injector drivers on 3x board keep dieing.

[shaodome]

So I am having a random problem with the 3X board and is injector drivers. This has now happened on 3 different cars in the same manner seemingly totally a random.

All of the injectors have been either the newer bosch EV14 style (9.5ohm) or the Siemens Deka units which ohm out a little higher(11.2-11.5). All high impedance. On my personal car I ran the bosch EV14 injectors in batch fire with MS2 via the v3 injector drivers as high imp with no problems for over a year. Switched to MS3x and the car drove flawlessly for 6+ mo every day, then one day driving home and injector channel got stuck open (channel C) (verified with LED tester on the side of the highway). That 3x board was from one of the original group of 50, so I just replaced it and called it a manufacturing defect and car had been fine since.

After several months of running it happened on my car again, this time channel A

Another customer's car of mine with a brand new 3+3x from DIY (prebuilt) and it worked for a few hours and then died on the dyno. channel D.

3 different sets of injectors 3-4 different boards, two cars

I did some poking around on the 3x board last night and I found the fault is with the transistor for the inj channel. The signal gets to the transistor gate just fine and the source ground is intact. If I lift the gate side of the transistor from the 3x pad so it is just connected to the source and the drain (using JimStimX) the LED remains lit. If I remove the ground (source) from the transistor, the led goes out. It is like the gate of the transistor is stuck in the ON state.

[shaodome]

using ohm's law....and the formula I=E/R (13.5v & 9.5ohm) I only come up with a current 1.42 amps, which is way below the limit of the VND5N07's 5 amp max.

[shaodome]

bump?

[jsmcortina]

I've passed this on to Bruce Bowling for comment.

James

[shaodome]

I've passed this on to Bruce Bowling for comment.

James

Thanks, as always, James. Would this thread (or any future 3X hardware questions) be better served on the MSefi forum as opposed to here?

[jsmcortina]

Would this thread (or any future 3X hardware questions) be better served on the MSefi forum as opposed to here?

I believe you'll probably get banned if you mention MS3 over on MSEFI.

James

[shaodome]

Would this thread (or any future 3X hardware questions) be better served on the MSefi forum as opposed to here?

I believe you'll probably get banned if you mention MS3 over on MSEFI.

James

Oh! Good to know Looking forward to see what Bruce thinks about this....

[wes kiser]

Just wanted to comment I have used the MS3x on my personal vehicle for many thousands of miles with no issue (which with 12 ohm injectors is admittedly not surprising).

Perhaps a more useful comment is I have a microsquirt module based adapter board, that must occaisionally support a pair of low z injectors per driver(which uses the same injector drivers). The way I handle this is with a series 3.3 ohm resistor, which allows right at 3.5 amps per channel when used in series with a pair of low-z injectors. I have several of these boards with quite a few miles as well.

Hopefully you find the root cause/resolution. Two different boards likely can't be attributed

[subwoofer]

From the schematics it looks to me as there is no overvoltage protection on the injector driver FETs. A voltage spike may or may not occur, but if it does there is nothing on the MS3X to arrest it. A set of Tranzorbs (BZX series or equivalent) would add peace of mind. I favour Tranzorbs over regular Zeners for this purpose, I have killed too many microphone preamp circuits that I had attempted to protect with regular zeners.

[Peter Florance]

From the schematics it looks to me as there is no overvoltage protection on the injector driver FETs. A voltage spike may or may not occur, but if it does there is nothing on the MS3X to arrest it. A set of Tranzorbs (BZX series or equivalent) would add peace of mind. I favour Tranzorbs over regular Zeners for this purpose, I have killed too many microphone preamp circuits that I had attempted to protect with regular zeners.

I believe the FET's are have built in clamping.

[Peter Florance]

I'm wondering if it is getting blown from the gate side.

Maybe some 5.6v zener diodes from gate to source? Cathode to gate.

[Peter Florance]

Just wanted to comment I have used the MS3x on my personal vehicle for many thousands of miles with no issue (which with 12 ohm injectors is admittedly not surprising).

Perhaps a more useful comment is I have a microsquirt module based adapter board, that must occaisionally support a pair of low z injectors per driver(which uses the same injector drivers). The way I handle this is with a series 3.3 ohm resistor, which allows right at 3.5 amps per channel when used in series with a pair of low-z injectors. I have several of these boards with quite a few miles as well.

Hopefully you find the root cause/resolution. Two different boards likely can't be attributed

I'm running 8 low z injectors with 3.3 resistors added to them using MS3X. It's a race car so it doesn't rack up large miles like a daily driver. It does get driven briskly for 20-30 minutes at a time.

[subwoofer]

I believe the FET's are have built in clamping.

Looks like you are right. I found the data sheet for VND5N07 and there is an internal clamp for V(DS). I guess that means that the problem is either that that clamp is for some reason ineffective, or that the problem is on the gate side as you suggest. But if the problem is on the gate side, the problem should manifest itself with all kinds of injectors, not just these particular ones?

[elaw]

Or... that the dissipation in the FETs when they're on, or the energy from the spikes, or both, is causing them to overheat.

Have you tried measuring the temperature of the drivers? It's not the easiest thing to do, but it might yield some useful info.

[subwoofer]

According to the data sheet, the FETs have overheating protection built in, so in that case the problem should not be persistent. They should also have failed open if the protection circuitry kicked in.

[racingmini_mtl]

If you look at the data sheet, there is protection on the gate side also. So the input and output sides are protected and there is thermal and over current protection. So it should be quite difficult to make them fail. And a single high Z injector should not be a big load unless these particular injectors have an unusually high inductance which would generate a lot of flyback energy.

Jean

[shaodome]

If you look at the data sheet, there is protection on the gate side also. So the input and output sides are protected and there is thermal and over current protection. So it should be quite difficult to make them fail. And a single high Z injector should not be a big load unless these particular injectors have an unusually high inductance which would generate a lot of flyback energy.

Jean

This has happened with 2 different types of injectors now (both newer designs)

One was this style (Siemens):

And This one (Bosch EV##):

When the last driver failed it died a little different than the other 3. I was driving along and the car dropped a cylinder. Then it came back a few seconds later...that happened a few times until I got about 1/2 mile from my house. That last 1/2 mile the car ran only on 5 cyl and it never returned. I figured from the intermittence that it was just a loose connection.

When I put that 3x card on the JimstimX, I could see channel A roughly 1/2 way lit. When I pulsed it using output test mode the led did get brighter, but clearly some current was being allowed to pass through that transistor.

[subwoofer]

Dropping a cylinder intermittently is consistent with the protection circuitry kicking in. I'll read the data sheet properly later tonight, I need to understand how the overvoltage circuit works.

[wes kiser]

My RX7 has the EXACT same Siemens units (4 of them) with MS3x. I have logged approximately 7k miles with the MS3x and these injectors (and 2 Megameets )

[shaodome]

Dropping a cylinder intermittently is consistent with the protection circuitry kicking in. I'll read the data sheet properly later tonight, I need to understand how the overvoltage circuit works.

that's what i figured, but then it just died.