throttle bodies spitting back

[se7ensport]

Hi folks

Throttle bodies are spitting back during tickover and low revs, when at full chat there isn't a problem and lambda seems about right, haven't had this issue before when the engine was on carbs so not sure where to start; valves? timing? other?

30 sec Video uploaded on youtube:

http://www.youtube.com/watch?v=O5aMb1XYrQE


seems to be mostly cylinder 2 with 3 and 4

All suggestions welcome

Alex

[mattias]

Vemslog  and link to project page (on this forum?) with full engine/setup specs would  be a good start.

[se7ensport]

Vemslog  and link to project page (on this forum?) with full engine/setup specs would  be a good start.

Fair comment, no project page unfortunately, setup is stage 3 head with fast road cam, pinto 205 block, gsxr750 tb's and of course vems.  Will take a vemslog later, was initially looking for general suggestions, I'm planning on retarding the timing a few degrees to see if that makes an initial improvement.

[mattias]

Always datalog, once you connect the computer press "Alt-L" (make it a reflex movement).

I can only guess that it's a tuning issue, maybe it's a bit lean. If you're running Alpha-N it can be tough to get a grip on how to make it run right at first.
Having cam sync with injector angle curve would help some symptoms, but we know way too little about your setup, etc..

[se7ensport]

link to vemslog attached http://www.box.net/shared/axzj7cztnx

tried at different rpms, around 12-1300 seemed to increase the occurance, but not by a lot.

edit: did wonder about a mechanical issue as its predominatly one cylinder, car hasn't been run much over the last year, just starting it up every couple of months.

[mattias]

The map doesn't look "tuned" to me, and too little resolution on the load axis (TPS) at low openings, seems like you're wasting the bottom 0% line. A simple "VE tune by statistics" on your log  will tell you that (display changes in % to easily see where the values are raised), or enable the "live VE analyzer" and see what it tries to suggest to you.

The rest of your config looks ok, here are a few points :
Set injector PWM % to 100, just to be sure.
Injector angle curve, load defaults for it (720 degrees).
I suggest using fuel only rev limiter, ignition cut is for people not afraid to loose their engine.
In "ECU calibrations" you should change to have "table kPa unit = 2", it will be the standard in later firmwares, you need to look over your Y-axis on all tables after you're done.

Press the "validate" button next to the log button to verify the sanity of the config, some do not matter for you, others do.

[z0tya]

Mattias: Why fuel cut is better than ign cut? I have read the fuel cut cause leaning out -especially with turbo engine- and it worse than unburned (no ignition) fuel go out from the engine...

[gunni]

Situation

With correct fuel : nice burn

Not enough fuel : lean

No fuel : no burn no leaning no anything.

Thats why fuel cut is better.

[z0tya]

Situation

With correct fuel : nice burn

Not enough fuel : lean

No fuel : no burn no leaning no anything.

Thats why fuel cut is better.

I found this:
Fuel cut is not realy favored in a race engine, particularly when the injectors are located a long way from the intake valve or where there are high levels of stand off. This is partly due to the fact that when you cut the fuel this only stops the fuel being injected and not the fuel that is still about in the intake so there is a risk of igniting a lean mixture and damaging the engine. The other problem is a lean mixture causing a backfire and igniting the fuel vapor in the air box. Having a Recovery Fuel parameter would only fix part of the problem.

Using an ignition cut only can sometimes cause a lot of unburnt fuel to accumulate in the exhaust which can explode and damage the mufflers. In some cases using the Fuel Cut with Ignition Cut 100 RPM above can help this problem.

Fuel cuts are often used by OEMs because they are smoother and less likely to cause catalytic converter damage but a production road car engine has injector location and fuel targeting that allows them to have greater control of the fuel compared to that of a typical racing engine.

When it comes to randomiser numbers values anywhere between 64 and 182 are fine - just choose something that gives a nice smooth cut. For a more random effect the randomiser should not be too close to a multiple of 255 / #cylinders. Numbers too close to 0 or 255 result in a long run of cylinders firing then cutting which probably isn't what you want.
MoTeC Research Centre

(sorry to offtopic)

[mattias]

Damaging the muffler and exhaust system from ignition cut is the easy problem, the hard problem is the valve train damage that can occur when the exhaust valves get pounded open by the huge pressure in the port. On any engine with rockers, valve  adjustment shims and other parts you can expect bad things to happen, you can snap the rockers or have shims break or come loose and disaster strikes.

Now on a race car engine it might be wise to disable the overrun fuelcut to always have a "wet" intake on engine breaking,  that is a different story - you can still cut fuel as a rev limiter.
There is always the possibility of using the soft ignition cut rev limiter, it's not as bad as a hard cut.

I always use fuel cut as the rev limiter and no engine has seen any damage from that, rather the opposite.

[se7ensport]

The map doesn't look "tuned" to me, and too little resolution on the load axis (TPS) at low openings, seems like you're wasting the bottom 0% line. A simple "VE tune by statistics" on your log  will tell you that (display changes in % to easily see where the values are raised), or enable the "live VE analyzer" and see what it tries to suggest to you.

The rest of your config looks ok, here are a few points :
Set injector PWM % to 100, just to be sure.
Injector angle curve, load defaults for it (720 degrees).
I suggest using fuel only rev limiter, ignition cut is for people not afraid to loose their engine.
In "ECU calibrations" you should change to have "table kPa unit = 2", it will be the standard in later firmwares, you need to look over your Y-axis on all tables after you're done.

Press the "validate" button next to the log button to verify the sanity of the config, some do not matter for you, others do.

Thanks for the info, made a number of changes to the VE config set the PWM to 100% and table kPa to 2, I also upgraded to the latest firmware and vemstune, since then I've cleared all of the errors post validation.  I ran out of time to do much more today as I was also sorting wheel bearings, hopefully I can use the analiser tomorrow to check map properly, as a side note i drained the tank and refilled with fresh fuel - this made the difference of about 5% on the cells tested!

[se7ensport]

A full fresh tank of Shell V-Power and 20mins of auto-tune and its running the best ever, still quite a few cells to hit yet but the difference is billiant, thanks mattias.