8 cylinder front engine iconic vehicle
  • User avatar
  • User avatar
  • User avatar
User avatar
By Crumpler
#154028
Dual wideband sensors, one on each side of an xpipe. Innovate system gauge on driver's side and VEMS ECU on passenger side. The Innovate is a 4.9 sensor and VEMS uses a 4.2 sensor FWIW.

Tuning ECU on the passenger side, things seemingly going well, but noticed AFR's on the driver's bank were running two whole points higher. 14's on Vems side was getting 16's on innovate gauge.
I will have to get more data but my memory is that the numbers would get closer together at high loads/WOT, but consistently off at idle and cruise.

Today I switched the sensors. Put the innovate sensor on the passenger side and VEMS sensor on the drivers side.
Unfortunately, the number's reproduced: driver side (now vems) showing 16's at warm idle and passenger side (now innovate) showing high 13's.

Because two independent sensors are giving me the same lean reading I am assuming it's real, and now engine related.
Which sucks, but what are my rule outs here?
Injector malfunction?
Vacc leak? unlikely, just did extensive smoke and boost leak testing on the top end which it passed.
Exhaust leak up stream from O2 sensor?
Compression/leak down issues on that side? Recent compression test was 'ok' on that side, the only lower numbers (155) were on the other bank. No leak down testing done yet.
Ignition or grounding on that side?
What else?
Thanks ahead of time.
User avatar
By worf
#154327
Crumpler wrote: Wed Jun 08, 2022 7:09 pm Because two independent sensors are giving me the same lean reading I am assuming it's real, and now engine related.
Yes. But, we need to qualify that with the knowledge that we've no baseline data for a "factory-fresh" motor.

And, I assume, you don't have data from "before supercharger."

The difference between 14 and 16 is big enough that we can assume, reasonably I think, that there's a problem.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Injector malfunction?
Possibly. Have the injectors been cleaned and balanced?

The other possibility is a rail-to-rail fuel pressure imbalance. I'd have to give some thought to how that could manifest absent something outwardly obvious (like a partially crushed fuel rail.)

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Vacc leak? unlikely, just did extensive smoke and boost leak testing on the top end which it passed.
False air for this type of side-to-side imbalance would be intake connections on one side. But, with a smoke test you've ruled that out.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Exhaust leak up stream from O2 sensor?
Possibly. You should be able to hear that though.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Compression/leak down issues on that side? Recent compression test was 'ok' on that side, the only lower numbers (155) were on the other bank. No leak down testing done yet.
Leak down would be the next data to acquire.

Have the heads ever been off?

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Ignition or grounding on that side?
Ignition is half of one side and half of the other side (1,4,6,7 and 2,3,5,8) so, probably not the "upstream" part.

What do the plugs look like?

Plug wires? Plug wire deterioration on one side could cause problems. But, if there's no obvious misfiring, then that's hard to diagnose unless you stare at 'scope traces. One easy way to kill plug wires fast is to have any section between "bolt down" points "tight"; expansion/contraction cycles will break the conductor internally and leave no trace on the outside.

Grounds would affect the whole thing. The LH is grounded on one side and the EZF on the other. (Not half and half on both sides.)

Crumpler wrote: Wed Jun 08, 2022 7:09 pm What else?
I assume that there's no air injection (air pump) into the existing exhaust and no "hole" in the cat/crossover for air injection?

O2 sensors are more-or-less the same distance from the engine?

At this point, I'd put my dual wide-band setup on the tail pipes and see if I got similar results. Even with a crossover you'd see an imbalance that big at the tail pipes.

After a leak down test.
User avatar
By worf
#154331
One other thing....

remember the multi-page thread on TOS where the guy replaced his LH harness with a new Kroon harness? He didn't "Ohm it out" and thus didn't find where a shield pin had been swapped with a signal pin. Car would only fire one side of the motor.

But, that was an S4 with shielded Hall and Knock Sensor branches. Your EZF doesn't have those.

On the other hand...

... what's the LH/EZF harness like? Any "repairs" to injector connectors? MAF?
User avatar
By smiffypr
#154370
I'd bet on an injector underperforming, because it happened to me. I was trying to set the idle on an '80 and if I stuck the CO probe in one tail-pipe I got a very different reading from the other. One injector was hardly delivering anything at idle, but with K-jetronic it's relatively easy to test the injectors out of the inlets but still connected to the K-jet.
User avatar
By Crumpler
#154747
worf wrote: Fri Jun 10, 2022 12:50 pm
Crumpler wrote: Wed Jun 08, 2022 7:09 pm Because two independent sensors are giving me the same lean reading I am assuming it's real, and now engine related.
Yes. But, we need to qualify that with the knowledge that we've no baseline data for a "factory-fresh" motor.

And, I assume, you don't have data from "before supercharger."

Sort of , the innovate wideband has been in this location for years. The lean reading are new

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Injector malfunction?
Possibly. Have the injectors been cleaned and balanced?

The injectors were cleaned and flow tested at witch hunter, but that was 8-9 years ago so far from "know-good" right now.


The other possibility is a rail-to-rail fuel pressure imbalance. I'd have to give some thought to how that could manifest absent something outwardly obvious (like a partially crushed fuel rail.)

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Vacc leak? unlikely, just did extensive smoke and boost leak testing on the top end which it passed.
False air for this type of side-to-side imbalance would be intake connections on one side. But, with a smoke test you've ruled that out.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Exhaust leak up stream from O2 sensor?
Possibly. You should be able to hear that though.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Compression/leak down issues on that side? Recent compression test was 'ok' on that side, the only lower numbers (155) were on the other bank. No leak down testing done yet.
Leak down would be the next data to acquire.

Have the heads ever been off?
No. Never been off.

Crumpler wrote: Wed Jun 08, 2022 7:09 pm Ignition or grounding on that side?
Ignition is half of one side and half of the other side (1,4,6,7 and 2,3,5,8) so, probably not the "upstream" part.

What do the plugs look like?
Ok., had them out when i did compression testing.

Plug wires? Plug wire deterioration on one side could cause problems. But, if there's no obvious misfiring, then that's hard to diagnose unless you stare at 'scope traces. One easy way to kill plug wires fast is to have any section between "bolt down" points "tight"; expansion/contraction cycles will break the conductor internally and leave no trace on the outside.

Grounds would affect the whole thing. The LH is grounded on one side and the EZF on the other. (Not half and half on both sides.)

Crumpler wrote: Wed Jun 08, 2022 7:09 pm What else?
I assume that there's no air injection (air pump) into the existing exhaust and no "hole" in the cat/crossover for air injection?
No. But here's where I need some help. There is a small exhaust leak where the x pipe meets the downstream pipe on the driver's (lean) side, probably a foot downstream from the sensor. This is under the clamp collar. System not welded there. I assumed, sounds like wrongly, that downstream of the o2 sensor could not effect sensor afr? I also assumed that after the x, the exhaust was co-mingled and would be pretty equal as well?


O2 sensors are more-or-less the same distance from the engine?
Yes.

At this point, I'd put my dual wide-band setup on the tail pipes and see if I got similar results. Even with a crossover you'd see an imbalance that big at the tail pipes.
See question above, but will do this.

After a leak down test.
smiffypr wrote: Fri Jun 10, 2022 3:16 pm I'd bet on an injector underperforming, because it happened to me. I was trying to set the idle on an '80 and if I stuck the CO probe in one tail-pipe I got a very different reading from the other. One injector was hardly delivering anything at idle, but with K-jetronic it's relatively easy to test the injectors out of the inlets but still connected to the K-jet.
I was hoping that it would be this. Is it possible that one injector could move afr so much, I guess it could. I can also try and IR gun each cylinder on that bank and see. I ordered the 5.0 motorsports 25 pound injectors, so i will know here in a week or so.

Thank you Dave and Smiffy for the help!
Image
User avatar
By worf
#154777
Crumpler wrote: Sun Jun 12, 2022 6:56 pm
worf wrote: Fri Jun 10, 2022 12:50 pm Possibly. Have the injectors been cleaned and balanced?

The injectors were cleaned and flow tested at witch hunter, but that was 8-9 years ago so far from "know-good" right now.
Ok. Two ways to do this.
Easy way: send the injectors out. I've decided that I like osidetiger better than witchhunter.
Hard way: follow WSM procedure for fuel pressure leakdown test. Only works if you have flexible (rubber/PTFE) return and supply fuel lines.

Me? After 9 years? I'd just send'em out.

Crumpler wrote: Sun Jun 12, 2022 6:56 pm No. But here's where I need some help. There is a small exhaust leak where the x pipe meets the downstream pipe on the driver's (lean) side, probably a foot downstream from the sensor. This is under the clamp collar. System not welded there. I assumed, sounds like wrongly, that downstream of the o2 sensor could not effect sensor afr? I also assumed that after the x, the exhaust was co-mingled and would be pretty equal as well?
Hmmm... no. Don't think that would be it.

And no, even with an x-pipe the exhaust will not be completely co-mingled. It will be somewhat co-mingled. This assumes no futher co-mingling downstream of the x-pipe.

smiffypr wrote: Fri Jun 10, 2022 3:16 pm I'd bet on an injector underperforming, because it happened to me. I was trying to set the idle on an '80 and if I stuck the CO probe in one tail-pipe I got a very different reading from the other. One injector was hardly delivering anything at idle, but with K-jetronic it's relatively easy to test the injectors out of the inlets but still connected to the K-jet.
Crumpler wrote: Sun Jun 12, 2022 6:56 pm I was hoping that it would be this. Is it possible that one injector could move afr so much, I guess it could. I can also try and IR gun each cylinder on that bank and see. I ordered the 5.0 motorsports 25 pound injectors, so i will know here in a week or so.
Yes. Even one bad injector can screw things up.

Note that "bad" can be a leaky injector or one with a bad spray pattern.

And the WSM pressure leak-down test will only diagnose a leaky injector. Hence why I'd just send the injectors out.

On the other hand, since the heads have never been off you might have some leaky valves. You mentioned one hole that was off? Might be a bent valve or sufficient carbon buildup to keep a valve from sealing well. I've seen both and most 32v 928s have at least one slightly bent valve at this point in their lives.
User avatar
By Crumpler
#155104
Appreciate it Dave.
I will hopefully get leak down numbers this weekend.
The new injectors showed up. I was going to replace the original OEM set anyway.
I was pleased that they test injectors before shipment.
Image
User avatar
By worf
#155109
Crumpler wrote: Tue Jun 14, 2022 3:36 pm I was pleased that they test injectors before shipment.
And they provide dead time! Sweet!
User avatar
By Crumpler
#155994
Went for jackpot with the new injectors.

They are pushing more fuel, I had to adjust req fuel down.
The preliminary verdict is that they made a difference, but not a fix.
Now the bank is running one full point leaner, not two full points.

I will bite the bullet and do leak down testing tomorrow.
User avatar
By Crumpler
#156102
I changed plans, I was slow walking leak down because any hand turning crank requires removing the radiator ( this was poor design on my part with the supercharger project) because of the long crank pulley.

I was due for new caps and rotors so I changed those along with new plugs (old plugs looked fine when I pulled them this morning).

I’ve learned the hard way to smoke test anytime I touch the manifold, and it passed afterwards.

Started car and both banks identical AFR’s.
Not exactly sure why. Either there was some ignition issues or I fixed a vacc leak i didn’t know I had.
It’s a fickle town boys. :drunk:
SeanR liked this
User avatar
By worf
#156108
Crumpler wrote: Sun Jun 19, 2022 1:51 pm Started car and both banks identical AFR’s.
Excellent.

Continue to monitor. Never trust measurement data where N=1.
Crumpler liked this
78 in Fort worth

Ed that would look good with new tan carpets and b[…]

Ineos Grenadier

It’s not an SUV. It’s a hardcore offro[…]

This upcoming DFW First Saturday Breakfast will be[…]

Looks like I have one tracked down. Thanks guys[…]