PWM Current Limit has no effect

[styxplo]

Im using Tunerstudio MS v2.6.14. Im also using megasquirt-II _2920.ini, Monitor_v2.920.abs.s19 I have a ford v6 171 with throttle body injection. I rebuilt my throttle body and got new injectors which are slightly larger than the old ones. If I do the initial tune of the Injector Characteristics I can adjust the open time and PWM Time Threshold until the idle changes, but I can adjust the PWM Current Limit down to Zero with no change. Also autotune is adjusting my midrange RPM bins at high vacuum to very high numbers 113-170 range and it studders badly when it's in these blocks. Not sure if this is related or a separate issue.

[Matt Cramer]

Please post an MSQ and data log.

[styxplo]

Here is the Data log and MSQ. When ever I go Wide open throttle is when it studders

[styxplo]

I am curious about the duty cycle and pulse width from my data log. Is there a target or an upper limit for theses?

[Matt Cramer]

I am curious about the duty cycle and pulse width from my data log. Is there a target or an upper limit for theses?

Anything over 100% duty cycle isn't happening, and past 85%, the opening / closing time is going to cut into the amount of actual opening and the injectors might as well be stuck open. So make sure the duty cycle stays under 85% - if it's over that and you're still too lean, you need larger injectors (or have a problem with the fuel system not delivering enough flow).

The target for the pulse widths is whatever gets you the appropriate air/fuel ratio for your engine.

Not sure why it's sticking the injectors open at 0% PWM current. That may require checking the outputs with an oscilloscope.

[styxplo]

From my data log it looks like the duty cycle is roughly high 60's to low 70's when my problem is occurring. The duty cycle spikes when I initially mash the peddle, but settles down relatively fast which makes sense because of accel enrichment. I made the assumption that because autotune was bringing those numbers so high that I was running lean when my problem is occurring. Is this correct. I can't come up with a way to check the fuel flow does do you have any advice?

[trakkies]

One comment I would make from your MSQ is that in general VE table values should start out lowest at the bottom left hand corner and gradually increase in both horizontal and vertical directions.

[styxplo]

I agree and that would make sense. The autotune is adjusting the ve table that way. If I watch my data log my o2 sensor feedback isn't really indicating a lean condition under WOT. It varies right around .78-.82. I'm not sure what my target is, but I believe it should be slightly rich under that type of load. It seems to only be behaving this way in the midrange RPM bins with high kPa numbers. It seems to run fine once it gets into the higher RPM range.

[trakkies]

I'm not really a software type, but your MSQ is showing a narrow band O2 sensor - and the log no actual AFR, only target AFR?

[styxplo]

I still can't figure out my problem. I read that with a narrow band O2 sensor the AFR table is ignored and it only looks at the feedback from the O2 sensor. I squeezed my return line and the fuel pressure doubled and I checked the flow of my pump. It filled a quart in less then 20 seconds. I'm also thinking that if it was a fuel supply problem it would act up in the high RPM area which it doesn't. My base timing is set up at 12 BTD per my sticker and it seems to advance appropriately when I hit the gas. When it does studder the feedback from my O2 sensor is around .80 volts which as I understand it is a somewhat rich condition. I don't understand why autotune is taking my numbers so high in my problem area. I've read in the tuning section that VE entries aren't typically over 130. Autotune is taking mine to a high of 173. I'm not sure what my next step might be.

[DonTZ125]

It is extremely difficult to do any tuning with a narrow band sensor. You'd be better off using the Mark 1 Butt Dyno.

The engine wants what the engine wants. The VE cells go up to 255, and sometimes that's what is needed. That said with "Incorporate AFR" and "Multiply MAP" turned on (or their B&G code equivalents), the VE map should be very flat.