ECU issue, Injectors lock open and hydralock motor

[lugnuts]

I'm having two major problems with an assembled V3 controller. The ecu "locks-up", loses COMM, and opens the injectors and fuel pump relay, causing the engine to hydralock (very quickly with 1000cc injectors!!!)

The second time this happened, the analog inputs (EGT, Misc 1 (was using for external wideband) and the Launch Control stuck ON)
- However the external MAP sensor is working, car still runs.
- What component has failed? Does this help diagnose the original problem?

Also, I was using my external wideband because the vems wideband was reading very erratically at medium - high RPM. I'd very much appreciate any insight into these problems.


Wiki Link:
http://www.vems.hu/wiki/index.php?action=find&find=MembersPage%2FKevinBlack%2FVWGTIAnother

[gunni]

Are you using a USB adapter to the vems??

or did this just happen out of the blue.

[lugnuts]

No USB adapter, just a serial connection. I was using Megatune and it happened right after a "Burn to ECU" and/or the first ignition OFF-ON cycle.   
Thanks for the reply. I'm hoping to get an absolute answer to this question so my customers do not lose confidence in this ecu.

[Denmark]

Are you running a duelcoil iac valve?

if so, then the vems automatic pick the channel after the one you use, meaning, if you pick channel 7, you will also use channel 0, were you probbely have a injector connected to?

I have had this before, and it will destroy the fets!!


/Skassa

[lugnuts]

The Dual Solenoid setting is "DISABLED" so it looks like I am OK there.
 
If Ch0 was stuck on, it would be possible to fill all cylinders with fuel (common plenum intake manifold) , however it would not explain the ecu losing COMM at the same moment.

Thanks very much for bringing it to my attention, I have added a note in my custom .ini file so this will be checked by others.

[rob@vems.co.uk]

Did you have the laptop connected to the car when you cycled the power?  Was the laptop connected to the car via an inverter or somesuch power supply?

It may be related to a ground-loop problem with the laptop and ECU.

Try and replicate the problem (remove your injector and ignition fuses) by keeping the laptop connected to the ecu during a power cycle (it may take several goes), and also without it connected.

[lugnuts]

Yes the laptop was connected.
Yes I was using an inverter.

I will be able to work on the car in 2 days. I can try to replicate the problem. I'm not sure how long it will take. I was using the power inverter for probably 50%-60% of the time I was working on the car.

How hard do you think it will be to replace the mcp3208 chip? (I do not have much exp. with small soldering, however on my other project I successfully soldered in the EGT chip)

Thanks, Kevin.

[rob@vems.co.uk]

This suggests a ground problem, this has happened in the past with dnb's car which has a fibreglass body and notorious ground issues as a result.

Jorgen says that the death of the mcp3208 chip is a further sign that there are grounding problems as even a tiny reverse voltage  (-0.3v) will kill a semiconductor, if its been connected to a ground who's voltage has gone high then there has been the potential to cause a reverse voltage. Its very unusual for this chip to die.

As for replacement if you've soldered on an EGT chip then you will be fine for installation - the issue is with the removal though.
When I replace a p259 chip  I start with a pair of microfine cutters and snip each chip pin so that I can remove the chip's body.  Next with an ultra fine soldering chip I heat up and solder suck each pad, and then finally use de-soldering braid to mop-up any old solder on each pad.

For soldering I take my normal everyday solder, and squash it flat using pliers, then cut along the flat to make a very thin piece of solder, starting soldering from opposite corners I then solder each leg in turn.

[lugnuts]

OK so are we talking about a broken ground 5 wire? (all 14 awg going to 10 awg to engine block)
Or, the power inverter? (to be clear it was an AC/DC Type.... what about DC cig. adapters are they OK?)
Or, the Air/Fuel meter? (analog gnd connected to chassis ground, mainly to not upset the meticulously connected gnd 5 / sensor gnd wires)

Thanks for the tips on desoldering/soldering.

[Jorgen]

Hi,

We aren't necessarily talking about broken main ground (the five wires you mention), the MCP3208 chip can brake if a switch is connected to it that will short an input to a generic ground point in the car. If you look at the different ground points in a car they can have significally different potentials. This means that if for example a launch control switch is used it MUST be connected so that it shorts to the ECU's sensor ground.

External sensors referenced to generic ground can also cause similar problems.

It is impossible to say if a certain DC/DC converter or a certain DC/AC converter will work, some of them works well and some doesn't. If the ECU is grounded in the chassi of the car it is far more likely that any of them would work. That would at least limit the risk of a ground loop.

If you ground the DC/DC or the DC/AC in the same chassi part as the ECU you are far more likely to make it work. Otherwise you will have to hope that the converter fully isolates the laptop from the cars supply.

Where does the air/fuel meter come into the picture?

Jörgen Karlsson
Gothenburg, Sweden.

[lugnuts]

The Launch Control switch is grounded to the car thru the e-brake harness, definitely not to the sensor ground.

The Autronic air/fuel meter has a 0-5v output with a sensor wire and a ground wire. I ran this ground to the chassis, normally I do run this to the ecu sensor ground but I admit I was in a hurry.

If this can not work, then how would one use other brands of wideband such as the 6 wire Innovate unit which only runs its grounds to chassis?

(I'd like to get the VEMS wideband working correctly if someone has any ideas about that, at mid-high rpm it is so spiky it is un-useable.)

The ecu is grounded to the engine block. 

[Jorgen]

It is very important for the Autronic WBo2 units to have their reference ground connected to the sensor ground.

If you read the innovate manual you also see that they suggest that you connect their WBo2 units to the same ground as the ECU. Or maybe that was something I got through their tech support when setting up one of their units to work with an Autronic SMC a few years ago.

Spikes in the WBo2 is usually a result of exhaust leaks or missfires. Do you have a log? Are you running an open header? Does the WBo2 work well at low-mid rpm? From who did you buy the ECU? Is the ECU factory assembled?

As the Autronic WBo2 has an output filter that can be set up in a way that would hide this. It would of course indicate the wrong Lambda on the voltage output when doing this.

The combination of having a few of the accessories connected to chassi ground while the ECU is connected to engine ground is very likely to cause serious problems. You need to choose a ground and then stick with it.

Jörgen

[lugnuts]

I agree about the Autronic sensor ground, I have been working with Autronic since 2002. Innovate changes tgheir minds about their grounds depending on the day of the week in my experience.

The exhaust system is fine, I removed the LSU4 and installed my sensor in the same bung, and the signal was fine.

The Autronic wbo2 settings are all default with the exception of 0-5v being set-up for 10-15 AFR. (Function 9 and 10). When used with other brands of ECU it is true you do see small fluctuations in the log but that is mainly because the visual display is dampened so it can be read. 

The signal I logged into "Misc 1" in MT is much,much smoother than the "lambda" from the LSU was.
The lambda logs made it impossible to tune the engine.

The ECU was bought assembled, serial number #1724.

What is your definition of a few accessories? Anything that connects to the ECU right?
Right now the FJO Injector Driver (was not sure if wise to gnd with ecu) and the Autronic wbo2(time, and did not want to upset the already joined sensor connections) are chassis grounded.
And can you please explain (I am obviously not an EE) how it can cause damage to the ECU? 

Thank you for the reply, 
Kevin

[lugnuts]

I have an update about the car, it is running well and made good numbers on the dyno.

The wideband started working well, I don't know how or when but it seem OK now and I tuned the engine using it.

I still do not know 100% what caused the mcp3208 chip to fail.  I did not use the external wideband or the DC-AC power inverter since the damage (out of paranoia, I really would like to know if this is going to be a problem for my customers) I have not replaced the mcp3208 yet.

Dynos:
http://www.vems.hu/wiki/index.php?page=MembersPage%2FKevinBlack%2FVWGTIAnotherDynos

Thank you to all of the forum members for your help.

[rob@vems.co.uk]

Don't come on here posting ~600bhp dyno charts...

...without  pictures!