breaking up at conisitan rpm under heavy load

[savagerocco]

Ok, I have been fighting this problem for a while and wonder if there may be more to it. Been dealig with Hayden having a similar problem with a bmw 325 and breaking up as well. He is using ls2 truck coils and I am using vag coils. I haven't finished the FET gate driver for my setup, but I want to bounce this off the group.
While the problem does seem to get better with load, ie it will pull clean with about 1/2 throttle but fulll throttle it cuts out almost exactly between 5500 and then comes back around 6500. I find this strange. I am hoping it is some driver weirdness that the FET mod will take care of, but I am skeptical it will.
I have a log and the msq for you to look at. I haven't really noticed anything amiss inthe log, and can only feel and hear the cylinder dropping out.

[Matt Cramer]

Have you tried a tighter plug gap?

[savagerocco]

I have, I took it down to around .020". I have also tried different plugs designs as well. I was using an NGk bkr8s gapped at .020" I nop have the bosch FDR05 tripple electrode plug. The tripple electrode plug is factory gapped and really cleaned up idle but really no change in the break up above 5k .

[billr]

To me, that VE map looks very "unusual", how are you certain this is an ignition issue, not fueling?

[savagerocco]

The maps on this car have always been fairly flat. The afr's look great on every data log. I have made posts about it before and it seems that with afr included in calculations the map can be quite flat.
I'm about 90% sure it is an ignition issue and it is consistent starting out at 5500 and cleaning back up at 6500. I might have to try getting a video and see if you can hear to tone change when the cylinder drops out.

[savagerocco]

Ok, I tried to install the FET drivers last night using led outputs. Ran into some issues. Space being one of them. There is a lot going on in that case. The second issue is knock module. It wants to use sp5. I also had some issue with baro correction pin usage. I disabled the knock module and turned off baro but wasn't able to get it to run and being 1am decided to throw in the towel and put it back on the msx outputs. I remember there being a pinout for the headers on the ms3 processor. I want to run the spark outs that normally go to the msx card out my 3.57 db15 and make the FET board external. Other than the list on the ms3 board itself where else Is documentation on pin usage?

[billr]

I'm confused, those are well-labeled for the MS3 JP2. You don't want to connect to that header (or the MS3X header or DB37)?

[savagerocco]

Was trying to bypass the msx card. I will try it with the msx outputs and see how it goes.

[billr]

If you take the spark outputs off of the MS3 JP2 header, you don't have to involve the MS3X...

Edit: I am poor with cut/paste or screen grabs, but are you familiar with page 194 of the manual below?

http://www.msextra.com/doc/pdf/MS3XV30_Hardware-1.3.pdf

[savagerocco]

I built a new driver this morning and installed it on the msx outputs and wired it into my coils. While it runs now it still cuts out around 5500 and cleans back up around 6500. I am ordering new 2 new coils and will swap them out 2 at a time until I find something.

[savagerocco]

Ok, got two new coils 06E905115e R8 coils just in case anyone asks, and swapped my coils out 2 at a time. Test drove after swaps and still have drop out.
So, new plugs, new coils, new FET tc4427 driver board. same break up issue. I guess it could be fuel. Unless anybody has made misfire detection, next step is egt probes on each cylinder to find the one dropping out.

[billr]

Do you have a sensor for logging fuel-pressure? I would try that first, a lot easier than adding the EGT stuff. Have you flowed the injectors recently, even a "quickie" test to see all are about the same and have similar spray patterns?

Edit: the VE map not only looks "flat", but seems low. I would expect any modern engine (like not a flat-head!) to have a VE better than ~70 in its power band. With such a short req_fuel value, I hope the dead-time has been carefully checked, that seems kind of low also. What kind of injectors, hi or lo-Z? If low, what kind of drivers?

[savagerocco]

I did flow checks in test mode last year. I was chasing an issue with detonation. Injector flow is pretty consistant. I have a pressure sender for fuel I just need to wire it in. I have a new in the last couple of years 044 Bosch pump in the car. I want to go with the egt per cylinder anyways. I have the probes and a tubular manifold already for install. Just need to make my own can interface or just get the diy one.

[billr]

That was last year... did you have this same problem last year?

[savagerocco]

I could not push the car hard due to detonation issues. So I'm not sure if this was an existing problem or a new one.

[savagerocco]

Also, why would it only start at the same rpm 5500 and go away at 6500? I can literally hear the change in tone of the motor as it drops a cylinder and then comes back.

[billr]

I could make up several different guesses as to that. Since you are headed toward installing the EGT probes anyway, get that done so that we can see if this is a problem on one specific cylinder or is random amongst all of them.

[savagerocco]

That was my concern, I want to narrow the scope down. It doesn't seem to change I am pretty sure it is only one cylinder having issues.