some running problems and MAF sensor 7A

[Obli]

On my 90 20v, I found two splits in the intake hose. I only saw the second split when I removed the hose.

I also found a Y piece connector (from inlet manifold to charcoal canister cut out valve) with a leak. As my car isn't a quattro and doesn't have a vac connection to diff, the Y piece was capped off but the plastic cap had deteriorated and had a hole in it. From your image, it looks OK (no Y piece on yours but a direct connection).

[Mikes2]

Have you verified the fuel pressure is correct or have you tried a different fuel pressure regulator ?

[Lago Blue]

I wouldn't fret over not being able to do Vagcom if as you say there are no blink codes to pull (i.e.: you received a 4444 code (meaning no faults stored) on your code pull, correct?). There are certainly instances when there is something amiss that is beneath the fault detection threshold, and the path forward is simply traditional trouble-shooting, not buying and replacing components on just suspicion. A DMM and a vac pump and gauge are your friends here.

With a simple hand-operated Mityvac tool, one can apply vacuum to the various branches of the vac. system individually to prove they are sealed to negative pressure (unlike what a smoke-test might reveal), and they can also be left under vacuum say overnight, to prove they do not leak down over time. Audi famously used often short lengths of a fabric-covered rubber hose for a good deal of the small bore vac system connectors (for the various hard plastic vac lines throughout the vehicle) which will all now be leaking vac. It takes about 3m of good BMW stiff synthetic thick-wall vac hose to replace it all.

That your idle issue does not appear until after warm-up may suggest that the issue is not the ICV, but is of a magnitude that the ICV simply cannot respond sufficiently to cancel it out. For instance, it may be that the fuel tank vent is not connected to the I/M; until after warm-up. If that vent line (or any other break in that system) where broken upstream of the vent solenoid valve, will then after warm-up you may have a switched on vac leak.

[Bowie69]

One is a hitachi valve, and one Bosch, so likely very very different. Most other bosch valves seem to work for me, when I tried it. Volvo is a possible source, I think?

[djpeck]

Although when serviceable the two ISV types would both likely respond equally to signal, I don't think they share the same hose or mounting arrangement.

Another way to discern whether there is a vacuum leak somewhere, is to temporarily attach a vacuum gauge (such as a simple MityVac) to the I/M for starters, to read the inches of water column and run the engine till it is warmed up and your issue appears.

It might be that there is a vac leak that wasn't revealed by the smoke test, such as CCV hose from crankcase to the I/M, the oil fill cap and/or dipstick seal(s). or maybe the 3rd fuel line (the vent line) back to the tank or the gas fill cap seal? Edit: Almost forgot, the hose back to the diff lock and the vac reservoir there too.

Hello, I don´t think ISV mounting could be a problem. My doubt is if it has the same electric values.

Regarding Vacuum leaks, yes... Maybe I have to check again...

Maybe the problem comes from the I/M or collector assambly, and leak appear whith metal dilation...
Or maybe it could be the ecu, or any sensor that goes out of values whith hit....


It is a shame that I couldn´t connect VAGCOM in this early 7As...
Maybe is eassier to swap ECU to 4 connector

[Lago Blue]

Although when serviceable the two ISV types would both likely respond equally to signal, I don't think they share the same hose or mounting arrangement.

Another way to discern whether there is a vacuum leak somewhere, is to temporarily attach a vacuum gauge (such as a simple MityVac) to the I/M for starters, to read the inches of water column and run the engine till it is warmed up and your issue appears.

It might be that there is a vac leak that wasn't revealed by the smoke test, such as CCV hose from crankcase to the I/M, the oil fill cap and/or dipstick seal(s). or maybe the 3rd fuel line (the vent line) back to the tank or the gas fill cap seal? Edit: Almost forgot, the hose back to the diff lock and the vac reservoir there too.

[djpeck]

Does this car have an idle switch on the throttle body?
If so, maybe it’s not closing.

Yes, I checked it and it is working correctly.

I used the car during this weeks and it works without changes. It is still increassing revs etc.

I would like to know if it is possible to swap ISV valve from old one to later one

• Early 1990 cars (CQ VIN <= 13610) have the Hitachi unit shown on the left -- Audi P/N 034 133 455 A
• Later 1990 cars (CQ VIN > 13610) and all 1991 cars have the Bosch unit shown on the right -- Audi P/N 034 133 455 B

A later one is easier to find it. So I would like to try it.

[newsh]

Does this car have an idle switch on the throttle body?
If so, maybe it’s not closing.

[djpeck]

Drive it, it needs to learn.

Okk, lets see what happen using it, thanks!

[Bowie69]

Drive it, it needs to learn.

[djpeck]

The fuse in the fuel pump relay fixes the idle valve current to 150mA, so that is why it fixes the idle issue.

The ECU remains in this mode until you cycle the ignition again, so that's why you are seeing the 'fix' stay.

You could well have idle speed set too low, give it a go, also, the ECU has to have time to learn properly, so don't expect perfection right away, you may get some stalling.

Something like OBD11 will read codes and the real time data, which can help in diagnosis.

At the end, we regulated the CO close to 1 with the fuse on, and after remove it, it falls to 0.75 aprox.
After this regulation, the engine goes better than before, but not as good as it could be. The revs drops so slowly, above all going down 2k revs.

Apart from that, I tried with some obd machines, and I couldn´t connect with ECU. In this toppic it is said that early 7A engines doesn´t work with obd machines. do you know if it is true? Mines is the old hitachi version ( 2 connectors).

https://www.s2forum.com/forum/technical/engine-and-turbo/38028-7a-diagnostic-code-reader-that-works/page2

Was an ordinary OBD1 blink-code pull done? As there are factory fault codes which list some of the symptoms you describe above (i.e. fuel consumption, high idle speed) and possible fixes, that may be helpful. Standard diagnostic procedure might start with a code pull, making note of any that display, erase them all then see if any re-appear after the engine is run again.

Yes, I have done that light diagnostic test and there are no faults. At this moment I don´t know what I can do...

[Lago Blue]

Was an ordinary OBD1 blink-code pull done? As there are factory fault codes which list some of the symptoms you describe above (i.e. fuel consumption, high idle speed) and possible fixes, that may be helpful. Standard diagnostic procedure might start with a code pull, making note of any that display, erase them all then see if any re-appear after the engine is run again.

[Bowie69]

The fuse in the fuel pump relay fixes the idle valve current to 150mA, so that is why it fixes the idle issue.

The ECU remains in this mode until you cycle the ignition again, so that's why you are seeing the 'fix' stay.

You could well have idle speed set too low, give it a go, also, the ECU has to have time to learn properly, so don't expect perfection right away, you may get some stalling.

Something like OBD11 will read codes and the real time data, which can help in diagnosis.

[djpeck]

I had similar issues with my 7A. After many hours of trying everything like MAF, lambda, throttle pot and idle switches, I moved on to ignition. It turned out that my distributor was (and still is) passing some oil through to the hall sender, then oil pools in the sender connector block and and affects the circuit. After cleaning out the oil, everything went back to normal. So I now just clean this out as a routine until one day I'll have to strip the cooling system down to get at it and do something with the distributor seals (my LPG vaporiser makes it very difficult to get to the clamp bolt).
So I'd suggest having a look in the dizzy cap and hall sensor plug for traces of oil.

I´ll check it. I don´t think it could be the problem because the distributor was fitted few years ago ( the engine was rebuilded in 2017).

Great work testing things! The smoke test result is encouraging.

Another unusual source of air leak is the small o ring on the idle adjust screw. Its well worth changing that as it will be squashed flat by now if orginal and will be letting air past. I found quite a strong vaccum around my idle screw, strong enough to feel with finger. = leak.

The fuse in top of fuel pump relay is to put the engine into CO and idle set up mode.

With the fuse in, the ecu does not get input from lambda, coolant temperature and idle valve.

This allows you to make changes without the ecu trying to control things. Its interesting to see that idle return is better with fuse in relay as it suggests that either lambda, idle valve or coolant sensor is confusing things.

When you checked the CO and got the low figure, was that with fuse in or out?

It will probably be different with the fuse in as the lambda sensor is not bringing it back under control.

Try setting CO and idle with fuse in and see what happens. 0.5-0.7% is factory figure.

When finished, remove fuse and rev engine to 3000rpm. This returns the engine back to ECu control and normal running. Hopefully the improvement stays!

I think what happened on my car is that somewhere in its history someone adjusted co and idle screws without the fuse being in relay and it got so far out that the ecu struggled to get it right.

Good luck!

Yes, I remember to change that o ring last year, but I will check it again.
At this moment, Lambda sensor is new ( bosch brand), coolant temperature sensor G62 is new too ( febi brand. I will change to the original one ( bosch) to try again). Fuel bump relay is new too.

I agree with you that the problem could be the CO and idle adjust. CO was readed without fuse (0.15% aprox) ¿maybe is to low?
My Mechanic told me that screwing the CO bolt, the value didn´t change ( without fuse). That why we think the MAF could be in bad shape.

The engine runs perfect when I put the fuse for the first time. After that, it runs as always with or without fuse.

At this moment, the main different when I use the fuse is the idle speed. Without it, idle goes to 2000rev, with it, it goes to 1200rev. I have to try again because maybe idle bolt could be to high.

Apart of that, rev goings down slowly in both cases at this moment.

We will check CO again with fuse switched on and see what happens...

[steve briance]

Great work testing things! The smoke test result is encouraging.

Another unusual source of air leak is the small o ring on the idle adjust screw. Its well worth changing that as it will be squashed flat by now if orginal and will be letting air past. I found quite a strong vaccum around my idle screw, strong enough to feel with finger. = leak.

The fuse in top of fuel pump relay is to put the engine into CO and idle set up mode.

With the fuse in, the ecu does not get input from lambda, coolant temperature and idle valve.

This allows you to make changes without the ecu trying to control things. Its interesting to see that idle return is better with fuse in relay as it suggests that either lambda, idle valve or coolant sensor is confusing things.

When you checked the CO and got the low figure, was that with fuse in or out?

It will probably be different with the fuse in as the lambda sensor is not bringing it back under control.

Try setting CO and idle with fuse in and see what happens. 0.5-0.7% is factory figure.

When finished, remove fuse and rev engine to 3000rpm. This returns the engine back to ECu control and normal running. Hopefully the improvement stays!

I think what happened on my car is that somewhere in its history someone adjusted co and idle screws without the fuse being in relay and it got so far out that the ecu struggled to get it right.

Good luck!