[N]VH Issue? Pump Voltage? Nope, MAF vs RPM vs VE!

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Okay, so I'm confident in the cause of the problem, theoretically speaking, and I do not have the really expensive equipment to directly test for it. That said, when I swapped my stock harmonic balancer out for the custom machined assembly that included the modified fluidampr, and then also a 55A polyurethane front engine mount, the problem went from random to regular.

I cannot do redline log pulls on the open road, but when on a dyno this isn't an issue.

Initially this started out in 2nd or 3rd gear, cold engine mounts, loading the engine mounts a specific way, and then it'd happen 50-75% of the time. Now, it is a definitive harmonic event from 2nd to 4th gear (1 & 5 are too extreme that I don't test for it) to where upon steady state or acceleration driving up to 4200 RPM at the soonest, but no later than 4500 RPM, the engine cuts out. I'm extremely confident that relay 109 is being vibrated open and the ECU power is cut.

New relay 109's have held up well for a few months and then fall suspect. The vibration is not perceptible and due to the size of the relay arm and the RPM this occurs at, I doubt I'd be able to sense it.

I've talked with my tuner in the past and he's never heard anything like this. I've not seen it reported here either.

The only other things I could think it might be is the ignition switch, which is "OG" but has given me zero issues.

The primary possible solution I have is a rubber "seat" that is between relay 109 and it's home in central electronics.

Maybe I revert back to my stock harmonic balancer and have my aluminum pulley hub that bolts over the fludamper be balanced at a shop?
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
I've got such harsh motor mounts at this point that the dash rattles, haven't had electrical issues yet, but now you've got me worried. I've never had reason to disassemble a 109, so no idea if they're an easy fix or some box of wizardry like the wiper relay.

Any changes to the tune file since it started happening?
Does it start right back up or take a bit of cranking?
 

adamss24

Veteran Member
Joined
Nov 2, 2005
Location
Great Britain
TDI
audi a4 2.5 tdi 98 quatrro 6speed
I think youre chasing a whild goose! Most likely youre running out of fuel and engine shuts down... Had same on a c5 before...
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
[486] large low frequency movements vs small high frequency movements may have similar memetic energy, but will affect large vs small objects, especially rotating/hinged, differently. It depends on their own natural resonant frequency. I’ve taken a couple apart and dig into relay design and longevity. There isn’t anything electrical I can do to beef it up aside from directly changing it out with an aftermarket one.

Adamss24, I’m smoke free and my SOI tracks perfectly so there is plenty of case pressure in my VP37. I have dyno videos of rolling coal and my VP37 fueling R520’s to the tune of 195 whp. That said, I’m de-tuned from level of power and have a fresh fuel filter in. Please note than this present if I try to drive in 4th, on the freeway, on level road, with cruise control on, right at 4200 RPM.



Additionally, if I’m in gear and rolling, I just need to cycle the ignition and restore electrical power to turn the ECU back on and resume normal driving. No need to use the starter. Maybe I should inline a resistor and LED to the load side of the N75 to confirm it’s cutting out? Can’t really check with a DMM while driving on the road...
 
Last edited:

john.jackson9213

Top Post Dawg
Joined
Mar 14, 2007
Location
Miramar, Ca. (Think Top Gun)
TDI
1996 B4V
Michael,


Try a direct solder connection to the relay and any different mount. If issue persists at same RPM - IMHO, it is not the relay or the relay mount.


Other way to test is to go back to your original set up and see if that solves problem. But I would suspect it is not really an issue with the relay. Just doesn't make sense to me.


But then, what do I know.
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
the immediate restart makes me smell some software limiter coming into play, any codes?

I ignored my QA overvolt codes for quite a while before I realized that the pump volt map is scaled by a factor so the values should never actually be 5000 in the map, was getting a limp mode from that that needed a keycycle.
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
Try a direct solder connection to the relay and any different mount. If issue persists at same RPM - IMHO, it is not the relay or the relay mount.
I also like this idea

Use a toggle switch, some wire and crimp on male spade terminals to stuff in the relay connector, rather than the relay itself.
eliminate the variable all together
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
the immediate restart makes me smell some software limiter coming into play, any codes?

I ignored my QA overvolt codes for quite a while before I realized that the pump volt map is scaled by a factor so the values should never actually be 5000 in the map, was getting a limp mode from that that needed a keycycle.
There is no immediate restart. I can immediately restart it. It’s behaves identically, sans interior/exterior lights, as if I turned off the ignition. As soon as it’s “on,” doesn’t need “start,” all computer controlled equipment functions again.

If I rapidly flip the key off-on, I will generate a code. If I leave the key off for a couple of seconds before turning back to on, it’s fine.

Never once had a code generate due to this.

Fuel map doesn’t even touch 4V iirc, MAP ‘linearization’ goes up to stock 4.75V, SOI dynamic range is under 18°, N75 is 25%-80%. MAF has easily been at its 1275 mg/str limit. Remember, this doesn’t occur *just* at WOT, but if I try and use cruise control at that engine speed on level ground.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Michael,


Try a direct solder connection to the relay and any different mount. If issue persists at same RPM - IMHO, it is not the relay or the relay mount.


Other way to test is to go back to your original set up and see if that solves problem. But I would suspect it is not really an issue with the relay. Just doesn't make sense to me.


But then, what do I know.
I may need to remove the fluidampr for another reason, but this is a decent idea. The relay doesn’t slide out of its socket so I know that socket/mount is fine. Your idea is not far off from my indicator LED idea. Don’t have any spades immediately on hand, but I have a nearby shop with some.

Maybe I have a wire break to the ECU loom due to flashing it so many times? I’ve dealt with data/power cables at my job for the past decade and know how to treat a cable. It doesn’t seem like 100-ish flexions is enough...
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
There is no immediate restart. I can immediately restart it.
nah, I meant the restart isn't involving a lot of cranking as it would if it were air intrusion, a stuck oil pressure regulator valve, or many other mechanical issues.
Meaning, it does indeed feel like an electrical issue from all the way over here.

It's times like this that I think to myself over and over about sticking a 1.6TD pump on my car rather than figuring out why the thing is acting strange. So far I've always found the trouble eventually, though.
Stick it out brother, the light of a correct diagnosis will shine down upon you at some point! :p
 
Last edited:

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Alright, so while doing some EGT vs IQ vs RPM logging, I stumbled across another clue and relay 109 is no longer being looked as as the one to blame. I've encountered the issue so much that I've gotten used to immediately flipped the ignition off after examining my surroundings and then back to power after a few seconds. This time I was logging. Well, my VCDS log data continued to capture data before I flipped the key off, thus the ECU still had power.

I don't have previous logs to look at in regards to this due to my own negligence that lead to overwriting a large collection of data while reinstalling an OS on my other laptop. Reasoning points be to the injection pump. Visual memory points to the fuel cut off solenoid as a replacement part on a number of websites over the years. That or the is a break in one of the wiring harnesses. I'll be inspecting the latter shortly.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
The center of the solenoid wanted to spin instead of loosen. Yay belleville spring washer? Pulled one spare from one of the two donor pumps I have lying around. I'll test drive it tomorrow night so see if it is fixed. Nothing seemed off with the wires.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Kinda have experienced a change. The cut out will occur earlier in lower gears, but only with higher loads. I can do cruise control now. WOT or near WOT in 2nd and third will cause the cut outs. However, now I still have throttle control. The accelerator is no longer dead requiring a power cycle of the ignition.

I have the replacement fuel shut off solenoid in, and I have remounted the fuse panel, basically back into it's stock position; it was facing straight down instead of ~30° down. I even was resting my finger on relay 109 for a couple of 2nd gear pulls.

I only have recent logs that have peaks at 4.54 V, 4.52 V for the pump/QA voltage. I swore my ECU reported a maximum voltage of 4.7V, but it's a bit late and I need to turn in. I decided to look at my logs while writing this post, fwiw.

How's this for some thought? I know the ECU sends a crude, modulated sine wave to the QA to monitor its position. I believe I posted gifs of it at some point since I looked at it through my oscilloscope. Yep, I did.

It is possible that the physical position is monitored by nature of the 10 kHz wave. I think I'll do a hammer mod in the near future to see if I can bring the voltage down and re-run to see if this is still occuring. It is weird (that is that this doesn't completely make sense) though, since 3rd/4th gear is a greater load, it wouldn't make sense to be able to load the pump to >4.5V at 3500 RPM in 3rd gear, when 4th gear will only go >4.5V at or above 4000 RPM. Also, despite not having fuel logs from the Colt vs OE/VW camshaft testing, there were many people who can vouch for the fact that those 12 runs on the dyno never resulted in a cutout like this.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
https://forums.tdiclub.com/showthread.php?p=4432860#post4432860

Last record of hammer modding. Pump showed the same now so I guess I haven’t touched it. I don’t have a precise before-after, but I dropped the pump voltage down to ~1.33V and adapted the IQ to its highest. Since this correlates directly with the pump voltage map, I checked the records I have that mostly closely match what I’ve logged with what I run and added a linearly interpolated one out to 5250 and some slightly different ranges.

Didn’t feel any harm in doing that because the pedal responsiveness after a cut-off seemed to strongly correlate to speed, thus I’ve had no change in symptoms.

The extra ~0.35V, if the QA is the issue, of dynamic range should help identify the problem, hopefully. I’ll correlate that post with my email correspondence with Malone to see if things jive properly because I did bring it up to him.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Solved: Hitting the upper limit on pump/QA voltage. I may hammer mod it lower if I can retrieve some old logs via sent emails.

I forgot a significant change: I'm running Diesel HPR, not normal D2. If memory is correct, it has ~5% energy per mass or volume in comparison to D2.

IQ is already at "50 mg/str" at voltages as low as 4.26V. How exactly do I use driver's wish, smoke map, or the torque limiter (first choice, methinks) to ensure I don't hit this limit?

I haven't finished going through ECU Connections to up this limit, though I'd keep it to 4.75 imho.
 

bhodgkiss

Veteran Member
Joined
Sep 11, 2006
Location
Banbury, UK
TDI
AFN Passat Wagon
If you're genuinely maxing out the pump voltage, then you probably need slightly bigger nozzles for your power goal?
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Fun part is that I hit the power goal once, but not anymore. Mostly concerned about turning down the fuel output to avoid the pump issue so I can do WOT pulls, diagnostics, and curiosity testing. From there I’ll change out the pump to get better pressure for the nozzles.
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
yeah, I had similar troubles when I used the wrong factor for the pump voltage map
it should not be 5000 in the binary, instead there's a factor of 1.221, so you should only have the map going up around 4130 or so to get 4.8ishV max
The whole time I was doggedly ignoring a code that kept setting for QA voltage.

Careful going bigger than a 11mm pump on r520s, any 12mm stuff does not survive without very large nozzles

ETA: oh, and sometimes logs don't show actual peak figures, as the datalog out of the OBD port is glacially slow in refresh rate, where the computer will internally see the full request for 4.9v near instantly, the vag-com is a second or more behind.
 
Last edited:

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Oh, I’m aware of how slow a 3 sample/sec sample rate of VCDS actually is. I need to have some board designed and fabricated for plugging into a $10 DAQ that I can crunch directly on a calculator with QWERTY and a decently functional CAS. I think FUB identified the QA voltage output being a differential signal, and I need to look into that myself.

When I go 12 mm, I’ll still have my 10 mm around. Plus, I won’t be using a DE143 camplate, but plan on sticking with the DE110. Even though it’d be less output, having that pre-fill ring milled into the plunger may help with longevity by lowering the pressure gradient. That’s for another topic though...

Back on topic, my factors are correct and nothing is off on the pump voltage map. However, the tune is capable of requesting enough fuel to require voltages off of its scale. That map though just serves to convert the voltage to an IQ number, from what I’ve read. Do I use the limiters to reduce this fueling, or is there a different map that I should adjust? I know that question is beyond the scope of this portion of the community in general, but I know some here have some pointers on figuring it out.

Yes, I could contact my tuner. However, I view this as “trivial” because I’m trying to figure this out for my own understanding. I didn’t pay my tuner to teach me, I paid him to tune the car.
 

cruiserboy

Well-known member
Joined
Feb 23, 2013
Location
Paris, France
TDI
´97 Golf MK3 TDI AFN
There is a 4013 value not so far from the pump voltage map. Increase the value to 4095 or so and it may solve your issue.

BTW, how come your tuner was not aware of this?! And also for the measuring groups... it’s known since decades that you can ask for whatever channel you want in the groups
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
My tuner was aware of my stall issue, but it was happening very sporadically under greater load cause by basically doing WOT around highly banked turns. I’ve changed fuel since then and have not asked or needed tube updates for about 2.5 years. Most of that time I was dealing with Life™️ and a turbo with a non-functional vane system. Also, before the turbo was caput, I was able to do 3rd gear, WOT, 1500-5000 RPM pulls without incident. New fuel has about 5% less energy density so my fueling requirement has gone up.

The EDC15 ECU is well documented. The MSA15 series shares similarities with its older brother, but they are not the same. Also, even though the info is around, you have to know where to look and what specific search terms to use to find it.

Is that value is decimal or hex? Also, I find that value appropriately close to the map, if it is decimal.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Glad you found the problem and solved it. In the early days this phenomenon was called Warp Field Collapse. :)
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Found the value before the map on the ECU, thank you. There is an odd habit in the tuning community to not specify the base of a number or always assume it is base 10. Everywhere else it is common practice to notate hex or decimal or, since this is the next level up from binary, assume hex.

Anyhow, the problem is still not solved. 3rd gear, 4200-4500 RPM presents the problem easily, with the inability to engage cruise control. CC is ECU only. Need another SO-8 packaged 24C memory chip on my spare ECU to see if I have a hardware problem in my original ECU. My grounds are cleaned. Need to attach a DMM/‘scope and see if I’m still experiencing noise between ground and the chassis.

I have some mean harmonics that sounds like a part of 2 body panels in the engine bay vibrating against each other. Need to check the runout of the fluidamper as I think the face of the crank sprocket isn’t 90° to the axis of the crank. Things became more consistent but problematic after it was installed.

I’m switching this back to a [N]VH issue with possible electrical (microphonic/piezo) involvement
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Yeah, just over three weeks of looking at this problem specifically and I've been able to identify that my cruise control issue is hardware related, possibly a broken wire, and that the pump voltage map only goes up to 4.58 V. Unfortunately, the problems are interrelated, it seems since CC doesn't like to engage sometimes with before "warp field collapse."

Unfortunately I had data loss so I lost most of my logs. Some surviving ones via emails indicate that I touched 4.5 V, 4.52 V, and once or twice, 4.54 V. Change to a fuel with 5% less energy per unit of mass, and then the question is if I exceeded a physical limitation or software one. My logic is this:

I need to do a function check on the QA in the pump, and similarly, if I'm extending the range of the pump voltage map.

"While I'm there" I might as well have an pump IQ more closely correspond to what I'm actually injecting. Since the smoke map(s), boost map(s), driver request map(s), torque limiter, and N75 maps(s) are all interrelated. If the maps don't cover the appropriate range(s) and WFC still occurs, then I have a hardware problem.

That hardware problem can be as simple as a broken wire, which is very hard to find. Or it can be my QA being lightly gummed up or restricted, but the MSA15.5 software isn't known for throwing codes for it like the EDC15/16 software is. It can be R&R's/cleaned in a about an hour.

I'll approach the problem with replacing just the pump voltage map. If that doesn't fix it, I've already adjusted [nearly] every other map I listed, which includes the additional MAF extension from 1275 to 1785 (255*5 vs 255*7) which covers more air than my turbo can output safely. If the problem is then fixed and flaky CC remains, I'll move back to my normal thread with hopefully a summary and condensed post-operation analysis.
 

[486]

Top Post Dawg
Joined
Mar 1, 2014
Location
MN
TDI
02 golf ALH
"While I'm there" I might as well have an pump IQ more closely correspond to what I'm actually injecting. Since the smoke map(s), boost map(s), driver request map(s), torque limiter, and N75 maps(s) are all interrelated. If the maps don't cover the appropriate range(s) and WFC still occurs, then I have a hardware problem.
I just kinda multiplied the x axis by 3, so now it goes up to 153 instead of 51
along with the DW, smoke maps, torque limit... Really only the ones that show up in vagedcsuite (and pump voltage which doesn't come up as labeled, of course) Could have probably left the start map alone, but I like how authoritatively it starts with the big fuel system parts. :D

it runs smoother, sure doesn't display more than 51 on VCDS but looking at the pump voltage and MAF actual you can tell it's going up in that corner of the map regardless.

Speaking of the pump voltage map, I've been holding off on refining my tune until I can get a function generator and a pump bench together to make up a calibrated pump voltage map. Run it at X RPM and give it Y volts for Z time and hey, that's 1000 strokes, can calculate mg/str from that easily enough. In reality it just means I haven't touched the tune since the idea came to mind. :c

Anyways, back to the topic at hand, your idea on the cruise control being a wiring fault, maybe the other fault is in the brake light switch or throttle pedal circuits, putting the car in limp mode as well as disabling cruise? Sometimes that stuff doesn't set reliable codes.
 

Matt-98AHU

Loose Nut Behind the Wheel Vendor
Joined
Apr 23, 2006
Location
Gresham, OR
TDI
2001 Golf TDI, 2005 Passat wagon, 2004 Touareg V10.
Another thing to think about:

Whenever you turn the key on, the QA does a "sweep" of its min and max positions. The min and max voltage readings of the QA position sensor are registered/viewable in measured value block 19.

If the max position during the sweep is less than the voltage you think it should be getting as requested by the tune, maybe that's the ECU limiting its movement as it previously recorded the max position as being lower than the voltage you're trying to specify.

So, for example, if the max position during the sweep only registers 4.5xx V, but you're trying to get it to 4.7 V by tune, I don't think you'll be able to get there.

The computer may also recalibrate its "mg/st" quantity readout based on the voltage registered during the key on sweep. So, whatever the max you can request as far as mg/st in the tune (60 mg/st?) may just be assigned to the max voltage recorded during the initial key on sweep.

In other words, without knowing what the computer thinks the max position of the QA is in volts, the max voltage reading during driving is kind of moot.
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
I’ll be using an WBO2 sensor for calibrating the map at some point. Have the hardware, just need the time.

Matt, yep, that’s exactly what I need to determine. However, the voltage map is only 40 mV off of the max voltage. Since they are so close, I need to change one (map in this case) to see what the behavior of the system is. You just happen to spell out what I am trying to discern ;)
 

Digital Corpus

Top Post Dawg
Joined
Mar 14, 2008
Location
Ontario, California
TDI
'97 B4 w/ 236K mi body, 46K mi soul
Btw, the CC issue is largely the clutch switch not being detected. Both brake switch are fine. It nearly always registers as open, where the pedal usually sits. This is regardless of which switch is there. I have continuity to pin 17 of the wiring harness and grounds are good. I have a ~4.8 mA drive current and 12 V like I should. The circuit looks the same on my DMM as our other B4 does. Comparing 3 ECU’s to each other, there doesn’t seem to be any fault unless it is in an ASIC, but I need a couple parts before I swap one in to test.
 

eddieleephd

Top Post Dawg
Joined
May 27, 2012
Location
Battle Ground, Wa
TDI
2002 jetta Wagon
If I rapidly flip the key off-on, I will generate a code. If I leave the key off for a couple of seconds before turning back to on, it’s fine.

Never once had a code generate due to this.
Now that's contradictory, what code does it throw when you immediately flip the key back on? Implausible 109, or something else? This is likely the only hint you will get.


As said before it may be limited fuel, or limited by ECU.


The symptoms oddly enough resembles the crank sensor failing from the cut out at higher temps and RPM's. Could be a wire is broken enough that when it gets hotter it just stops conducting, or a bad ground that is causing a loss of connection. I suspect a lot more things than the 109 because of when it happens

Somethings phishy about the situation and you accusing the 109, if it were the 109 RPM's would have nothing to do with it, only temperature and time would vary. Don't solder it in, pull the plug from the rack it's clipped into and let it hang. The only way a good "updated" relay should fail is if too much power is being drawn through it and it's fouling. It's a normal 4 contact relay with a 5th small signal wire connected, the updated ones were to withstand higher loads. Technically you could split the output and use 2 to increase life and reduce the load through the relays




This vibration and the modifications involved only intensified a situation.
 
Top