PD Injector Issue - Advanced Troubleshooting

[A5INKY]

Car is a customer's '05 Passat with a BHW PD engine. Customer bought this car after it had changed hands several times with 3-4 different shops not being able to fix it. No surprise, tons of work had been done and most of it of very poor quality. After a general detooefing of substandard previous work, issue still plagues it, a cylinder #3 misfire. This fault shows up in measuring block 23 only intermittently under load while driving and values for that injector only go sky high when the fault is occurring.

To try and solve it a bunch of work has been done:

• Found loose bolts under the VC including injector hold-down bolts after other shop put in a Colt cam, so pulled the head, new correct gasket and checked what can be checked. Reassembled properly with proper new parts.
• While head was off, found that previous shop had replaced #3 injector with new of different P/N. So found old injector in box of provided parts and sent all 5 to Drivbiwire for new nozzles and full rebuild. He used old #3, said it was good now.
• Swapped head injector harness loom with known good used one while head was off just because.
• Had ECU tuned to Stage 1 from TD Tuning with EGR corrective "stuff" done
• Checked wiring harness back to 60 pin plug on ECU with aggressive wiggle test, not open

So, I have done a process of elimination and I think am down to a failing injector #3 driver in the ECU itself. Only thing I can think to do before swapping ECUs is swap #3 and #4 injectors to make sure that (if by some astronomically unlikely probability) the bad injector has gone undetected and keeps ending up in the same #3 hole.

Anyone seen a failed ECU in one of these yet? Anyone see some troubleshooting step I am missing?

Thanks in advance

 

[peiphil]

I came accross that same issue with a Bosch ecu in a different vehicle.
The injector drivers were larger and easy to see in this case and it was a broken solder joint.
Had to use a magnifying glass to see the small ring crack in the joint and like yours the problem was intermittant
I dont know if you can get inside of these ones or if they are potted inside but might be worth a look
The one I fixed never gave a problem since 2 years now

 

[ryanp]

So is it actually showing an Injector Fault on Cyl3? Or just dodgy readings in Ch23?

When the Injector Bolts were loose, has it damaged the Port?

Ry

 

[A5INKY]

OK, since the ECU was out for the wiring loom test I went ahead and pulled the ECU cover as Peiphil mentioned - seems plausible. It was not potted. Careful inspection of the injector control chip showed no bad solder connections.

Pulled injector 3 and 4 and swapped them (thankfully I happen to have a spare set of TTY bolts ATM). Put it all back together and nothing. No running issue, and group 23 looks great. ***!

I hate it when I take something apart to find a problem, change essentially nothing and then it works... I want to know what was wrong.

Ryan, no codes, just the dodgy group 23 cyl #3 reading before. The value would jump massively relative to the others when I could catch it misfiring - no mistaking the correlation.
Edit: I checked the injector bores carefully when the head was off, no damage.

Guess I will hand it back over and see if it returns.

 

[thundershorts]

When the tuner had the ecu, any fault of that type would most likely show up. You didn't mention a compression test??

 

[A5INKY]

When the tuner had the ecu, any fault of that type would most likely show up. You didn't mention a compression test??

That is one of the issues, the misfire does not trigger a fault condition in the ECU. So many things do trigger a fault it is hard to understand why such a significant running issue as this does not. I can only surmise that it is because the reduction in fuel must not cause dirtier emissions so is therefore not part of the ECU's CEL setting scrutiny.

I had intended to compression test the car after a test drive at the end of the day. Unfortunately, I had not moved the tester to my shop out of storage yet. With all the trouble I didn't find and after taking with a few trusted tech friends, low compression on #3 was really the only remaining likely cause of the misfiring. However, the car never missed a beat during the long and fairly hard driven test drive. I still intend to test compression, but that being low enough to cause a misfire would not come and go like this. In fact, low compression from anything short of a crack should make for a worse cold misfire which is not this car's symptom.

 

[thundershorts]

driving this unit fairly hard would indeed keep boost up which in turn raises the relative compression in #3. Other possibility is Colt cam lobes on #3

 

[A5INKY]

driving this unit fairly hard would indeed keep boost up which in turn raises the relative compression in #3. Other possibility is Colt cam lobes on #3

Checked cam carefully when swapping injector positions yesterday, looked really good.

I will be checking the compression soon just because it is a good idea on a car with nearly 200k miles. The misfire literally comes and goes and does not seem temperature dependent at all. Compression tends to improve on a warm engine, if something was causing a lack of compression misfire then I would expect the misfire to to correlate to cold engine, it doesn't. On the contrary, the misfire (when present!) is only under load and even then it is only intermittent.

I just can't imagine a scenario where a low compression misfire would disappear for days and weeks at a time.

I am still wondering about a dodgy injector solenoid. Now that I've moved it to #4 hole I hope it will show up there next time it crops up. A least I will then have it nailed to a part that can be fixed.

 

[peiphil]

You might check the 12 volt supply to the injector bank paying close attention to #3
The ecu sends a negative signal to each injector when injecting
There will be 4 larger transisters in the ecu that fire the injector coils
Follow the curcuit in from pin 60 and double check as well.
the ucu determins a fault by measuring voltage drop
A bad conection in the negative curcuit to the injector or a bad 12 volt supply to the same injector will make the ecu sense the wrong voltage drop value.
It might also be as simple as a bad plug on #3 which is now working after being disturbed!
Happy troubleshooting You will get that gremlin!

 

[otty]

I had a very similar misfiring issue and to everyone's surprise after swapping injectors etc it turned out to be an intermittently bad tandem pump. Maybe do a pressure test on it while you are in there.

 

[Growler]

MoGolf had one like this a couple of years ago. IIRC he got/borrowed the special tool for checking the tandem pump pressure from Oilhammer to verify running pressures inside the tandem pump.

perhaps you could rig something up to check with? I know that the area is super tight back there on the Passats.

 

[A5INKY]

You might check the 12 volt supply to the injector bank paying close attention to #3
The ecu sends a negative signal to each injector when injecting
There will be 4 larger transisters in the ecu that fire the injector coils
Follow the curcuit in from pin 60 and double check as well.
the ucu determins a fault by measuring voltage drop
A bad conection in the negative curcuit to the injector or a bad 12 volt supply to the same injector will make the ecu sense the wrong voltage drop value.
It might also be as simple as a bad plug on #3 which is now working after being disturbed!
Happy troubleshooting You will get that gremlin!

This car has EDC16 which (found using my Google fu) has a single chip driving the injectors in the ECU. I always hesitate sticking a meter on circuits I don't completely understand, but the common wire coming out of the head is brown, VW's universal color for ground on these cars. I did a thorough job of checking that harness all the way the the ECU connector though. Wiggled the hell out of everything I could touch (which is pretty much the whole run) with a circuit tester connected. Never blipped open no matter what I did.

Thanks for the help and vote of confidence. The gremlin will show itself. More on that at the end of this post.

I had a very similar misfiring issue and to everyone's surprise after swapping injectors etc it turned out to be an intermittently bad tandem pump. Maybe do a pressure test on it while you are in there.

MoGolf had one like this a couple of years ago. IIRC he got/borrowed the special tool for checking the tandem pump pressure from Oilhammer to verify running pressures inside the tandem pump.
perhaps you could rig something up to check with? I know that the area is super tight back there on the Passats.

Sorry, I forgot to mention the fuel supply testing I did awhile back. I didn't check pressures but I did check volume from both the lift and tandem pumps per DanG144's experienced guidance. I thought perhaps the car may have had a failing older vane style tandem pump, but no, it the newer Bosch unit and it tested to flow within the volume spec Dan provided.

I won't rule the tandem pump out entirely though. I may check pressures before it is all over. I will say the car does not want to prime it's fuel system as easily as every other PD I have worked on. Took forever to get it to fire after the injector work in spite of both pumps testing like they did. It was a real head scratcher at the time.

So, a little update. Drove the car again today after the 3-4 injector swap and could not (once again) get it to miss a beat. I did play with VCDS graphing function a bit though and got it adjusted to see group 23 behavior in a different light. Injector #4 (same suspect, different location) is slightly operating in a different range. I don't think this is significant though as it is within less than half the +/- 100 spec VCDS gives and much tighter than so many other good running PDs I have observed. What did catch my eye is when I changed the graphed pixel shift to "speed up" the display and see the resolution (for lack of a better term). Injectors 1 and 2 had a much higher rate of change than 3 and 4. It is a big difference, perhaps 3X faster.

I spoke to Drivbiwire since he rebuilt the injectors and he mentioned a solenoid part (a pin IIRC) that can be damaged (slightly bent) that can cause an intermittent issue. Apparently this cannot be found in bench testing. With all the evidence of unskilled repair efforts, I would not be surprised if some knuckhead didn't try and take an injector or two apart not knowing what a bad idea it is on a PD.

I wonder if this group 23 slower rate of change for two injectors may not be those solenoid pins hanging up slightly and every now and then one really hangs enough to cause starvation on a cylinder. May be ringing DBW back to discuss this possibility, he was happy to help.

Thanks all who are following this thread. I LOVE a good troubleshooting challenge

 

[Frans]

Which parts are replaced inside the rebuilt injectors? Only the nozzles or also the pump elements? OEM bosch parts used?

 

[oilhammer]

Sounds like you are on the right track since the suspect cylinder moved/changed.

 

[thundershorts]

At this point, other than find another injector, driving the car a thou or so miles to see if it gets better, might be an option.

 

[imahaare]

I too have misses on #1 and #4 in BEW PD. I pinched the connectors a tiny bit to get them to connect better. All pretty snug. No change resulted. #4 stopped working altogether. I have now replaced #1 and #4 (yesterday) and after 30 minute test drive all is well.

In my case the failure initially only occured after getting the car fully warmed up over 190 degrees. Never missed below that temp. When it first started missing it would be after 30+ miles of highway driveing in the mornings and maybe 10 miles in the afternoon. Then it degraded from light miss to persistant miss to full on failure.

at 285k miles I wonder if its clogged nozzels or exceeing mean+6sigma life expectancy?

I ohm'd both injectors and got 0.2 and 0.6 respectively.
also applied 12v from fullwave power supply and both injectors buzz (60 cycle chatter) although #4 is louder than #1.