erratic n75

I was going to follow up on another thread that had the same faults as my ecu, but failed to subscribe, and can't find it. So I will start anew, his was solved anyway.
Open or short to grounds for egr, glow plug relay, n75, n108, n239, and fuel solenoid.
The other person's issue was resolved by tightening the fuel solenoid connection. I cleaned that connection and that code disappeared.

To summarize, I was losing boost, and after troubleshooting led to nothing definitive, I switch ecu's and had no issues during a 45 mile drive. However, I've heard the saying ecu's generally don't die, they are murdered so I have been searching for a cause. I have cleaned grounds and going to check continuity in related wiring to see if I can find something compromised.

I am going to have an ecu tuned tomorrow for it and hate to blindly put it in and always wonder if it will fry. I am going from 012 bd to 012cp and will post how that works as no one I asked seemed sure if they are compatible, both from '00 or '01 alh's. EGR, ASV, and possibly MAF will be deleted.

Now, to what I found to be strange. I ran n75 output test, with vacuum gauge set up, and the system worked with the engine off. Gauge bounced between 2 and 20 in., and actuator and vnt did it's thing(until vacuum depleted). Running the same test in basic settings with the engine running, the n75 did not cycle, but the signal spike was on the graph. Changing n75's makes no difference. It actually works stone cold, the turbo spool can be heard at idle when first fired. Once, after sitting over a week, it worked for about 10 miles.

My thought was maybe a diode in the n75 went bad and backfed the ecu as the signal pulsed off. Assuming there is a diode, can it be checked? Are there any other diodes that could cause an issue?

Any one with any thoughts or similar experience? Thanks.

I bought a golf from an auction and it had these same codes, the previous owner had replaced EVERYTHING. N75 was new, all vacuum lines were new, egr, injection pump etc... I could clear codes and unplug and then plug in ecu and would get maybe 5 miles before it wouldnt drive anymore. I found a thread on here where the guy had same codes and it was fixed with a ecu. I had an ecu on hand so I sent it out to get immo delete and havent looked back since. My friend has put about 10k on the car since and has not had one hiccup.

It would be unusual for a solenoid valve to have a diode included.
They work on AC or DC pretty much the same.
The electronics should all be in the ECM.
Now, a dead short could cause some harm upstream but wouldn't function.

flee said:

It would be unusual for a solenoid valve to have a diode included.
They work on AC or DC pretty much the same.
The electronics should all be in the ECM.
Now, a dead short could cause some harm upstream but wouldn't function.

Click to expand...

The diode could be to catch inductive kickback spikes when a solenoid is shut off.
The magnetic field collapses and causes the pulse.

Many years ago I saw a spike a couple hundred volts from a starter.
Used a scope to catch that one.
It was burning out the radio, alternator, and some bulbs.
Replaced the starter & the car was OK.

I always thought that was why the radio shut off while cranking the car.

A DC solenoid has EVERY reason to have a diode (reverse-biased) across the terminals, as it captures and shunts the spike from the collapse of the magnetic field.

An AC solenoid CANNOT have one for the obvious reason; those circuits need some other sort of protection. The voltage spike produced from field collapse is pretty impressive; if you have an oscilloscope and set it up to trigger on excursions above VCC+ you can easily get an image of it. The correct place for such a diode is AT THE COIL; you can place it at the switching transistor (e.g. power MOSFET) but that's really a "last ditch" sort of thing; DC solenoid coils should have it at the coil itself because you want it right at the inductor; closer is better.

For AC circuits the correct approach is to use an SCR or Triac and only shuts off at zero-crossing -- ergo, no spike. If you must use a pair of MOSFETs or similar for drive current so you can shut off mid-cycle then special considerations are required as the back-EMF HAS TO be handled since even a 12V coil can produce ~100V+ in back-emf, and if whatever is driving it can't handle that then it will be destroyed. The quicker the switching cutoff the more back-emf is produced, all things being equal. The usual answer for physical contacts that cannot be synchronized with zero-crossing (e.g. a relay) is a capacitor across the contacts which acts as a snubber and absorb most (but not all) of the back-EMF spike.

In the U.P. With no phone service, getting breakfast in Gladstone this morning. I will address this when we get back home in a few days. Good to hear the ECU solved the issue. Yes, voltage spikes from a bad diode near a coil is my concern. The n75 or maybe fuel solenoid? So a scope may catch that, had not occurred to me, good point, thanks. Had my ecu remapped with Maf, egr, asv, and immo deleted. From same year ALH so I will post how that comes out too

Update on ecu compatibility, the 012cp works in my car. I don't know if I can access a scope somewhere, but if I can, I will start another thread on checking the n75 diode with it. I did not find anything wrong in the wiring, so I took all my unused plugs, squirted dielectric grease in them and sealed them shut with electrical tape.

Fantastic thread! OP keep us posted on the cause and fix


Sent from my iPhone using Tapatalk

I haven't seen this one in a while, but sometimes there are weird error messages caused by a blown fuse #34. I realize this isn't the answer for this issue, but once-upon-a-time I pulled my hair out chasing what I thought was an N75 error but the fuse was the real culprit.

Glad this is of interest. Something else hopefully is useful to someone.

Some other info we dug up while on a web search, a website that claims to have ecu repair manuals. Short version, Google electronicrepairegypt, and it will take you to "Electronic Repair Company. A login is required to view the inventory. If I were to do that I would make a specific email for it. I did not bother because it does not look like there was much in the ecu that we would repair.
The circuit boards were laminated together, so no following paths by eye. There was a slight odor in one area, and the top of a processor had a change of finish on the top. Some was satin and some of it shined, indicating a probable change in temp. Not something I will ever fart with.

If anyone out there has info on "scoping" the n75 to check the diode, sure would like to here it. Meanwhile, I will try to find someone with a scope. It would be nice to know the cause.

I did delete the MAF, figuring one less common component. Haven't noticed that it has changed anything. Fuel economy will be a while before I can compare to anything.

I see no data here (yet) showing the actual presence of a diode in a N75 valve.
When I bench tested mine I paid no attention to the polarity of the 12VDC I fed it.

Hmmm, I noticed no difference in ohm readings regarding polarity either, though with pulse width modulated supply controlled by a ground, It seems a diode would exist to prevent the possibility of power returning via the supply wire back toward the ecu when the ground is interrupted.

It appears that this issue is not uncommon, several instances showed up during my search. It wouldn’t surprise me if someone has already gone down this path and figured it out.

I guess VW really dropped the ball by not specing a diode in every N75 valve.
Practically speaking, I have purchased and used literally hundreds of solenoid valves.
Many low voltage miniature valves like the N75 and many not so miniature, too.
There has never been a need for a diode, particularly when controlling them with PWM.

I assure you that back-EMF protection is present somewhere on that circuit. It takes only seconds with a Oscilloscope on an unprotected drive circuit for any sort of inductive load to understand WHY it has to be there. It may be at the ECU but as someone who designs and builds circuits like this for both fun and money, as well as someone who has worked on a hell of a lot of them from pinball machines to various industrial devices to diving rebreathers to cars and more you can bet your last nickel that it's present -- it's just a function of WHERE it is.

flee said:

It would be unusual for a solenoid valve to have a diode included.
They work on AC or DC pretty much the same.
The electronics should all be in the ECM.
Now, a dead short could cause some harm upstream but wouldn't function.

Click to expand...

Genesis, at least we agree that there is no diode in the valve, as I said before you found a podium.

Flee, are you positive there is not a diode in the valve, I am asking because I do not know, not to disagree.
I have a friend who designs and builds controls for automated packing lines. He thought there would be one on the circuit, and the most common spot would be at the coil.
Now if I can get someone with a scope and test both ecu's, it may show something conclusive. If the replacement ecu shows a diode and the previous one does not, then it would be in the ecu. If it shows none, I purchase a new n75 and try that.
At any rate, probably won't happen very quickly, I have house project I am determined to finish before Oct 1.

You guys really threw me when you started mentioning diodes. I don't really know, but I'm pretty sure it's a simple solenoid activated valve.

jmodge said:

Hmmm, I noticed no difference in ohm readings regarding polarity either, though with pulse width modulated supply controlled by a ground, It seems a diode would exist to prevent the possibility of power returning via the supply wire back toward the ecu when the ground is interrupted.

It appears that this issue is not uncommon, several instances showed up during my search. It wouldn’t surprise me if someone has already gone down this path and figured it out.

Click to expand...

I found the same ohm readings on an old N75.
The same reading both ways about 20.2 ohms.

The diode (alone) theoretically should be infinite one way and 0 ohm the other way.

Yes, my thoughts also. My thinking is that it may be in the ecu. That’s why I am going to run a scope on both ecu circuits. Just like to know instead of guess and wonder

jmodge said:

Flee, are you positive there is not a diode in the valve, I am asking because I do not know, not to disagree.
I have a friend who designs and builds controls for automated packing lines. He thought there would be one on the circuit, and the most common spot would be at the coil.
Now if I can get someone with a scope and test both ecu's, it may show something conclusive. If the replacement ecu shows a diode and the previous one does not, then it would be in the ecu. If it shows none, I purchase a new n75 and try that.
At any rate, probably won't happen very quickly, I have house project I am determined to finish before Oct 1.

Click to expand...

A diode on either terminal in series would stop current flow in only one direction.
A diode across both terminals would read less resistance in one direction than the other.
I think it would be hard to simply test the ECU to verify a diode's presence.

The usual back-EMF protection is to place the diode across the terminals reverse-biased. There is no flow of current during normal operation, but when the field collapses the EMF spike is in fact the opposite direction of the normal current flow, the diode conducts and shunts it harmlessly.

A meter MAY NOT detect this, depending on the resistance of the coil and voltage used by the meter for resistance measurement, as the usual silicon diode has an 0.7v voltage drop. Depending on the coil resistance the drop across the junction may not be enough to bias the diode on (irrespective of which direction you connect the leads.)

I hope this thread doesn't turn out like Hiraldo Rivera's opening of Al Capone's vault, but I have not had much luck locating someone with a scope to decipher this. Busy on my daughter's place and getting ready for hunting season. Haven't given up, maybe someone coming to Kirk's GTG will have some rescources to look a little deeper.
I do have some electrical issue though (seems eleven out of every ten of these cars do!). The other evening my instrument and dash lights would flicker. Did not seem to be anything else, headlights were fine. Wiggling the dimmer and headlight switch made no difference. I have seen others that found switches, grounds, voltage regulator, and even a sloppy alt bearing to be a culprit.
Need to check into that, but I am off to help set up the hunting camp tent and beat the crap out of my quad.