VE numbers - big jump from idle to cruise

[BennVenn]

Hi everyone!

I've been adding fuel injection to my 70's superbug mostly from scrapyard parts. At the moment it is throttle body injected from an early 90's Mazda 121 (here in OZ) and I've managed to tune it well enough for it to be a reliable daily driver. Idle AFR's are in the mid 13's, cruise low 14's and WOT high 12's/low 13's. It is a low compression engine so it prefers to run and idle a little on the rich side (the long cold intake runners don't help much)

I am looking for some advice in regards to my VE Table. Although running and driving fine, the sharp increase in VE from idle is a little confusing. I would expect the transition from 30 to 65 % to display some unusual AFR's but instead I get a very smooth AFR transition. My only thoughts are perhaps my idle PW is bordering on the unreliable non-linear region of my injector?

There is no data on this injector and the dead time was calculated from sampling dynamic flow rates at various PW's. Also the theoretical max PW was bumped up .7ms to smooth out the idle (suggesting that the Idle PW is approaching non-linear region)

This TBI setup came originally from a 1.3L engine, and is now on a 1.6L which idles around 1000rpm so I would assume if the Mazda managed to avoid the non-linear range of the injector, a 1.6L would be well above it.

VE.JPG

I should add I have designed, built and written all the firmware/software for my ECU - which is why the VE table will not be familiar to you, but the theory is all exactly the same. I'm hoping someone could offer some insight into the VE curve.

This engine is never revved past 4000RPM, and much of the VE table is never accessed during daily driving.

Additionally, the Injector is low impedance, driven by a peak+hold driver. And raising the idle VE results in a proportionate drop in AFR. Also, Cold start enrichment (130% at 20degrees) increases the injector PW as you would expect but the AFR's don't become excessive as I would have thought they might if the injector is on the edge of the linear region...

Look forward to hearing your theories!

build info at http://www.bennvenn.com

[billr]

I'll bump this up for you...

[BennVenn]

Got it sorted, Others might find this interesting.

I've read quite a few posts about injector dead time and peoples varied opinions about how accurate it needs to be. It turns out that an inaccurate dead time can be corrected with a bit more work on the VE tables. In my case, having the dead time larger than required will increase the range of VE numbers.

For example, in my case - A dead time of 1.1ms and Theoretical max PW of 5ms results in a VE of around 30 at idle and 90's at WOT (for a given RPM)

If I were to drop .4ms off the dead time and bump up the max PW by .4ms I would end up with a VE of 60 at idle and the same VE(90's) at WOT.

Why bother? Well a larger VE gives you a finer resolution when tuning. 1 VE unit change at 60 = 1.7% variation of PW whereas 1 VE unit at 30 = 3.4% variation.

Of course, having your dead time and max PW set incorrect will throw out your calculated pulsewidths as the PW will change non-linearly with manifold pressure (due to pressure density formula).

Looks like I'll be pulling the TBI off and re-calculating flow and dead time...

As for the requirement of an MSQ and Datalog file attached to a post, It is not always required. I would have thought someone with experience in tuning VE tables would have picked up the inaccurate Dead Time just from the range of numbers.

Hopefully someone will find this information useful when getting their VE tables sorted.

-Ben

[kjones6039]

About to admit to being a dumb a$$ here.........

What program is that screen shot from? I have never seen it before......

Ken

[racingmini_mtl]

The non-linearity will be affected by more than just manifold pressure. All the corrections will be affected and will result in a wrong fueling except for a very narrow range of conditions.

If you absolutely want to increase the VE range, change the req fuel value. That is a multiplicative factor so that will keep everything working as it should, at least for most things now.

Jean

[BennVenn]

Ken, it is one of the many entry tables I've written for my ECU. I've custom built the hardware and software to suit my engine/needs. Running a 32bit ARM MCU @ ~90mhz leaves me with a LOT of processing power spare for all sorts of future trickery.

Jean, AFAIK the VE table is used for fueling corrections throughout the entire load/rpm range. What other corrections will be offset by a skewed idle PW value? Accel enrich is a % scalar, as is temperature enrich. Closed loop mode shouldn't be bothered too much as its % based?

[racingmini_mtl]

I'm sorry but this is a MegaSquirt forum paid for by the MegaSquirt people. So this is not intended as a generic EFI forum supporting other units.

Jean

[kjones6039]

Ken, it is one of the many entry tables I've written for my ECU. I've custom built the hardware and software to suit my engine/needs. Running a 32bit ARM MCU @ ~90mhz leaves me with a LOT of processing power spare for all sorts of future trickery.

Jean, AFAIK the VE table is used for fueling corrections throughout the entire load/rpm range. What other corrections will be offset by a skewed idle PW value? Accel enrich is a % scalar, as is temperature enrich. Closed loop mode shouldn't be bothered too much as its % based?

I had a sneaking hunch that it wasn't MS.........


Ken

[BennVenn]

Apologies Jean, I didn't mean to offend. I did mention in my original post that this was an ECU of my own design. I was just looking for some general wisdom from like-minded enthusiasts. I'll take my questions and experience elsewhere.

Regards,
Ben

[billr]

Right at the beginning I suspected this wasn't an MS system, but gave the OP the "benefit of the doubt" and asked for the msq/msl; a subtle way to find out if this thread was bogus. And no, Ben, it wasn't clear that your system is totally non-MS. You hinted at that, but a lot of folks do software and hardware mods to the basic MS systems. However, no hard feelings, good luck with your project.