| Author |
Message |
nelinor
Samba Member
Joined: March 21, 2008
Posts: 98
Location: Ventura County / SoCal
|
 Posted: Fri Nov 21, 2008 1:14 am Post subject: So close . . . (idle issues) |
 |
|
1983.5 Westy
A few weeks ago, my engine suddenly stopped running. I was driving at about 40mph when the engine stopped. I pulled over and the van wouldn't even start. I had it towed and here's what I've done so far:
DAY 1
1) changed the battery
2) replace air filter
3) replace fuel filter
4) changed the motor oil
5) changed the spark plugs and ignition wires
6) replaced the coil and the wire from the coil to the distributor.
7) replaced the cap and rotor
 checked the timing with the gun
I put the key in the ignition and the engine started up. However, the idle was extremely rough. The motor would surge and nearly cut out at idle. It would die as soon as I opened the throttle. Did more research here on the samba.
DAY 2
1) Replaced the oxygen sensor
2) Replaced the coolant temp II sensor
3) Replaced the idle stabilizer control box (the old one had stuff rattling around on the inside)
4) Checked the rubber boot that connects the throttle body to the air intake box.
I put the key in the ignition and the engine starts right up. Idle sounds perfect. No hestitation or surging.
Here's the problem that I can't figure out. After the motor warms-up (about 3-5 mins.) the engine starts surging and coming back down, surges and comes down. The idle is not stable.
The same thing happens if I open the throttle and then close it. The engine surges and comes back down again. Over and over and over.
I think that I'm close to getting my westy back on the road. But I'm confused right now. Where did I mess up? Is this a timing issue or vacuum issue?
Oh, sages of the samba, please help. You are my only hope!
CLIFFNOTES: motor stops, change out parts, idle is fine until the motor warms up or until the throttle is opened and then closed. After the motor warms up or the throttle is manipulated, the idle starts surging.
_________________
1983.5 VW Vanagon Westfalia - manual - EJ22
2002 Subaru Impreza Wrx
1990 Nissan 300zx tt |
|
| Back to top |
|
 |
hansh
Samba Member
Joined: January 13, 2004
Posts: 615
Location: Bend, OR
|
| Posted: Fri Nov 21, 2008 8:27 am Post subject: |
|
|
I know this answer may not be what you want but, I don't think this is anything other than a vacuum or timing problem. When the idle is too high, the computer compensates to bring it down, then the idle raises, ECU brings it down again, over and over and over.
Carefully double check your timing procedure as that is the most likely suspect...that the timing is too advanced or the timing is too retarded and the idle has been set to compensate.
If you have a dual advance distributor, make sure you haven't switched the hoses.
Another possibility I see from what you've done so far is that the mixture could be too rich causing the engine to love what's going on while it warms up but then disliking the extra fuel once the extra air allowed for warm-up is cut off. A cause for this would be someone having set the mixture to compensate for an air leak problem. When problem is fixed, the idle is still set too rich.
_________________
‘85 ej22
'84 Westfalia |
|
| Back to top |
|
|
tencentlife
Samba Member
Joined: May 02, 2006
Posts: 10147
Location: Abiquiu, NM, USA
|
| Posted: Fri Nov 21, 2008 9:46 am Post subject: |
|
|
| Quote: |
When the idle is too high, the computer compensates to bring it down, then the idle raises, ECU brings it down again, over and over and over.
|
No, sorry, there's no subtractive function in the Digijet ECU that "brings the idle down". But it could fluctuate idle if the TB switch is opening and closing, because then the ECU is going into and out of its idle injection mode. Or it could be getting an erratic signal from the Temp2 coolant sensor, so the ECU is in idle mode but it makes idle mixture rich, then lean, etc. Or the ignition could respond to fluctuating idle vacuum and cause the surging up and down. But the ignition and FI ECU are totally independent systems on Digijet, only sharing the rpm signal from the Hall gen in the diz via the Digital Idle Stabilizer.
The single item in nelinor's parts list that conceivably could have killed the motor is the Digital Idle Stabilizer box. The test to have done at the time would have been to simply unplug the two round connectors from the DIS and plug them together. If it was the DIS at fault, the engine would have started and run again, albeit with a possibly low or weak idle. You can bypass the DIS and run like that indefinitely if you like; everything downstream is blind to whether it is actually there or not.
The DIS is an additive device; it senses engine rpm because the Hall signal from the diz passes thru it. If it senses the engine is at idling speed but the speed dips too low, it phases the rpm signal waveform forward so the outgoing rpm signal is a little ahead of the base signal it is receiving from the diz, effectively advancing the idle timing a bit by making the ignition transistor control unit do what it's already doing, but a little bit earlier relative to the diz position. It varies the phasing dynamically to keep the engine idling within a programmed range, but it is only additive, meaning it can only advance and speed up, not retard and slow down the idle speed. That's why the base idle speed is set with the DIS bypassed, and it is lower than the ideal speed; the DIS can only add, not subtract, from idle speed. The ECU is, again, blind to what the DIS does to the rpm signal.
The Digijet ECU is pretty passive in all this. It's reading the rpm signal being passed on to it after the DIS, comparing it to the load and intake air temp signals from the AFM and engine temp signal from the Temp 2, and pulsing the injectors to give the appropriate amount of fuel. If the TB switch is properly closed at idle, the ECU is using it's idle quantity program, but it's incorporating those signals nonetheless and even if the TB switch is open, the ECU will still integrate those signals to determine the appropriate injector pulsewidth. The Digijet ECU gets no info about ignition, it only knows the present rpm.
The problem is that the rpm signal passes thru the DIS on its way to the ignition transistor control unit and on to the ECU. When the DIS starts to go, often the signal is degraded or interrupted by the DIS. This can make for wanky idle, eventually progressing to missed injection pulses and missed ignition events by both the ECU and ignition control module. Sometimes the signal can be interrupted completely, so the engine dies. In any of these instances, if you bypass the DIS and the engine runs normally again, that shows that the DIS is at fault. DIS's fail frequently by this age, so this isn't at all unusual. Luckily you can bypass and run without it with no problems, and new ones aren't very expensive.
So, for the surging idle, you should rule out vacuum leaks, but given the work you just did I would first pay attention to a few other things:
Make sure the Temp2 sensor is firmly seated in the water housing and it's connector is very clean and tight. Check continuity between the sensor body and chassis ground; there should be no measurable resistance.
Then make sure the TB switch is closed and staying closed at idle; try unplugging the two wires on either of the TB switches and patch them together. If it now idles smoothly without surging then the problem is that the switch isn't staying closed at idle. This can be due to wear in the throttle body shaft or just misadjustment of the switch itself.
The time delay after startup before the idle gets bad suggests trouble in the O2 sensor signal. This is very common when replacing the sensor because the signal wire gets manipulated which often shows up unseen damage in the signal coax wire where the signal-carrying center core gets shorted to ground. If it starts OK, runs fine for a short time after startup (from up to 10 minutes when cold to only about one minute when the engine is already warm), but then runs very rich, this could be the problem. Confirmation of it is when you restart the engine, it goes back to running fine, and gets bad aain after the same short time delay. If this is the case, the easy test is to let it idle when it is running badly, with everything connected as normal, and put a jumper between the o2 sensor connector and ground. If the idle degrades when you ground the signal wire, then you do not have a shorting problem there. If there is a change in idle to rich when you ground the connector, you want to test further: disconnect both the o2 sensor and the ECU, and test for continuity between the sensor signal connector at the free end of the green coax cable and ground. Wiggle the cable and make sure the signal core never shorts to ground.
I would rule out these things because the work you did could have induced problems in any of them. If all these things check out, then spend some time hunting down vacuum leaks.
_________________
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 |
|
|
hansh
Samba Member
Joined: January 13, 2004
Posts: 615
Location: Bend, OR
|
| Posted: Fri Nov 21, 2008 10:11 am Post subject: |
|
|
Tencent: I have a hard time believing that you could be wrong on this however, I spent a lot of time tracking down my surging idle and learned a lot about these early WC vans now. I think an academic debate on the issue is healthy and intend no disrespect.
While looking for a TB switch in my vicinity (Bend and PDX), I was told by all of the reputable shops I contacted that it is an impossibility that the switch on the early WC vans could cause this problem and that it is the ECU compensating for an idle speed that is too high. I was leery but, it spurred me to look elsewhere and actually find the problems.
I have verified this myself several times. The switch is capable of leaning the mixture or richening the mixture. If it is richened, the idle speed could increase and cause surging. This signifies that there is some other problem allowing the idle speed to increase significantly and that the problem is being flushed out via the idle switch's function.
Edit above: Actually, I can't remember if it just increases idle speed or changes mixture but, regardless, either method would just flush out another problem.
I am able to cause a cycling idle by, opening my idle screw excessively, advancing my timing excessively, removing a vacuum hose, or switching the the advance and retard hoses on the vac can, or anything else that increases the idle speed. I am not however, able to replicate this issue by disconnecting the idle switch, connecting it's wires together, or misadjusting the screw. Any of these methods would disable switch or change the signal of the switch to replicate a non-functioning switch. Other than a loose contact point within the switch, which I'd assume to be extremely rare (as none of the shops I talked to had ever seen this), I can't imagine how the idle switch could switch itself on and off. A loose contact point would not connect and disconnect the switch in a rhythmic pattern.
Another edit: I was also told that the DIS is capable of this symptom if it is malfunctioning. It is very quick and easy to rule this one out.
_________________
‘85 ej22
'84 Westfalia |
|
| Back to top |
|
|
nelinor
Samba Member
Joined: March 21, 2008
Posts: 98
Location: Ventura County / SoCal
|
| Posted: Sat Nov 22, 2008 2:59 am Post subject: |
|
|
Thanks for all the input. You guys rock!
I check the temp II sensor again. I also checked the O2 sensor. There was a loose connection.
Set the timing again.
Van is purring!!!!!
Thanks again sages.
-nelinor
_________________
1983.5 VW Vanagon Westfalia - manual - EJ22
2002 Subaru Impreza Wrx
1990 Nissan 300zx tt |
|
| Back to top |
|
|
ftp2leta
Samba Member
Joined: October 11, 2004
Posts: 3271
Location: Montreal
|
|
| Back to top |
|
|
240Gordy
Samba Member
Joined: May 15, 2008
Posts: 2354
Location: Vancouver, BC
|
| Posted: Thu Dec 11, 2008 6:13 pm Post subject: |
|
|
brilliant, thanks to Ben and Tencent for this. I have removed the DIS and cranked down the idle screw and will go for a drive to check out the results.
_________________
Tencentlife said,
"So, now that you know what you're doing, go to town."
2010 GOLF TRENDLINE 2.5
1985 GL now with more! a 2.1L
H&R SPORT(RED) Springs FRONT , SLAM SPECIALTIES RE6 AIRBAGS REAR |
|
| Back to top |
|
|
|
