Update: Troubleshooting EFI issues on 1.7l D-jet
 

Ok, I found another culprit.

First, my head temp sensor was bad: was never allowing the mixture to lean.

Second, I sent my fuel injectors out for a rebuild/flow, one was completely bad: no electrical activity within with power, so, for $60, Autozone has one coming for me.

I could blow through it barely, which means it was dripping, barely spraying all the time, which is why the car was still running, but since heat expands items, it might also be the reason the car ran excessively rich once warm (15-20m into driving). We'll see, it's going all back together with new braided 7mm EFI hose, EFI hose clamps, & everything painted. :)

Mark

so your Head temp sensor didn't vary the resistance as it warmed Up?

a leaky injector would likely make just 1 cyl rich, not all of them.

a leaky cold start valve could make all burn rich though, as will a faulty MPS (ask me how I know) . So check both of those as well.

MPS appears ok, decel valve appears within spec, AAR valve appears to be working ok, no resistance on a hot head temp switch, & a internally leaking injector: I had one cylinder running lean, that was probably it. I'd lose a cylinder or two & the car would then die at idle, once the car was warm. Fuel consumption was suggestive of a EFI'd V8 rather than a flat 4

Looks like you have def. found your trouble then.

I hope so! I need to start working on my neglected 930 that's hogging my lift.

Mark

Look at the 7mm ID, 13 mm OD line they have here. Spendy, but better than the cloth braided stuff, IMHO. That clth covered stuff is so 1950's, which is okay for music, but, maybe not fuel hoses.

I believe there was a recall of the cloth braided stuff for 914's, but none of the North American VW products.

The fabric hose soaks up any water that hits it including the acid water. Rots it out quickly so Porsche came out with the rubber coated acid resistant fuel line to solve this problem.

People seem to think the fabric stuff it cooler or more OEM but its not correct unless you like the problems of the fabric hose.:confused:

Crap, ok, next problem: I've completely, completely gone through the whole top of the engine: all new fuel lines, rebuilt fuel injectors (one was bad, replaced with an Autozone "GP Sorensen" injector, all new hoses, etc.

My rebuilt distributor wouldn't fit due to the condensor being in the way, so I had to move to the opposite side of the disty to get it to seat. Wierd.

Anyway, engine fired up, but still running very rich (haven't verified fuel pressure, nor the "base setting" on the potentiometer on the ECU, but was running on 3 as one of my rebuilt injectors became seized within the last week. Ok, wierd.

Un-seized the injector, car runs ok, however, my new injector is flowing less than the others as I actually achieved some detonation while test driving. My new injector's ground strap was white, while the other plugs were about as black & sooty as I've ever seen them.

Anyone else use the GP injectors before? I just double-checked the PN, it is indeed correctly stated on a second part number source for the 1.7l: 800-1108N. I understand the 2.0 injectors are larger, wondering if I have 2.0 injectors in my car...the part numbers on these yellow top injectors are: 022 906 031 A and the Bosch # is 0280 150 009.

:(

the yellow injectors are 1.7's. 2.0 inj are green i think.

i would be inclined to put a bosch oem injector in there. 1.7 ones are plentiful and cheap, I have some extras if you need some. That way you'd know voltages and resistances would be to spec. The fact you're having as much trouble with ind. injectors makes me wonder if you have some grit in the system that's clogging them.

disconnect and plug the fuel line to the cold start valve, in case it's leaking.

I tried the cold-start disconnect: didn't do much, but #3 (new injector) was very lean compared to the remaining 3 which were freshly cleaned though black & sooty (the spark plugs).

Last week I received back the original 4 injectors, which had been balanced & flow-matched, though #4 was dead, hence the new injector from Autozone. I'm semi-inclined to go buy 3 more GF injectors. :)

Mark

Busdepot.com has the Bosch 1.7 injectors on sale for $39 which is a steal.
you have to look hard to find them on their site but use the 022 part number to search.

Forget using Autozone injectors as many people have had problems with those injectors

Excellent advice, thanks!

to me, everything points toward a bad MPS if everything else is as you say. find someone with a spare and try it on you car.

well, doesn't explain the one lean cylinder with the new injector. I'll buy one from the before mentioned bug place to replace my dead one, as the other 3 are freshly flowed.

MPS holds vacuum pretty well, not sure what other tests to throw it's way.

mark.

Ouch, $196 per injector as busdepot.com currently....hmm, might be better off buying 3 more new injectors from Autozone :)

wow I have a gold mine in parts in the back, then.

one lean cylinder means the injector is not working right, because the ECU sends the signal to a Pair of injectors at the same time, and if the "lean" one were working I suspect they would all be too rich. You've checked the CHT, and cold start valve, it's a rebuilt distributor (trigger points), so that leads to the MPS, in my opinion. check the part no's, and check to see if the plug has been drilled out, maybe someone, like myself, has adjust it to meet their purposes.

I emailed BusDepot just to clarify, but have not heard back yet.

If the banks of injectors are bank-fired, then I can swap the leads to the injectors to verify that I don't have a weak signal in my one lean cylinder bank. I can also swap the plugs to eliminate the possibility of having a poorly-conducting plug (I've had bad plugs before).

What's the CHT? Sorry, still learning my 914 acronyms.

My rebuilt distributor cam with used Trigger points, vacuum canister, points, & condensor: not bad for $69 right?

Drilled out which plug, the one on the MPS? I remember the epoxy being intact.

Mark

they fire like 1 and 3, 2 n 4, together, not each side but across the motor. and yes, you can switch the plugs to test the electrical signal.

you have me wondering though, you said you'd lose 1, then 2 cyls and it would die. Have you checked the grounds on the top of the engine block? particularly the one connected to the injector that is lean firing.

The epoxe plug yes. If it's intact, I would just cross check that it's the right part number, mostly.

CHT = CYLINDER HEAD TEMP sensor... you replaced it.

yes, new cylinder head temp, new oil pressure switch, cleaned the ground wires (3??) that connecto to the splade connectors at the top bolt under the intake manifold. CHT was dead with heat, car now idles, though idles at about 1800rpm.

TB is shot, lots of play in the shaft, so ordered a good used TB from 914-parts.com

Hmm, I hate to keep throwing $ at this: I feel I need more testing.

BTW: is there a default setting to the ECU's potentiometer? My car is still running very rich, even with all the new parts. (verifying Fuel pressure tonight).

Mark

if you mean the knob on top of the ECU, that is only for idle mixture, doesn't affect the rest of the range.

Oh excellent. Ok. Well, My fuel pressure is sitting at 32psi in lieu of about 29, so once I have an OEM injector back in the saddle, I'll leave it there.

BTW: I'm running so rich that the car starts & runs perfectly w/o the cold start valve connected...eek.

Mark

The cold start valve only works at freezing or below, not likely this time of year in Austin. That may be an indication that things are working well as they should.

you mentioned idling at 1800 rpm, that could be the Aux Air Regulator allowing extra air for higher cold idle speeds. That should stop after a few min. They tend to stick, or leak air that causes high idle. I have mine blocked off because of that. you might pull off a hose and block it to see if it makes a difference in the idle speed while warm.

You could also pull the throttle body plug completely to see how it runs when not on the idle circuit, which is switched in by the throttle body when the throttle is closed. That is when the ECU "potentiometer" comes into play.

Well, the idle is now between 1800-1400, but consistent throught the time the engine is running. I have 4 AAR vavles (3 from the local salvage yard) to test, so I can put another in there, see what that does. :)

Is there a write-up on testing the TPS switch on the TB anywhere? My manuals appeared vague in that area.

Thanks again, Mark

pretty much everything about DJET is here. We are all indebted to Pat for this page... Thank you Pat!

http://members.rennlist.com/pbanders/DJetParts.htm

Ha, surprise: as of lunch today, I confirmed that my "new" Autozone injector is not firing. I have a weak cylinder somewhere too, so currently, I'm running on 2-3 cylinders depending on the wind direction. :)

On the weak cylinder, if I unplug the injector, the idle rises for a 1/2 second, then falls again. when I reconnect it, the idle does the the same thing: rises for 1/2 second, then falls. Wierd. Usually, there's a definitive & continuous fall of rpm followed by off-cylinder vibration when you unplug one injector.

For now, I just unplugged the Autozone injected cylinder, I might do the same on the so/so cylinder: wonder if I can make it home on 2 cylinders??

I have a very steep 200' hill to scale on my way home today: this is gonna be interesting!!

TPS-http://www.pelicanparts.com/techarticles/914_cooley/914_cooley_throttle_switch.htm