I am the new owner of a 1985 Cabriolet with 46,000 miles. Although the car has low miles it has it's share of troubles that I am working through slowly. We are in the Phoenix Az area and it is July so the temps dont get much below 90 F now in the mornings. I am having trouble getting the car started in the morning. It runs very rough for the first minute or so often requiring multiple starts until it is warm enough to rev. Once warm the car runs great and will start again well while warm. Once it cools down completely we go through the same rough start problems.
I have investigated the Cold start valve (CSV), Thermo Switch (TS) and Hot Start Pulse relay (HSPR) and have found the following. 1) The cold start valve is not operating with the electrical pulses being sent through the circuit from the HSPR. 1) Wires from the CSV to the TS are good, (I checked them with my ohm meter) 2) I receive about 12 volt pulses to the Temp Switch from the HSPR while turning the key and operating the starter. The Cold Start Valve only recieves 5 volt pulses from the HSPR while turning the key and operating the starter. I have tested the CSV and it operates great with a 12volt signal and I have tested it using voltages down to 7.5 volts using a small battery and it still opens OK. I have started the car easily in the morning using a 12 volt hot wire from the battery to the CSV giving it a couple of touches while cranking cold. The car started and ran well. So I think I have narrowed the start problem down to low voltage (5-6 volts) being sent from the HSPR to the CSV which is inadequate to activate the valve. I have ordered a new HSPR and hope that this will do the trick to increase the voltage and make the CSV open and give me that little boost of fuel needed when the engine is cold (90 F).
I was hoping that someone could tell me if they have had similar problems with low voltage to the CSV and if so am I on the right track replacing the Hot Start Pulse Relay?? Should the Cold Start Valve be working on 5 volts?? if so I guess I should pull that next.
Any help at all with this starting problem would be much appreciated.
Last edited by jfknudson; 07-27-2010 at 04:17 PM. Reason: clarified text
I'm in Phoenix too!
Anyway, it sounds like you might be on the right track. However, the fact that it starts warm/hot without issue indicates that the relay is actually still good.
When the engine coolant temp cools off to around 30°C or below, the thermo-time switch should be closed, completing the ground path, and telling the CSV to operate between 3-8 seconds at start-up. The switch opens at around 35°C (95°F), interrupting the ground path, preventing the CSV from operating and, instead, diverting power to the CSV from the HSPR in pulses.
Another possibility: the auxiliary air regulator is not opening. However, since jumping the CSV when cold gets the car to start, I'd test the thermo-time switch:
1) Engine coolant temp should be below 30°C (86°).
2) Disconnect coil wire from distributor cap and connect it to ground.
3) Disconnect CSV harness connector and connect a test light across the terminals.
4) Have helper actuate the starter. The test light should light up, then go out. The amount of time the light remains lit depends on the coolant temperature. At 68°F, the switch is closed (test light lit) for up to 4 seconds. At 86°F, the switch is closed for up to 2 seconds.
Thanks for your response on this. Since I posted the thread I have replaced a faulty Auxilary Air Regulator and looked into the Air Flow Sensing Plate on the Fuel Distributor. When I cracked that thing open I found an old air cleaner that needed to be replaced and the air sensor plate seemed to set too low in the bowl. My Haynes manual says that the plate should ride up about 1.9mm (for 1985 model) up in the bowl to allow air to pass. I figured out how to adjust the spring clip (PAIN IN THE BUTT) and after a few tries I got it to where it gaps between 1 - 2 mm. This is really tough to do and it took quite a while. I set it once with a wide gap and the car started beuatifully but on the road test it would cough and sputter when I tried to accelerate. I came home and lessened the gap to close to 1 mm and tried again. This adjustment caused the car to accelerate well again and the next morning when I tried the cold start it fixed that too.
Now only one problem left. When cold starting (Ambiant Phoenix morning temp about 90 F) The car fires beautifully idles at about 900rpm briefly the the idle drops to less than 200 rpm where it almost dies than immediatly picks up again to about 900 - 1000 rpm where it idles perfectly. It needs than to warmup for about 1 minute before you can rev the engine without it sputtering. After the one minute warmup it runs great. Now is there some kind of vacuum advance or cold idle compensator thing to look for now or is this as good as it gets.
I shouldnt complain too much, after adjusting the Air Flow Sensing plate the car runs much better on cold startup than it did before but if there is something else to check please let me know
Last edited by jfknudson; 08-10-2010 at 09:07 AM. Reason: misspells
Idle setting is out of adjustment
Lamba system not operating properly (system tests are at the above link)
Auxiliary air valve not opening
Leaking cold-start valve
"Cold" control pressure is out of tolerance
Injectors aren't spraying evenly
A "hunting" idle usually indicates a leak in the vacuum/air intake system. Since yours "hunts" only until it's warmed up, I'd test the Lambda system and check/adjust the air-fuel mixture and idle.
The info you linked me to talks about a thermoswitch, is this different than the thermo time switch that operates the Cold start injector or is it one in the same. Either way which one should I be running in our Az climate, the 77+°F, 104+°F, or 177+°F, That looks like this setting could be pretty important. What do you recommend?
Thermo-time switch = cold-start valve (on top of coolant hose flange)
Thermo switch = oxygen sensor (on underside of coolant hose flange)
Thermo switch = cooling fan (lower left corner of radiator)
The thermo-time switch comes in only one temperature: 35°C.
The thermo switch for the oxygen sensor system comes in two different temperature ratings: 25°C and 45°C. The standard switch is 25°C; it's what it is in my '86 and is the factory-original switch.
The thermo switch for the cooling fan comes in various temperature ratings. The standard is 95-84°C. If you're wanting to install a lower-temp thermo switch for the fan (making it turn on sooner), you must install a matching lower temp thermostat. Unless the car is having cooling issues, it's wise to just leave everything as-is.
After reading the idle setting piece on the Cab website I have one last question. (well maybe more) When starting my beautiful 1985 Cab at 95 degrees F (morning cold start) It sounds like from the article that it should be in "Open Loop" phase. Would changing the Oxy Thermo Switch to the 104 degree F version extend the cold running enrichment phase of the start sequence longer for warm climates, and would extending this cold enrichment phase help the car rev better without coughing in the first few minutes of operation? (Long question but I used only one question mark) I will be buying a dwell meter to check all the other things as well but the car runs perfectly now except for this annoying cold cough. By the way, The auxilary air regulator was replaced and the cold start valve works perfectly without leaks when it receives more than 7 volts and as you suggested I will be checking the Lambda, vacuum leak and air fuel mixture that you suggested earlier.
Your thoughts on this would be appreciated
Last edited by jfknudson; 08-10-2010 at 09:16 AM. Reason: add detail
Yes, the car should be in open loop upon start-up. If the car is sitting in a 98°F garage (like mine), the coolant temperature is also 98°F upon "cold" start. That means the thermo switch should be open and the system is just waiting for the oxygen sensor to heat up and kick in its signal to the ECU. The cold-start enrichment phase is skipped altogether because the system's components have deemed it unnecessary.
The switch currently installed in your car is perfectly fine temperature-wise and is actually better for warm/hot climates. The higher temp version is more for cold climates. The 45°C OXS thermo switch will extend the time the system runs in cold-enrichment mode, which means the car will be running slightly rich for a longer period of time. In other words, here in Phoenix the car might still be running in cold-enrichment mode even after the OXS has kicked in, which is not a good thing. In warm/hot climates, the system reverts to closed loop quicker, as you've experienced, because the coolant temperature (and OXS) gets a head start on heating up to its "operating" temp.
The issue your car is having is that it has an irregular idle during open loop, when there should be no issue at all given that the ECU should be forcing a 50% air-fuel mixture. Therefore, since the AAR and CSV have already been replaced and are confirmed as functioning properly and the only issue is during warm-up, something else could be amiss somewhere in the system.
Therefore, do what you're planning to do: test the OXS system (you may find the OXS and/or other component have gone bad), check the air-fuel mixture/idle and adjust if need be and make sure there are no vacuum leaks (CIS hates vacuum leaks!). I'll be air-fuel checking myself this week because my '86 gets to go in for an emissions test, hoorah!
Now back to the Cold start problem after last weeks Instrument cluster failure diversion.
I am preparing do the CIS Air/Fuel & Idle Adjustments as described on you website. but I do not have a dwell meter and have been unable to find one to borrow one so I thought I would ask. I do have a very expensive Fieldpiece Multimeter (the type used by HVAC repair techs) This meter is able to read frequency. So I thought that I would hook it up to the dwell port on the car and see what kind of a reading I get. Well when starting the car cold and idleing at 90 F The meter moved betwenn .0069 and .0070 and was very stable even while the rpm was cycling a few 100 rpm's. After idleing for a few minutes the cars idle rpm stabilized but the readings from the meter still were reading the .0070 as before. The specs for the meter are below
FrequencyRanges:4k, 40k, 400k, 4M, 40MHz
Accuracy:±(0.1% rdg + 3 dgts)
Sensitivity:10Hz ~ 4MHz: > 1Vrms, 4MHz ~
40MHz: > 2Vrms, <5Vrms
Minimum pulse width:> 25ns
Duty cycle limits:> 30% and < 70%
Overload protection: 500VDC or AC rms
After all that the question is... will this meter do me any good and exactly what am I reading here?? Is this .0070 = to 70 degrees that you would see on a Dwell meter?? If not, is there a conversion I should be using?
Thanks for any info you can give
Well, what you're actually measuring is not dwell, but the duty cycle of the frequency valve, which is read as a percentage; however, a dwell meter can be used and it will simply display the reading in degrees. It would appear your meter would be appropriate, but I cannot confirm that; it seems your meter displays the reading in Hz rather than percent.
I'm not sure what range the car's frequency valve operates in (all I know is it's resistance is 2-3 ohms), but I would equate your meter's reading as 70%. However, you need verify that the meter is reading correctly (the reading not fluctuating is not a good thing*): Actuate the full-throttle switch; the meter should read 65%. If it reads 35%, reverse the test leads at the test port and actuate the full-throttle switch again; it should now read 65%. If actuating the full-throttle switch shows no change, there is a problem with either the meter** or the test ports.
*The duty cycle, under normal conditions, usually fluctuates between 5 and 8 percent.
**The meter itself is probably fine; it simply isn't reading the car's duty cycle as a regular duty cycle/dwell meter would.
Jettincab! Thanks for all your help on this.
I purchased a nice little dwell meter and was able to get some good readings from the test plug. At ambient temp about 88 F it takes a couple of times to start the car. It idles up and down to where it dies. after a couple of starts you put a little gas to it on the down swing and you are able to stablize the idle as it warms up. after about a minute of idle pushing the gas to keep it alive you can let it idle by itself. It continues to pulse up and down. during this pulsing the dwelll readings had a large flucuation from around 38 degrees to the high 60's. the idle stabilizes and the dwell range reduces as the car warms. I then took car out for a run to get it up to operating temp. once the car is warm it runs great. Upon return I plugged it in and measured the dwell it was ranging from 54 degrees to 64 degrees in warmed up condition with the fan turning on occasionally. From these readings it sounds like the car is running lean (> 45 degrees or 50%) I also flipped the full throttle switch and got a constant 59 degrees with no variation on this reading. The full throttle reading looks dead on (58.5 +/- 2) degreeas. Do you have any thoughts on this idle behavior?? I am preparing to adjust the air fuel mix and have attempted to pull the tamper plug as you describe on your site. I tried using the punch method from the inside and got it to move a little but my punch tool was not very good. Its a bitch to get around the airflow sensor arm to get a good angle for the punch. Now I have drilled in and installed a 1.5" sheet metal screw from the top but am unabe to pull hard enough to get it out. Do you have any suggestions or any other tricks to get this *&%##?< thing out??
Again any ideas you may have would be appreciated.
Last edited by jfknudson; 08-29-2010 at 10:37 PM. Reason: added info
Sorry, didn't see you had replied until now!
If you haven't gotten the plug out by now and the screw method is just not working, it's time to go to method #2: Take the air-flow/distributor assembly off, flip it over and punch the plug out from the bottom.
The car is indeed running on the lean side when warmed up. So, you'll want to richen the mixture just a bit, to where your meter is reading 45° with a swing in either direction of 7°.
The full-throttle switch is clearly working properly. The issue is that warm-up period. Have you checked for vacuum leaks and tested the oxygen sensor system (as outlined on the CIS page)? If not, you'll want to do that too.
I did get the plug pulled and thouroghly tested and tuned the Air/fuel/Lambda system. It all is working great. 45.6 degrees solid cold, 45 +/-7 normal operating temp and 58.6 full throttle swithch. The dwell meter showed that I was running rich with # less than 45 degrees so I turned the adjustment screw counterclockwise to lean it out raising the # to 40 - 50 swing or about 45 +/- 5. I also checked the Oxygen sensor by the Haynes method and both your method #1 and #2 and it was all good. Auxilary Air regulator as well as the Hot start pulse relay are new. I also tested the resistance reading on the Frequency Valve as Hynes suggests and got a passing 2.7 ohms. The Idle air stabilizer was also checked and operational by the hose pinch Haynes method.
Now for what is still suspect. I checked the Warm up regulator as Haynes suggests and got a good voltage of 13v to the regulator but the resistance measurement was low at only 12.6 ohms. The manual says it should read between 16-22 ohms.
Today when I ran the car this AM it started poorly still requiring a few starts before it would idle. It idled at about 500 rpm and sputtered for about 1 minute then I was able to put a bit of throttle to it to get the idle up and rev'd it a few times. Then it seem to warm up and runs great. I parked it at work where it sat all day (105 degrees F in Chandler today) I came out a 4:00 and started the car. I did not touch the gas at all. The car started then idled down to 300 - 500 rpm and just barely ran. but it ran steady. After a minute or two I was then able to give it some throttle without it bogging down and was able to get it rev'd up. Then it runs great. On hot starts it cranks but wont fire until you give it alot of throttle, then it catches, then after just a few seconds it again runs great.
What do you think about that Warm Up Regulator resistnace at 12.6 ohms?? Should I replace it, they are tough to find and expensive. I plan to hunting again for vaccum leaks with my carb cleaner and I thought that I would pull the injectors to check for any leakers. Is there anything else that you can think of that would be helpful?? I am really running out of ideas, beer, and things to tinker with. These startup problems are a pain. I am beginning to think this is how they all must run and just live with it
Thanks for any help you can give.
Oh, excellent!! Definitely check for and fix any vacuum leaks and pull the injectors to check their spray/clean them/replace O-rings, if need be. CIS hates vacuum leaks, even small ones.
And, I think you've now discovered the root of the problem: The warm-up regulator (WUR) is also known as the control pressure regulator (CPR) and, actually, that's the more proper term for it because it doesn't operate just during warm-up. Two more possible issues are: 1) control pressure is out of tolerance; 2) faulty injector(s). That resistance value of 12.6 appears to be the tell-tale sign: According to the Bentley Manual, when measuring the resistance across the regulator's terminals (harness unplugged), "If the resistance is not 20 to 26 ohms, the control pressure regulator is faulty and should be replaced." You might want to double-check the resistance, just to be sure.
A quick search revealed this place (probably the cheapest around that might actually have it in stock):
http://www.evwparts.com/Merchant2/me...ry_Code=A1Fuel ($145 +$100 core charge)
And, no, they all most certainly do not run like yours and they aren't supposed to. They all should fire right up and idle smooth. You can live with how it currently is... until your patience runs thin.
Last edited by jettncab; 09-03-2010 at 05:53 AM.
Still working to stabilize the startup idle. I did a very thourough check for vacuum leaks with the Carb cleaner and found none, but I did find something new. With the engine idleing low I checked the incoming voltage on the idle air stabilizer plug. I was reading about 12volts as I adjusted the Idle up the voltage dropped off, so it seems like the solenoid is getting power at the correct time. I removed both Idle stabilizers (regular and AC) to clean and degunk. After cleaning the solenoids, I attempted to blow air through both valves and both were closed. I put 12 volts across the Idle air stabilizer and got no activation I then checked resistance across both solenoids and got 25 ohm reading across the AC booster and 0 ohms across the Idle air stabilizer. I think It dont work!!! If this is the case it does make sense that the car fires up then idles down and stalls. The Idle stabilzer is not there to catch it. This happens routinely four times on cold start until it warms up a bit then it runs at a low idle increasing slowly until the engine warms completely. When you attmpt to push the throttle while in this cold condition it will sputter (lean??) until it warms up a bit then it runs great. Does this make sense to you guys or is it still something else??
Thanks again for any advice you can give.
Control pressure has a lot to do with idle and general running condition. As said earlier, if the control pressure regulator (warm-up regulator) is reading below 20 ohms when checking resistance, it is faulty.
The idle speed boost valves (proper name for them), are controlled by the Idle Stabilizer Control Unit on the relay panel (position 18; verify with part #811905343). This relay turns the black valve on/off according to an engine speed signal from the coil. Below 750 rpm, the relay sends a voltage signal which opens the valve (there should be an audible click). When engine speed exceeds 1050 rpm, the relay cuts the voltage signal to the valve and it closes. The second boost valve kicks in whenever the A/C compressor engages.
Check the boost valve as follows (engine must be running): If the valve is receiving a voltage signal below 750 rpm and not opening, the valve is faulty. If the valve is not receiving voltage below 750 rpm, check the signals to and from the relay: With the ignition on, check for battery voltage at the socket for relay terminal 15. With the engine running, test for voltage between the socket for relay terminal 1 and ground. Also check the continuity of the wiring between the panel and the valve's harness connector. If all of that passes inspection, the relay is most likely faulty.
It's uncommon for a boost valve to fail, but it can happen.
I am looking at replacing the control pressure regulator (warm-up regulator) soon.
I did check voltage and the idle speed boost valve was receiving 12V when the idle speed was under 750 rpm when I increased the rpm to above 1000 rpm the voltage went to 0 volts. So I think all of the electronics are good. The solenoid itself did not open when 12 volts was applied and It appears to be shorted out when I checked It's resistance. This I think tells me the solenoid valve is bad. I will go out and check it all again however. I looked around online for the idle valve set (Idle boost and AC combo) and have not found any for sale yet. any ideas where to look for that item??
A new one runs around $100. I've not come across one on any "aftermarket" part store web sites, so you may have to go through a local VW dealer or online at 1stvwparts.com .
The cheaper route is to get the boost valve used from a local VW junkyard (Affordable German and German Auto Salvage).
For ease of getting/looking for them, here are the part numbers:
026906283C -- white, A/C
026906283B -- black, regular
HELLO, This thread has been quiet for a long time. I now have my 1985 Cab running perfectly hot or cold. It took a lot of work and investigation but it is worth it. How have you tested the cold start valve and the thermo switch?? I used the proceedures as written in the Haynes Manual and I was able to follow most of it and I am not a mechanic. Let me know in detail what you have done and I may be able to help.