one controller it's probably good. Ergo problem in controller.
Yes Roger has them, but I'd like to try to figure this out on at least one core. I'll take the most dead one and look it over. In theory....
I should be able to put a 1khz square wave on the input signal line, +12 and ground on the battery connections, and the cruise should come to life.
If I short 9/12 briefly it should "set" the value in the computer. Then as long as the frequency doesn't change it should do nothing on 3 and 7.
The question is what does it do when the car slows down? I'm guessing it applies 12v between 3/7 to engage the servo to pull in the throttle till the frequency matches what is set. So if I go to 1.1khz then the controller should do nothing. If I go to 950 hz, then the controller should power 3/12 to engage the servo to bring up the motor speed till the frequency input matches 1khz. Likewise if I short the accell lines again the controller should close the servo line to once again speed up the car.
A question is does pulling back on it open the servo line to allow the car to "coast" then close the line again to allow the CC to re-establish the set speed?
So a CC that doesn't have a working memory could still work the accell line and move the servo, but would not have any set point. Likewise a CC that goes to full power may think the speed is set to infinite frequency which is why cars sometimes accelerate out of control when the CC is pushed to set/accell.
Hm.
Yes Roger has them, but I'd like to try to figure this out on at least one core. I'll take the most dead one and look it over. In theory....
I should be able to put a 1khz square wave on the input signal line, +12 and ground on the battery connections, and the cruise should come to life.
If I short 9/12 briefly it should "set" the value in the computer. Then as long as the frequency doesn't change it should do nothing on 3 and 7.
The question is what does it do when the car slows down? I'm guessing it applies 12v between 3/7 to engage the servo to pull in the throttle till the frequency matches what is set. So if I go to 1.1khz then the controller should do nothing. If I go to 950 hz, then the controller should power 3/12 to engage the servo to bring up the motor speed till the frequency input matches 1khz. Likewise if I short the accell lines again the controller should close the servo line to once again speed up the car.
A question is does pulling back on it open the servo line to allow the car to "coast" then close the line again to allow the CC to re-establish the set speed?
So a CC that doesn't have a working memory could still work the accell line and move the servo, but would not have any set point. Likewise a CC that goes to full power may think the speed is set to infinite frequency which is why cars sometimes accelerate out of control when the CC is pushed to set/accell.
Hm.
The following users liked this post:
TED WATSON (06-13-2023)
Trending Topics
#8
Rennlist Member
Recent experience showed me that having a significant vacuum leak in my HVAC system (recirc flap actuator bad) will cause the cruise control to misbehave. Cruise depends on vacuum to actuate, accelerating reduces vacuum available, cruise looses control of speed.
Check your vacuum system for leaks.
Check your vacuum system for leaks.
#9
Rennlist Member
Thread Starter
That is a really good point! I fixed the major leaks in my HVAC system by cleaning up the lines and plugging the three bad servos (footwell, comb, and defroster). It's possible I don't have the vacuum set up quite right to hold on hard accel, that would cause the symptoms.
Will hook up the Mity Vac this weekend and take it for a drive and watch.
Will hook up the Mity Vac this weekend and take it for a drive and watch.
#10
Rennlist Member
Join Date: Aug 2014
Location: Adirondack Mountains, New York
Posts: 2,398
Received 309 Likes
on
161 Posts
I read a very detailed description of how the brain functions, somewhere. Hairy analog stuff.
I used a guy who rebuilds your unit. He had trouble sourcing the driver transistor that had failed. More often it’s a capacitor.
Would have been faster and cheaper to use Roger’s.
I used a guy who rebuilds your unit. He had trouble sourcing the driver transistor that had failed. More often it’s a capacitor.
Would have been faster and cheaper to use Roger’s.
#11
Rennlist Member
Thread Starter
Well, from reviewing the schematics of the car it looks like the CC unit has only two inputs from the rest of car. They are:
Input from Speedometer: This is probably the pulses from the speedo sensor, so if the speedometer is working then the CC should be getting the right data
Input from the Lamp Check line. That's a bit more interesting, I have read anecdotes that CC doesn't work if the lamp check is on. So that might be something to keep in mind.
The servo likewise has a single plug input and a single vacuum line going to it. Also has what appears to be a vent as well. My *guess* is that the voltage to the servo can be reversed and it's a motor of sorts that can either set the diaphragm to vacuum (accel) or vent/atmosphere (retard). Or it's a single valve to add vacuum to the diaphragm to hold position against the normal throttle spring. Hm. Given that Porsche used a 3 wire servo motor for the idle air bypass, I can't see them switching to a polarity motor here, and I don't recall seeing an H bridge of FETs in the cruise control box, which would be needed to reverse voltage. Hm again.
In my case the oscillations could be caused by a leaking diaphragm in the AC unit itself, I'll have to figure that out. Or the computer, I have another one coming from the seller to cross-check: If I get the same problem I know it's the servo.
On the unit where nothing works, I might start by fuzzing out where the drive transistor/FET is for the servo and see if that is open/shorted.
Input from Speedometer: This is probably the pulses from the speedo sensor, so if the speedometer is working then the CC should be getting the right data
Input from the Lamp Check line. That's a bit more interesting, I have read anecdotes that CC doesn't work if the lamp check is on. So that might be something to keep in mind.
The servo likewise has a single plug input and a single vacuum line going to it. Also has what appears to be a vent as well. My *guess* is that the voltage to the servo can be reversed and it's a motor of sorts that can either set the diaphragm to vacuum (accel) or vent/atmosphere (retard). Or it's a single valve to add vacuum to the diaphragm to hold position against the normal throttle spring. Hm. Given that Porsche used a 3 wire servo motor for the idle air bypass, I can't see them switching to a polarity motor here, and I don't recall seeing an H bridge of FETs in the cruise control box, which would be needed to reverse voltage. Hm again.
In my case the oscillations could be caused by a leaking diaphragm in the AC unit itself, I'll have to figure that out. Or the computer, I have another one coming from the seller to cross-check: If I get the same problem I know it's the servo.
On the unit where nothing works, I might start by fuzzing out where the drive transistor/FET is for the servo and see if that is open/shorted.
#13
Rennlist Member
Thread Starter
In terms of the chips, this is a pretty much analog device: The two DIP chips in there are LM2901's, which are 4 channel voltage comparators. My guess is it operates as follows:
The pulses coming in generate a variable signal that can be rectified into a DC bias voltage. As long as the car goes a certain speed, you have a certain voltage. The first step is to set the voltage into a memory, and my guess is they have 2-4 of the comparators in a feedback loop that is set when you press the accell button. Once "set" that voltage stays on a bus.
Meantime the drive system checks the current voltage coming from the pulses coming in and compares it to the bus reference voltage. If the voltage is higher, the car is going faster and it opens the FET that powers the servo (letting vacuum out and slowing the throttle). If the voltage is lower it closes the FET which powers the servo, adding vacuum and pulling on the throttle cable. When in equilibrium, things stay the same.
With that kind of set-up you can "coast" by turing off the output to the throttle cable, and you can "accel" by turning on the outout to the throttle. Note that set and accel are the same thing: This opens the throttle while at the same time setting the reference voltage to the speed sensor voltage (which goes up). Turning off the CC by resetting it simply drops the feedback loop in the comparators, returning that voltage to zero.
Interesting.
The pulses coming in generate a variable signal that can be rectified into a DC bias voltage. As long as the car goes a certain speed, you have a certain voltage. The first step is to set the voltage into a memory, and my guess is they have 2-4 of the comparators in a feedback loop that is set when you press the accell button. Once "set" that voltage stays on a bus.
Meantime the drive system checks the current voltage coming from the pulses coming in and compares it to the bus reference voltage. If the voltage is higher, the car is going faster and it opens the FET that powers the servo (letting vacuum out and slowing the throttle). If the voltage is lower it closes the FET which powers the servo, adding vacuum and pulling on the throttle cable. When in equilibrium, things stay the same.
With that kind of set-up you can "coast" by turing off the output to the throttle cable, and you can "accel" by turning on the outout to the throttle. Note that set and accel are the same thing: This opens the throttle while at the same time setting the reference voltage to the speed sensor voltage (which goes up). Turning off the CC by resetting it simply drops the feedback loop in the comparators, returning that voltage to zero.
Interesting.
#14
Rennlist Member
Thread Starter
My review here is
#15
Rennlist Member
Just got a cruise unit rebuild from Roger...slipped it in and works absolutely perfectly...set/accel, reset, cancel...beautiful! Car hasn't had cruise since I bought it...so nice I won't have sore ankle on my next car trip!