| Author |
Message |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Tue May 27, 2008 5:08 pm Post subject: Rough, rich idle revisited. |
|
|
It's been a few years now and I still can't get rid of this stinky, fuel rich, loping idle on my 1985 Digijet Vanagon. I've been here many times dicussing it, things have gotten better but the problem is still there. Everything electrical in the FI system tests fine or has been replaced. The whole engine was replaced about a year ago. She runs extremely well with the new 2.2L. Great gas miledge (22mpg) and nice power.
The TPS (single switch design) tests fine at the connector, and at the ECU terminal. Adjustment is right on. I noticed though, that the idle doesn't change if I disconnect it while running. Does this indicate there could still be a TPS problem even though it tests fine according to Terry Kay's book and Bentley's? If the signal is getting to the ECU, does that tell me that the ECU is the problem? Funny thing is, I have 2 used ECU's and AFM's and the problem exists on all of them. Fuel pressure is perfect, the injectors are new. This idle problem is my "white whale", can anyone help me? _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
keithwwalker Samba Member

Joined: May 30, 2005 Posts: 886
|
Posted: Tue May 27, 2008 8:35 pm Post subject: |
|
|
Here's my idea, unpop that seal on the CO adjustment screw near the AFM. Count the turns all the way in. Now go back where you started and turn it out one turn. If still rich, go out more in 1/4 turn increments.
Don't go beyond 13 turns, or the screw will disengage from the thread.
My 1.9 failed the idle only emissions test until I turned the screw out. Now it runs fine at idle.
IMO, this is the FIRST adjustment you should do for a bad idle, AFTER you are sure that there are no vacuum leaks in the distributor lines or the intake boots and runners.
 |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Tue May 27, 2008 9:48 pm Post subject: |
|
|
I've tried tweeking the AFM. I can screw it all the way out without changing the idle much. If I screw it in past a certain point the idle slows and gets even richer. I keep a spare with the stock adjustment. No difference if I switch AFM's. Both test perfectly. I had a Harley guy stop me at a gas station the other day. He complained about the rich smell of my exhaust. Works great at all speeds but idle. Plugs are nice, clean, and tan. _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Wed May 28, 2008 11:41 am Post subject: |
|
|
Tell the Harley guy I hate his stupid pipes, and if he doesn't like the smell he should just come around. So there.<insert sarcastic bastard emoticon here>
Matt, if your o2 sensor is working right, the mixture should be lambda at idle once warmed up. The way the ECU knows to get out of warmup mode and into closed-loop is by the sgnal from the Temp2. So, assuming that your o2 is working (test it by seeing how it responds when you disconnect it, or by probing the voltage at the sensor connector), it may be that the Temp2 isn't lowering resistance enough to boot the ECU out of warmup. But I'm assuming you put in a new T2 in all your travails. Nonetheless, see if you can check its resistance when the engine is warm, to be sure it's not out-of-scale according to the resistance/temp table in Bentley. You're not using an 80C tstat, are you?
Also, rather than go on your own impression, or worse, those of some troglodytic thumper-jockey, have you actually quantified that it is running rich at idle, by attachment to a CO meter or a WBO2 controller? I've learned over the years not to put too much faith in my sensory impressions, except as a prod to gather some actual data.
I can disconnect my TB switch circuit at idle and normally I see no change in idle quality. If the system is set up to idle, it will do so just fine without needing the ECU to go into idle mode. That's actually a sign that the tune isn't too far off.
Lastly, have you used a timing light to verify that the ignition advance, and especially retard, work as specified in the book? Too much retard at idle could cause the sysmptoms you describe, even though it runs fine off-idle. _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Wed May 28, 2008 2:17 pm Post subject: |
|
|
| tencentlife wrote: |
Tell the Harley guy I hate his stupid pipes, and if he doesn't like the smell he should just come around. So there.<insert sarcastic bastard emoticon here>
Matt, if your o2 sensor is working right, the mixture should be lambda at idle once warmed up. The way the ECU knows to get out of warmup mode and into closed-loop is by the sgnal from the Temp2. So, assuming that your o2 is working (test it by seeing how it responds when you disconnect it, or by probing the voltage at the sensor connector), it may be that the Temp2 isn't lowering resistance enough to boot the ECU out of warmup. But I'm assuming you put in a new T2 in all your travails. Nonetheless, see if you can check its resistance when the engine is warm, to be sure it's not out-of-scale according to the resistance/temp table in Bentley. You're not using an 80C tstat, are you?
Also, rather than go on your own impression, or worse, those of some troglodytic thumper-jockey, have you actually quantified that it is running rich at idle, by attachment to a CO meter or a WBO2 controller? I've learned over the years not to put too much faith in my sensory impressions, except as a prod to gather some actual data.
I can disconnect my TB switch circuit at idle and normally I see no change in idle quality. If the system is set up to idle, it will do so just fine without needing the ECU to go into idle mode. That's actually a sign that the tune isn't too far off.
Lastly, have you used a timing light to verify that the ignition advance, and especially retard, work as specified in the book? Too much retard at idle could cause the sysmptoms you describe, even though it runs fine off-idle. |
_________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Wed May 28, 2008 3:00 pm Post subject: |
|
|
Oops, just lost my big long winded reply.
I'll run another check on the Temp 2. I think I may have assumed that it was good because it's new. Can't remember if I've checked the wiring on it. O2 sensor, seems to have no effect when disconnected. Hmmm... it used to run better when disconnected. This is the second one I've bought in a year. Do they burn out when the engine runs rich?
I already swapped the dizzy and the vac advance.
Thanks _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Wed May 28, 2008 3:25 pm Post subject: |
|
|
Yes, I think rich running can really shorten the life of o2 sensors. But it's easy to check if it's working. Disconnected or connected, you should read between 0-1 V at the sensor signal connector, with your voltmeter's other lead grounded. Connected, it will range up and down, and won't settle at any point; disconnected, it will still vary, but not in such a wide range, and may settle around a certain point, usually near one end of that range or the other. If it's putting out voltage, then it's functioning. As they age, what mainly deteriorates is the response time. Dead ones produce no voltage. They can be killed by chronic overly rich running, and especially by byproducts of silicone sealers, so avoid using silicone in the air tract (intake and exhaust) unless it's "sensor safe". And, like every electrical part, there are a lot of bum ones on the shelf, so even though you buy a new one, you should verify its function. _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Wed May 28, 2008 7:44 pm Post subject: |
|
|
I ran some tests and this is what I found.
The O2 sensor seems to be functioning. 0 volts when cold, it steadily works it's way up as the engine heats up. It levels out at around .75 volts.
Temp 2 reads 2376 ohms at 68F and 230 ohms at operating temp. All connections check out at the ECU. I ran all the ohm tests.
AFM and temp 1 are functioning within limits throughout the temp range.
TPS is perfect.
The engine does run differently when the O2 is disconnected, not necessarily worse, but different.
Here's the interesting thing. When I disconnect the Temp 2 at operating temperature while the engine is running, the engine instantly achieves a much smoother idle. This seems strange since doing this fools the ECU into thinking the engine is cold...right? That would mean that the engine wants an even richer mixture...? The strong smell of the exhaust always made me feel it was too rich. Either I need to rich out the AFM (I think I already tried that) or I must have a big vacuum leak thats leaning things out too much, or.......? _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Wed May 28, 2008 10:15 pm Post subject: |
|
|
I seem to remember saying something like this:
| Quote: |
| Also, rather than go on your own impression, or worse, those of some troglodytic thumper-jockey, have you actually quantified that it is running rich at idle, by attachment to a CO meter or a WBO2 controller? I've learned over the years not to put too much faith in my sensory impressions, except as a prod to gather some actual data. |
_________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Thu May 29, 2008 9:52 am Post subject: |
|
|
Yes, I know, assumption is the mechanic's worst mistake. Once again your words of wisdom ring true.
I'm pretty sure enriching the setting on the AFM isn't going to do much. I've been down that road. I can find no vacuum leaks, all fittings and hoses have been replace.
Does the Temp 2 signal tell the ECU to do anything else other than to increase the pulse rate at the injectors? Is timing linked to temp 2 in a direct way? I feel like I'm getting close. _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Thu May 29, 2008 10:13 am Post subject: |
|
|
It is on Digifant, but not on Digijet. Digijet's ignition system is completely independent of the ECU. They both read the Hall's rpm pulse, but share nothing else except power supply. _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Thu May 29, 2008 10:39 am Post subject: |
|
|
Thanks. That simplifies it. If the timing is working properly (better check that again) then it's just running too lean. Only problem is, I don't know why yet. I think there's more to it than the AFM. When I put the AFM with the stock adjustment and cap in place I still have the same idle issue.
By the way, I like your engine building add. If you had been building them a few years ago I definately would have bought one. _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Thu May 29, 2008 6:31 pm Post subject: |
|
|
Thanks. The things are keeping me very busy these days.
It's funny: the AFM and ECU are the first things blamed a lot of the time, but they're almost always the last things that are actually at fault. I'll bet 95% of the replacements of those sold over the years have been unnecessary. And most of the mechanics who sold them found that out pretty quick, because they didn't fix the problem. Both units are tough as all get-out. And, the AFM signal can be all over the place, I mean out to friggin lunch, and the things will run really good so long as the lambda system is working. That's why you'll almost never read me recommending someone replace either, and if the ECU is suspect, I try to make sure they're really really sure before they go buying a new one, because most of the time they're throwing their money away.
But, the system is very sensitive to those temp signals. The rpm signal has to be perfect. It hates bad signal from the TB switches, which makes sense because that throws the ECU into a completely different program. And the actual down-the-road mixture is only as good as the o2 sensor. _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Fri May 30, 2008 9:19 am Post subject: |
|
|
| tencentlife wrote: |
It's funny: the AFM and ECU are the first things blamed a lot of the time, but they're almost always the last things that are actually at fault.
But, the system is very sensitive to those temp signals. The rpm signal has to be perfect. It hates bad signal from the TB switches, which makes sense because that throws the ECU into a completely different program. And the actual down-the-road mixture is only as good as the o2 sensor. |
I totally agree. It's like saying "if the car won't start it's the starter", "the battery is dead, must need a new battery."
In my case, I've tested all the components several times. I've learned a lot along the way. Many things have improved. The van now runs as well as it can with the wrong idle mixture. I'm still looking for the silver bullet. Getting close. _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Fri May 30, 2008 11:08 am Post subject: |
|
|
Great, keep at it. This is how you become "An Expert": by banging your head against the wall long enough that everyone can see the bumps! _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Sat May 31, 2008 6:55 pm Post subject: |
|
|
I bipassed the idle stabilizer and attempted to richen the idle with the AFM. I ended up screwing the adjuster all the way in. I backed it out about a half turn. That's the best I can do. It runs a little bit smoother. The idle surges so much with the IS bipassed that it's hard to tell when you've got it dialed in.
Again, if I disconnect temp II while it's running at operating temp the idle smooths out nicely. Can't get rich enough. Any comments? Can't find a vacuum leak. Timing issue..? Hall sensor...? _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
Pascal Samba Member

Joined: January 13, 2006 Posts: 836 Location: Fredericton, NB, Canada
|
Posted: Sun Jun 01, 2008 4:38 am Post subject: |
|
|
I have the exact same thing with my 1.9, and have tried pretty much exactly what you have. A few weeks ago I started swaping parts with my spares and when I tried the idle stabilizer, that smoothed the iddle out. I have it at 1000rpm now as that is the best I can do. I still think something is not quite right but the iddle is now solid.
I was thinking of trying a new breather tower after reading some of the posts on those and how they deteriorate over time. _________________ '84 Westy - 2.2wbx, GT Torque-Biasing Differential
'91 EA Cabby |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Sun Jun 01, 2008 12:25 pm Post subject: |
|
|
I've been doing a lot of spare parts swapping. The main thing I've confirmed is there was usually nothing wrong with the original parts.
| Pascal wrote: |
I was thinking of trying a new breather tower after reading some of the posts on those and how they deteriorate over time. |
That's one I haven't heard. Is there a valve in that thing? If that is the case a stuck valve could cause vacuum leak symptoms when crank case pressure is low and manifold vacuum is high.
I just checked out this post.
<www.thesamba.com/vw/forum/viewtopic.php?t=295812&highlight=breather+tower>
That fits my van to a tee. The idle smooths out when I disconnect and cork the breather tube inlet. Why didn't anybody mention this in the 100 some posts I've made on this subject. Thanks Pascal!
Tencent, did you ever end up using an American PCV valve? If so, how did you configure things? Which one did you use? _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
tencentlife Samba Member
Joined: May 02, 2006 Posts: 10156 Location: Abiquiu, NM, USA
|
Posted: Sun Jun 01, 2008 2:48 pm Post subject: |
|
|
That was where I got stalled. Which one? I was staring at the 80-some-odd different PCV valves at Autozone and didn't now how to differentiate apart from hose sizing and such, but it wasn't a high priority that day, so nothing so far. I do want to try a thru-breathing setup. Just too much else that needs doing first. _________________ Shop for unique and useful Vanagon accessories at the Vanistan shop:
https://intrepidoverland.com/vanistan/
also available at VanCafe.com!
Please don't PM here, I will not reply.
Experience is death to doctrine. |
|
| Back to top |
|
 |
mattcfish Samba Member
Joined: October 24, 2006 Posts: 565 Location: Bellingham, WA
|
Posted: Sun Jun 01, 2008 8:03 pm Post subject: |
|
|
$88.00 is a lot for the plastic VW part. PCV valves might work but I'm worried about proper breathing, PCV's are restrictive by nature. _________________ 71 Bay Window
75 BMW 2002
76 BMW R90S
85 Vanagon with 2.2L upgrade |
|
| Back to top |
|
 |
|