Random injector spikes in Experimental Aircraft LS7 with crate motor E67

[Keith@Dyno-Time]

Both positive and negative injector spikes appeared randomly when the MAF was removed from the installation and changed to SD tune. Any time above 2000 RPM and 30% load, including steady state 100% load. Once in a while spark advance spikes at the same time + or - 10+ degrees. The spikes are 125 to 400ms wide,(200% duty cycle!), engine drops 50 to 300 RPM before recovering. No cats, no O2 sensors, burst knock disabled, transient fuel disabled, dynamic airflow and predictive coefficients disabled. Everything that can be disabled is disabled one step at a time, even oil starvation. No change. Many iterations of VVE tables from very smooth to very rough have been tried, even the stock GM table generates spikes when using speed density tunes.

Seven years ago Tick posted about this same issue, with no resolution found at that time. He was running turbocharged on alcohol, bending connecting rods from hydro locking with the large injectors required. This application is so basic that it is hard to imagine that HP Tuners can't handle it. Besides, many others are using SD tuning without this problem. What am I missing? What else can I try? Any suggestions are welcome.

LS7 crate motor and E67 controller, low rpm cam requiring starting and idle tune modification, otherwise stock. Knob adjustment of IAT for leaning at altitude. Using 77 degrees setting for tuning. Tune and log files attached.

[Ben Charles]

I’ll check this out!! I have a p51 replica in my garage that’s fuel injected

[GHuggins]

Just a thought, but you might want to check the mbt table. Only reason I say that is because to me it looks like "throttle" is floored - it's stuck at 3400 rpms or an 80 cylinder airmass - it "see's" rpm's aren't increasing, so it resorts to "adding torque" by dumping fuel and dropping timing to the 24 degrees in the mbt table. Don't think it would do much for the fuel dump - possibly, but if you make timing the same or even slightly higher like OE, it might take care of the timing swings. A stock ls7 uses a much higher mbt table - not sure if that would fix your issue or if it would just dump fueling and timing in...

Only reason I suggest this is because I ran into some issues a few years ago where the mbt timing table was causing fuel to be dumped in during clutch release events causing engine stalls. I had to go a few degrees above the main timing table to fix that issue...

[Keith@Dyno-Time]

Thank you Greg, for taking your time to look at my problem. It's obviously not a common problem, but not unheard of either. Your suggestion could only come from someone very experienced at fixing things that don't work as they were intended to. I will give it a try tomorrow and report back. I filled the MBT table from 2000RPM and .32g down to the bottom with 36 degrees. I'm not worried about knock since the timing table was developed on my engine dyno with 93 octane fuel. In the plane it runs 100 octane Av Gas and the knock retard system is activated.

I truly appreciate your input, I have tried hard to find a solution, and have run out of ideas. I suspect the solution, if found, will be something unexpected, like you suggest.

On a different topic, your personal Corvettes, have you needed to upgrade the rubber doughnuts to put 1,000 + ft. lb. through them? I am using the Corvette doughnut to connect the engine and prop gearbox and wondered about their durability when taxed like that.

[Keith@Dyno-Time]

Well, the MBT table change didn't make it better, but didn't have any obvious negative effect.

So I tried smoothing the VVE table, again, concentrating on the zone boundaries. That only screwed up the A/F ratio uniformity, and richend everything. Still hiccups.

Anyone know of any other fixes for obscure problems that might apply that I could try?

[GHuggins]

I didn't figure on it helping the fuel dump, but did it change the timing pull any? You may need to try a different OS or even use something that gives you access to more parameters to look at like tunerpro if your able to pull and read with it. Sorry it didn't work for you.

[Alvin]

This looks 100% like the VVE bug some cars have. I've dealt with it multiple times.

You can fix it by applying the 2 bar OS. You can then rescale it to have great resolution 0-100kPa

[Keith@Dyno-Time]

I didn't figure on it helping the fuel dump, but did it change the timing pull any? You may need to try a different OS or even use something that gives you access to more parameters to look at like tunerpro if your able to pull and read with it. Sorry it didn't work for you.

After reviewing some recent logs, the wild timing variations seemed to stop when I disabled burst knock. Timing now has a consistent drop of 1 1/2 degrees at each spike. The following post by Alvin confirms there is a bug in HP Tuners original VVE table computations. I will try the operating system change. Thanks for your keen observations.

[Keith@Dyno-Time]

This looks 100% like the VVE bug some cars have. I've dealt with it multiple times.

You can fix it by applying the 2 bar OS. You can then rescale it to have great resolution 0-100kPa

Thanks for sharing your priceless experience! I've literally spent months trying to fix this. Do I rescale by halving the Map Characteristics under Airflow, or is there a better way? I see that the 2 bar SD is the only available option for my ecu.

[Ben Charles]

You can manually scale the VE table To whatever you like.

[GHuggins]

Let us know if the 2bar fixes it. Usually when what he's referring to happens it causes just the opposite. You'll usually see dynamic drop to 0 along with a fuel cut or lean out.

As for the other you should be able to set map zones although I don't remember it being necessary on the 2bar OS upgrade.

[Alvin]

Thanks for sharing your priceless experience! I've literally spent months trying to fix this. Do I rescale by halving the Map Characteristics under Airflow, or is there a better way? I see that the 2 bar SD is the only available option for my ecu.

Apply the 2 bar operating system. You'll get normal VE tables after that. You can then click on the axis's and change them to whatever you like.

Tip. if you log your GMVE and plot it on a histogram with the same axis's as you intend to run before you apply the 2 bar OS you can drop that into the 2 bar OS VE tables (they start completely zeroed out) it will give you a decent starting place.

[hjtrbo]

If you are going to apply the 2 bar patch let me know. I have a spread sheet that directly converts your existing VE table over to the 2bar. It uses the co-efficients as opposed to interpolation so is far more accurate than other sheets I have found.

[Keith@Dyno-Time]

Wow! Two different methods to make the fix less of a pain. What a great forum! Thanks, guys.

Keith

[Keith@Dyno-Time]

FIXED! Alvin was right on. The random hiccups are caused by a bug in the VVE Table. No change other than switching to the 2 Bar SD version software and rescaling for 1 Bar resolves it.

Thanks to all who responded.

Keith

[Alvin]

Awesome.. Glad to hear.

[GHuggins]

That's good to know. All I've ever seen or heard from it was the dynamic air dropping to 0 and then fuel leaning out or cutting because of it - kind of like cutting the key off. Never seen it actually cause an over fuel dump.