1 2
deveous9
deveous9 Reader
11/30/11 5:38 p.m.

Have been working on my newly acquired 83 MK! Rabbit and it does need some work. It will start and idle fine but it will not drive right. I can only get good acceleration under WOT and when I attempt to cruise at lower speeds I am unable to pick up speed. It will die out when accelerated slightly but the funny thing is that when I placed the car on jack stands it can go as fast I want. After spending time reading the hayes manual and various online resources, I have to suspect a faulty fuel delivery system. I am also suspecting the o2 cold start system. Here are a list of things that I will look at:

fuel accumulator
fuel pump o2 sensor pressure relief valve fuel distributor fuel injectors

Any advice on something I should be looking for?

Cone_Junky
Cone_Junky HalfDork
11/30/11 5:51 p.m.

As with all CIS vehicles, especially VW, check for intake leaks. Air boots, vacuum lines, injector o-rings, etc. Also check for proper operation/adjustment of the idle and full throttle switches. The fuel pre-pump (in tank) is also a common failure and gives similiar symptoms.

Curmudgeon
Curmudgeon SuperDork
11/30/11 5:53 p.m.

(Previous '83 GTi CIS owner) First thing: check for vacuum leaks! A big leak between the throttle body and the CIS plate/injector 'box' will make it do weird things. About the only real way to test the other stuff is with a good fuel pressure gauge and a Bentley manual.

deveous9
deveous9 Reader
11/30/11 6:36 p.m.

So there is a pump in the gas tank? I was under the assumption this model vehicle only had a single electric fuel pump sitting under the vehicle next to the fuel tank. I am looking into making my own fuel pressure tester because it looks like fuel pressure might be a culprit to my problem. I will check for possible air leaks.

vwcorvette
vwcorvette HalfDork
11/30/11 7:26 p.m.

83 rabbit/Gti has one pump as you found outside the tank by the pass rear tire. There are a myriad reasons for CIS to have a problem. As a matter of fact I took the title of this thread to be a statement of fact! Others are correct to check for vacuum leaks. Find a good quality fuel pressure gauge--you'll need one that can allow for system and control pressure readings. Bentley manual is your friend here. And the little blue book from Bosch. Have you replaced the filter? Checked the screen in the warm up regulator? Make sure the intake manifold gasket is not blown out.

turboswede
turboswede SuperDork
11/30/11 7:28 p.m.

I would buy the proper CIS gauge set. They aren't expensive and it's a chance to buy another tool :)

Flogger00
Flogger00 New Reader
11/30/11 10:02 p.m.

Good advice in this thread. I'll add, though, that the symptom of revving on jackstands working better than while on the road suggests even more strongly a vacuum leak at the boots, especially at the throttle body. A split boot can be hard to spot. You really have to flex it to see if it opens up. Also, in looking for vac leaks - don't forget the injector o-rings. They're cheap to replace, and you'll probably want to check the injectors for volume and pattern anyway.

porschenut
porschenut New Reader
12/1/11 5:36 a.m.

I second the vacuum leaks or possibly low fuel pressure under load. On jack stands you have no load on the engine, either of these may not be observable.

wclark
wclark New Reader
12/1/11 6:00 a.m.
deveous9 wrote: ...I can only get good acceleration under WOT...

Excellent advice so far. However this part suggests to me that a vacuum leak may not as likely as fueling issues. Once a few years ago before I came up with a wideband fuel enrichment system for my race (CIS-E) GTI, I was at a dyno looking to tune for max power. At WOT the stock enrichment was nowhere near enough so we started adjusting the CO control ( the adjustment that sets the relationship between the airflow sensor plate and the primary fuel metering valve) so at WOT we could get enough fuel to produce power (.87 Lambda was the point we found). However at part throttle acceleration the engine was horribly rich and it ran like a pig because the CO adjustment was way beyond what the DPR could correct by going lean and the Lambda was in the .7-.8 region. At idle it ran sort of OK.

Anyway my point is, it is worth checking your AFR (even a narrow band will give you a clue) to see if you are rich or lean when the engine is stumbling. Just attach a voltmeter to the O2 sensor and go reproduce the symptoms while observing it is all that is needed for initial diagnosis. This will also give you some insight as to whether the O2 sensor has gone bad.

Curmudgeon
Curmudgeon SuperDork
12/1/11 7:07 a.m.

Yeah, maybe someone has been fiddling with the mixture screw (plate stop screw).

On the VW setup, there's an aluminum plug where the brass one is in this pic. If it's missing there will be a big vacuum leak. If the hole is full of sealant, someone's been fiddling with it. If there are vacuum leaks, it will be impossible to adjust the thing properly.

16vCorey
16vCorey SuperDork
12/1/11 7:31 a.m.
Curmudgeon wrote: Yeah, maybe someone has been fiddling with the mixture screw (plate stop screw). On the VW setup, there's an aluminum plug where the brass one is in this pic. If it's missing there will be a big vacuum leak. If the hole is full of sealant, someone's been fiddling with it. If there are vacuum leaks, it will be impossible to adjust the thing properly.

Not true. Only the CIS-E has the aluminum plug. The CIS cars had a little rubber stopper that is on a wire with a loop for a handle(that almost none of them have anymore). Either way, that won't cause a vacuum leak. The air that it will suck through that hole is on the intake side of the airflow plate, so it will be metered. The only thing that hole will cause is a little unfiltered air to get it. If air gets in after the plate, it's unmetered, and that's bad.

7pilot
7pilot Reader
12/1/11 7:48 a.m.

It's been a while since I played with these cars, but I'd suggest removing the boot over the air flow sensor and using a pen style magnet to pull up on the sensor plate via the hex bolt head in the center. As you pull up , you want to feel uniform resistance to the pull. The resistance feels kinda like wading in molasses. If the resistance to the pull is not uniform, clean the sensor plate and its hinge. Then undo the fuel lines and remove the fuel distributor. Remove the plunger and clean it and the orifice that it rides in. m

16vCorey
16vCorey SuperDork
12/1/11 7:52 a.m.

Bad picture, but this is what the mixture adjustment plug looks like on the early cars.

But yeah, they're almost extinct.

Curmudgeon
Curmudgeon SuperDork
12/1/11 8:11 a.m.

My '83 GTi (US built) had a tapered aluminum plug. I had to drill into it, run a screw in then pull it with Vise Grips. You are definitely right that once removed it's not a metered air leak.

car39
car39 HalfDork
12/1/11 8:14 a.m.

Plugged injectors used to be a big problem with the early Volvo CIS cars. That, and the millions of vaccum leaks

16vCorey
16vCorey SuperDork
12/1/11 10:49 a.m.
Curmudgeon wrote: My '83 GTi (US built) had a tapered aluminum plug. I had to drill into it, run a screw in then pull it with Vise Grips. You are definitely right that once removed it's not a metered air leak.

Weird. I've only ever seen the aluminum plug on the CIS-E cars, because you weren't supposed to adjust CIS-E.

But back the the original question, here's what I'd do.

  1. Check VERY thoroughly for vacuum leaks.

  2. Check the fuel pressure at the fuel distributor inlet.

  3. Pull the injectors, take the intake boot off of the fuel distributor metering plate, hotwire the fuel pump relay, turn the key on, and raise the metering plate with a magnet and see what kind of spray pattern you have.

  4. If that doesn't tell you anything, get the CIS fuel pressure gauges and the Bentley.

deveous9
deveous9 Reader
12/1/11 11:30 a.m.
16vCorey wrote:
Curmudgeon wrote: My '83 GTi (US built) had a tapered aluminum plug. I had to drill into it, run a screw in then pull it with Vise Grips. You are definitely right that once removed it's not a metered air leak.

Weird. I've only ever seen the aluminum plug on the CIS-E cars, because you weren't supposed to adjust CIS-E.

But back the the original question, here's what I'd do.

  1. Check VERY thoroughly for vacuum leaks.

  2. Check the fuel pressure at the fuel distributor inlet.

  3. Pull the injectors, take the intake boot off of the fuel distributor metering plate, hotwire the fuel pump relay, turn the key on, and raise the metering plate with a magnet and see what kind of spray pattern you have.

  4. If that doesn't tell you anything, get the CIS fuel pressure gauges and the Bentley.

Going to pull the fuel pump and bypass the accumulator and test the injectors at the same time. Going to buy myself a fuel pressure tester also. I forgot to mention that the ac compressor has been removed and from what I have been reading the warm up cold start system works with conjunction with the ac. I have to look at a wiring schematic and look if anything has been tampered with. the fuel distributor does not have that key attached to the screw, it does have a tiny hole where a hex size driver fits.

Man I sure hope its not the fuel distributor!!!

noddaz
noddaz Reader
12/1/11 2:10 p.m.

Go with simple first. Check the airboot at the air box for cracks, dents, splits, missing plugs ect.... Then check the boot at the throttle for the same... Don't jump to the expensive stuff first. Let us know what you find...

Curmudgeon
Curmudgeon SuperDork
12/1/11 5:41 p.m.

That tiny hole right between the big air boot and the fuel distributor is where the allen wrench fits. Mine used a 3mm.

Bosch says you cannot rebuild the distributors. I know for a fact they can be rebuilt as long as the castings are good, there's a couple of places here in the States that sell them.

This guy rebuilt his own:

http://forums.bimmerforums.com/forum/showthread.php?t=1123823

chrispy
chrispy New Reader
12/1/11 6:21 p.m.

In addition to what was stated here, there is a great K Jet resource on bimmerforums.com in the E21 section http://forums.bimmerforums.com/forum/showthread.php?p=15904137#post15904137

benzbaron
benzbaron Dork
12/1/11 9:00 p.m.

This sounds like a vacuum leak but hell if I know.

A couple of guys on the benz forum rebuilt their fuel distributors. I guess sometimes the fuel metering plunger in the distributor will get plugged or bind from crud. The big issue with the distributor is the two mating surfaces don't use a gasket so you have to be carefull when reassembling. I guess some people use indian shellac to reseal them. Another issue the fuel distributors have is there are little tiny screens at the top of the distributor where the hardlines attach which can become clogged and prevent fuel from getting to a cylinder.

If the car runs and idles I would throw a couple of cans of fuel system cleaner in and run the car, it could be a stuck injector. You can take a long handle screw driver or stethoscope to the injectors at the head and listen, the injectors should make a buzzing sounds and all injector should sound similar. I actually found a nasty miss on my car this way by listening to the injectors.

Another thing to look at is the fuel pressure regulator(warm up regulator). Controls fuel pressure to the distributor. There is a vacuum enrichment on it, I don't know much about the FPR but it plays an important roll in the CIS system.

Flogger00
Flogger00 New Reader
12/2/11 12:11 a.m.

First off, don't mess with the idle mixture screw. That's all it affects - idle. Not your problem, and adjusting it without having everything else exactly right first can have you chasing your tail. Also, benzbaron seems to have forgotten that CIS injectors are mechanical - no pulses to listen for. IME people freak out over CIS because it's different. When it goes wrong, though, it's usually something simple. Start w/the easy/cheap stuff (and oddly, twisting the idle mixture screw is not "easy" since it can mess up other diagnostics and would barely affect the problem you're actually chasing anyhow). What is easy and cheap? Check for vacuum leaks. Check volume and pattern of injectors (noting metering plate smoothness while doing so). Check fuel pressures. Check O2 with VOM and frequency valve with dwell meter. Along the way you'll find the problem.

benzbaron
benzbaron Dork
12/2/11 1:45 a.m.

I only learned this way to check injectors from the Probst CIS troubleshooting book so it must be bull, right?

Have you ever actually tried listening to a CIS injector? Take an injector to 80psi with compressed air and then tell me it doesn't make a sound. The CIS injector do indeed produce a buzz frequency and if you haven't tried listening to them don't tell me they don't make a sound! The CIS injector uses a spring and stop valve to make sure the nozzles don't open until 20psi. When the injector opens fully you will hear the spring buzzing with fluctuations in fuel pressure.

Honestly I took my leak down tester to a spare set of injectors with compressed air. A good injector won't open until 20psi, after that the injector will open partially and start buzzing. At 80psi an injector make a hell of a buzz.

fornetti14
fornetti14 HalfDork
12/2/11 5:50 a.m.

Former CIS '83 GTI owner here.

My advice - get with the local VW community and swap your CIS unit with one that you know works. I ended up bringing my car to a local VW specialist and for $250 they found two bad injectors plus the main source of my problem back then. I had a bad fuel tank. The OEM tank had a rubber baffle in the bottom that had rotted away. $160 for a new fuel tank and it ran perfect.

Curmudgeon
Curmudgeon SuperDork
12/2/11 8:11 a.m.

My GTi Bentley manual showed a way to test the injectors: pull them out of the intake, stick each one in a container then turn the ignition on. Use a magnet to pull up on the metering plate, the injector flow should rise and fall with the up/down movement of the plate.

I always wondered if it were possible to just remove the frequency valve completely on the later cars, that would take the electronice completely out of the picture. Never got around to trying it, mostly because that damn car kept trying to kill me.

1 2
QmKN5nRtqtW0HkSNJ5XtfPRGZw5yyYUG