Hello to all, this is my first experience megasquirt. Need some help. I recently got my V3.0 board with MS2 running, fuel only for now. I'm using some VERY low-Z injectors, these are 1 Ohm. It's a Toyota 2.4L getting a Nissan v6 TBI. Fuel only, I'll do ignition after the basics of the fuel side are worked out. Triggering is on the coil (-). Injectors were rebuilt by witchhunter.
So I'm working though the idling first, and am having some behavior that I feel needs to be understood before continuing to longer pulse widths in other parts of the tune. Basically it idles great, although a bit rich, with 4 squirts per cycle simultaneous, but goes into a bad lean misfire if I change it to 4 squirts alternating. With the throttle plate closed I wouldn't have expected such a difference due simply to fuel distribution. Is this somehow reasonable or acceptable? I'd like to have better control over my idle mixture by using alternating squirts for the longer pulse widths.
The PWM feature appears to be working fine, see screenshot. Ignore the file name, I had a huge vacuum leak at first.
Also I made a video of the behavior, although somehow amazingly my phone filtered the engine noise to the point that you can't even hear it change from running well to poorly.
http://youtu.be/ReJcvzCw0WY
Low-Z, PWM, alternate vs simultaneous... What's wrong here?
Moderators: jsmcortina, muythaibxr
-
Matt Cramer
- Super MS/Extra'er
- Posts: 17507
- Joined: Thu Apr 16, 2009 8:08 pm
Re: Low-Z, PWM, alternate vs simultaneous... What's wrong h
Do you have data logs of the issue as well?
We have not been very impressed with TBI fuel distribution; that could easily be the problem.
We have not been very impressed with TBI fuel distribution; that could easily be the problem.
Matt Cramer -1966 Dodge Dart slant six running on MS3X
Re: Low-Z, PWM, alternate vs simultaneous... What's wrong h
Matt, Thanks for being willing to help. I figured out this problem, and will give details here which might help someone in the future.
I had one setting masking a different problem. The underlying problem was that the VE table was just extremely lean everywhere for my setup. So lean that when calculating pulse widths predominantly from the VE table it would go into a very lean misfire. The VE table I started with was for a toyota 22RE, but was for stock + MPFI instead of performance + TBI.
When I was in simultaneous injection mode the lean problem was being masked by my having the injector opening time set too long. I was messing with the injector opening time (and PWM threshold) to smooth things out, because it seemed like these injectors would be the first unknown I needed to tackle. I'll throw out some numbers here for example, probably not what my numbers will actually be: If I had my injector opening time set to .6ms longer than what it needed to be, then that's .6ms longer pulse width then what's called for by direct calculations from the VE table. That will mask a lean problem if you happen to be within .6ms pulse width of having a lean misfire. With the expected low pulse widths of idle a change of .6ms is a LOT.
The low VE problem becomes unmasked when switching to alternate injection mode. The injectors fire half as often, with double the part of the pulse width calculation from the VE table. However, half as many injection events meant the extra fuel from the extra opening time occurred half as often as when it was in simultaneous mode.
Hope that makes sense and helps someone else out.
I had one setting masking a different problem. The underlying problem was that the VE table was just extremely lean everywhere for my setup. So lean that when calculating pulse widths predominantly from the VE table it would go into a very lean misfire. The VE table I started with was for a toyota 22RE, but was for stock + MPFI instead of performance + TBI.
When I was in simultaneous injection mode the lean problem was being masked by my having the injector opening time set too long. I was messing with the injector opening time (and PWM threshold) to smooth things out, because it seemed like these injectors would be the first unknown I needed to tackle. I'll throw out some numbers here for example, probably not what my numbers will actually be: If I had my injector opening time set to .6ms longer than what it needed to be, then that's .6ms longer pulse width then what's called for by direct calculations from the VE table. That will mask a lean problem if you happen to be within .6ms pulse width of having a lean misfire. With the expected low pulse widths of idle a change of .6ms is a LOT.
The low VE problem becomes unmasked when switching to alternate injection mode. The injectors fire half as often, with double the part of the pulse width calculation from the VE table. However, half as many injection events meant the extra fuel from the extra opening time occurred half as often as when it was in simultaneous mode.
Hope that makes sense and helps someone else out.