Reasons To Cut Injection/Ignition With Hysteresis

Official FreeEMS vanilla firmware development, the heart and soul of the system!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Reasons To Cut Injection/Ignition With Hysteresis

Post by Fred »

Sorry for the slow reply, this tab hasn't closed since you posted, and the machine has crossed 4 countries since then! :-)
BenFenner wrote:
Fred wrote:
BenFenner wrote:103% pulse with is a result of that math, however wrong it may be.
No, it's not wrong, what it means is that the pulse width that you requested is 3% longer than the pulse width being delivered (by virtue that you can't run at more than 100% duty).
I assumed the logged pulse width would be actual pulse width and not some stupid theoretical value that I told it to perform. Under that assumption, I thought the math was bad.
Your assumption makes more sense.
There is another possibility, and considering what you saw, it's quite likely, I guess, and that is that its an electrical duty cycle. IE, dead time + fuel pw = 103% of avaliable time. In that case you'd get upto about 10% more fuel by exceeding 100%. I would do it as I said, which is "fuel duty" = (fuel pw/available time), however the secondary figure of "electrical duty" = (dead time + fuel pw / available time) would also be useful, of course. That one would tell you where you should have stopped because as soon as (REAL dead time + REAL fuel pw > available time), the injector will hang on and suddenly step up in flow slightly, which is bad. There is no more flow past that point, though, provided that your dead time curves are correct.
When you rear my concern under my assumption, you'll see I am making good sense.
Yes, sure!
Maybe I forgot to mention that 100% injector duty cycle at power peak did not, in fact, provide even stock levels of fuel. Only going up to the 103% value by modifying the fuel table (wasn't even aware we were going "over" 100% at the time obviously) would give us the stock fuel delivery. I maintain it is likely a bug in VEMS. I know over 100% duty cycle is impossible, and said so many times.
It's probably because of what I wrote above. It could also be = (fuel pw/ available time - dead time), hard to say without code!
Fred wrote:
BenFenner wrote:VEMS fuel table is tuned with volumetric efficiency representation or approximation (as I'm sure we are all at least somewhat familiar) and we were around 80% VE at power peak. Putting in higher values didn't result in any more fueling even though the pump was up to it.
The pump has fuck all to do with this.
It sure does.
Lies, but....
And I mentioned it because if you have an inadequate fuel pump you can also run into situations where modifying the fuel table produces zero results. I was just making sure to eliminate that variable. The injectors had reached their limit. Not the fuel pump.
That is true, just almost certainly not relevant when crossing the border between "AC" and DC injection :-)
Fred wrote:
BenFenner wrote:RPM cut and others being always "on" personally won't affect me any.
Tell me, who *will* it affect? No one with any automotive engine.
I guess it would affect those who believe things should be done a certain way, regardless if the immediate result is the same as doing it another way?
Maybe they are confused?
It's only affecting them psychologically, though, not in any practical or pragmatic way!
Like I said before. I'm just so glad I could get the fueling I needed from VEMS despite what seemed to be an apparent bug (to me). 100% duty cycle wasn't enough. 103% was enough. And the mixture readings actually changed between the two settings. Over 103% gave no more however.
See above.
In the real world, I was able to get the fuel I needed without some VEMS developer putting some immovable cap on injector duty cycle.
In this case, it's not the dev putting the cap on it, it's physics. Greater than 100% isn't possible, if your demand exceeds that supply, you'll get a warning that you can then tune around, accurately.
Yah, maybe 1 in 1,000 would have reported that bug
There are more than 1000 VEMS users??? Really??? :-p
I can and will give this project the benefit of the doubt and assume it will avoid many of the pitfalls of other, more poorly designed and implemented projects. I will be a convert. Lead and I'll follow.
Good man! :-)
BenFenner wrote:it's a slippery slope. Not a bad one... It just is. So, you have a spectrum. Crappy defaults on one end, and fully sanctioned default configurations on the other end.
OK, understood, and I am in full agreement with your sentiments. The only sanctioning I'll be doing is "typical turbo jap 4 cyl" such that when I do a setup I have to change the least stuff possible, just to get it running at all. Past that, no, no sanctioned tunes, just one big set of defaults that is semi reasonable for the majority of cases but needs adjustment in almost all of them.
Getting sick of reading my shit yet?
Never! :-p
Fred wrote:If a default rev limit was 500rpm, that would be stupid as it would prevent the thing from starting and idling...
Who cares if the default is 500 rpm? The user should go over the config and catch that.
Right, but when you're struggling in the rain trying to get that weird config starting/idling/running for the first time, the last thing you need is some weird default that stops it from happening buried at the back of some screen that you never find without a 2 hour search.
Or how much do you expect to hold their hand with sane defaults? Just a thought is all...
See above.
And if you're holding their hand a bunch, why not just make the small leap to sanctioned tunes?
See your own post! And yes, I'm aware that I'm being slightly contradictory :-)

Cheers Ben, keep it coming, no matter how much I kick it back ;-)

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
Post Reply