Summer build thread: Megasquirt and more...
#76
Rennlist Member
Thread Starter
Batch fire is usually just fine...the only real differences are at idle. Batch fire cars feel the effects of intake pulse reversion and a little fuel can get splashed into adjacent cylinders. You would also see measurable gains if you also went to more modern ignition, that's good for more top end.
Speaking of replacing old Bosch parts... the speed/reference sensors got the boot today The hall effect sensor transmits a superior signal (square wave 12v, free from EMI) and is very accessible - no more messing around over the bellhousing and playing with connectors. Even if this sensor goes south, it's $40, instead of 2x $60 (or more).
#77
Rennlist Member
Thread Starter
Wiring is more or less done:
...and the injectors and coils are good to go! Short video showing how MS allows diagnostic testing through a configurable GUI. No starts (knock on wood) will be a breeze to troubleshoot.
Currently installing the new injectors, have it all apart right now so I can fully clean and paint the fuel rail. It looks disgusting at the moment.
...and the injectors and coils are good to go! Short video showing how MS allows diagnostic testing through a configurable GUI. No starts (knock on wood) will be a breeze to troubleshoot.
Currently installing the new injectors, have it all apart right now so I can fully clean and paint the fuel rail. It looks disgusting at the moment.
#79
Rennlist Member
Thread Starter
All in good time...
More importantly, it started today just on MS
The tune is still really rough. The first video shows the honest-to-goodness first start. The second video shows the idle once I messed with the tables a little.
The DME had it idling at ~13.3 AFR so I tried to aim for that. I think I overshot the warmup enrichment, so it was rich. Will be playing with all this stuff a lot, and will share tables to those who are interested when they're presentable.
The car already idles better than stock with 5 minutes of tuning, unbelievable. COP for the win! And they're not even dwell optimized yet.
More importantly, it started today just on MS
The tune is still really rough. The first video shows the honest-to-goodness first start. The second video shows the idle once I messed with the tables a little.
The DME had it idling at ~13.3 AFR so I tried to aim for that. I think I overshot the warmup enrichment, so it was rich. Will be playing with all this stuff a lot, and will share tables to those who are interested when they're presentable.
The car already idles better than stock with 5 minutes of tuning, unbelievable. COP for the win! And they're not even dwell optimized yet.
#82
Freedom Enthusiast
Rennlist Member
Rennlist Member
#83
Rennlist Member
Thread Starter
Thanks! and do it!!
Yeah, in fact I need to find where to hook that into on the early gauge. All the info online seems to pertain to the late car, shouldn't be a big deal though.
It's semi-sequential, so it's wired like batch fire but configured to run 2 squirts, alternating. Not sure about mileage. Even with the most crude option (batch fire), the "not on time" fuel just sits in the intake valve until it's open. That's why idle is a little less refined, but once it gets revving most people can't even tell. I should do some data logging between batch fire and semi-sequential to see about economy though.
There are only two injector channels in MicroSquirt, so I couldn't do fully sequential fuel anyway.
There are only two injector channels in MicroSquirt, so I couldn't do fully sequential fuel anyway.
#84
Freedom Enthusiast
Rennlist Member
Rennlist Member
It's semi-sequential, so it's wired like batch fire but configured to run 2 squirts, alternating. Not sure about mileage. Even with the most crude option (batch fire), the "not on time" fuel just sits in the intake valve until it's open. That's why idle is a little less refined, but once it gets revving most people can't even tell. I should do some data logging between batch fire and semi-sequential to see about economy though.
There are only two injector channels in MicroSquirt, so I couldn't do fully sequential fuel anyway.
There are only two injector channels in MicroSquirt, so I couldn't do fully sequential fuel anyway.
I wonder what happens to the fuel sitting and waiting for the valve to open. What happens to the atomization level? I wonder if it leads to accelerated accumulation of "stuff" on the valves compared to a system where the injection is better controlled. So many questins.
#85
Rennlist Member
Thread Starter
Big shout-out to Dougs951S for helping me out today over the phone and helping me dial in the idle, it's rock solid and it better than the DME ever had it idling. The idle bypass screw on the throttle body is almost shut but not quite.
To all the people who are commenting on here about potentially converting to MS, you really need to go through Doug's thread, link here. I sincerely believe it's the way forward for our aging cars - they're already set up for EFI and the Motronic tune is pretty lackluster. Just look at how much people are willing to pay for M-Tune and whatnot.
At idle, it literally "splashes" and pools there...hence why full sequential is a lot nicer at low speeds. Plus you can adjust for minute cylinder variances, such as head temperature. At higher revs it all happens so quickly that it doesn't pool - it just lingers around the valve as a "mist" until it's sucked in. Or at least that's my understanding, someone please jump in if that's not correct.
To all the people who are commenting on here about potentially converting to MS, you really need to go through Doug's thread, link here. I sincerely believe it's the way forward for our aging cars - they're already set up for EFI and the Motronic tune is pretty lackluster. Just look at how much people are willing to pay for M-Tune and whatnot.
Ah, the DIYPNP allows for other things to be added to the configuration. I have added a sequential injection board.
I wonder what happens to the fuel sitting and waiting for the valve to open. What happens to the atomization level? I wonder if it leads to accelerated accumulation of "stuff" on the valves compared to a system where the injection is better controlled. So many questins.
I wonder what happens to the fuel sitting and waiting for the valve to open. What happens to the atomization level? I wonder if it leads to accelerated accumulation of "stuff" on the valves compared to a system where the injection is better controlled. So many questins.
Last edited by odonnell; 07-10-2015 at 12:50 AM.
#86
Congrats on a successful install! I am still on the hunt for a cheap 944 to finish my MS build, when that day comes I will be picking your brain on your sensor mount for the crack trigger. Hell I should just MS one of my turbo cars but they both run so good the way they are at the moment...
#87
Rennlist Member
Thread Starter
Have been tuning the VE table and messing with enrichment curves, it idles like a boss and starts up by itself now. Still a long way from being usable across all load/rev conditions, but the ball is rolling.
My fuel rail was disgusting - there were random discolorations and I decided to give it a fresh coat of paint when I had it off to fit the new injectors. A few tips for anyone wanting to do this: tape off all connection ports (duh), use oil based engine enamel, multiple light coats, use ATF on seals when installing injectors, use new M6x1.0mm hardware for mounting the rail. Also, tape off the surfaces that will be up against the cam tower. You don't want paint transferring onto it.
The gauge is mounted just to verify rail pressure and leakdown:
If you have multiple 944s, what are you waiting for?
You'll be shocked how much "upward mobility" there is in terms of smooth running and a good fitting tune, even for a 944 that's running fine now. Mine was running as good as new on the DME, and if I plugged all the Motronic stuff back in, it would start right up. But that would be a downgrade.
Regarding my sensor mount, it's not really something the average 944 guy can replicate. That's because it's mounted where the lower balance shaft roller used to be I used a Chris White sensor bracket, modified if for my particular sensor, and used a M10x1.5 nut + loctite to keep it on there.
My fuel rail was disgusting - there were random discolorations and I decided to give it a fresh coat of paint when I had it off to fit the new injectors. A few tips for anyone wanting to do this: tape off all connection ports (duh), use oil based engine enamel, multiple light coats, use ATF on seals when installing injectors, use new M6x1.0mm hardware for mounting the rail. Also, tape off the surfaces that will be up against the cam tower. You don't want paint transferring onto it.
The gauge is mounted just to verify rail pressure and leakdown:
Congrats on a successful install! I am still on the hunt for a cheap 944 to finish my MS build, when that day comes I will be picking your brain on your sensor mount for the crack trigger. Hell I should just MS one of my turbo cars but they both run so good the way they are at the moment...
You'll be shocked how much "upward mobility" there is in terms of smooth running and a good fitting tune, even for a 944 that's running fine now. Mine was running as good as new on the DME, and if I plugged all the Motronic stuff back in, it would start right up. But that would be a downgrade.
Regarding my sensor mount, it's not really something the average 944 guy can replicate. That's because it's mounted where the lower balance shaft roller used to be I used a Chris White sensor bracket, modified if for my particular sensor, and used a M10x1.5 nut + loctite to keep it on there.
#89
Rainman
Rennlist Member
Rennlist Member
Ah, the DIYPNP allows for other things to be added to the configuration. I have added a sequential injection board.
I wonder what happens to the fuel sitting and waiting for the valve to open. What happens to the atomization level? I wonder if it leads to accelerated accumulation of "stuff" on the valves compared to a system where the injection is better controlled. So many questins.
I wonder what happens to the fuel sitting and waiting for the valve to open. What happens to the atomization level? I wonder if it leads to accelerated accumulation of "stuff" on the valves compared to a system where the injection is better controlled. So many questins.
#90
Rennlist Member
Thread Starter
Set the timing today. Although the advance is tunable through the GUI, there's still a value called the "tooth #1 offset" that tells MS when to expect TDC relative to the missing tooth. I'm going to lay off on posting videos directly here, but if anyone is interested, here's how to set this with a timing light. My 'OT' mark is really bad, so I did it from the bottom marking. Laziness spurs on creativity, so I used a mirror to see what was going on.
Wiring is also done, save for a few odds and ends. I posted the last several videos, so likewise, here's part 5 for those who are interested.
I decided to start taking videos about the stuff that would be helpful to anyone else who wants to convert, seeing as how there are a few interested people floating around. I won't post them here - instead, I decided to make a YouTube playlist for this project. Whenever I make a video, it will go there.
Link to playlist
Edit: Just took it for the first drive! It pulls! Still need to work on accel enrichment, but it was rock solid. It's smoother than the factory setup by a mile, and it's still a rough tune.
Wiring is also done, save for a few odds and ends. I posted the last several videos, so likewise, here's part 5 for those who are interested.
I decided to start taking videos about the stuff that would be helpful to anyone else who wants to convert, seeing as how there are a few interested people floating around. I won't post them here - instead, I decided to make a YouTube playlist for this project. Whenever I make a video, it will go there.
Link to playlist
Edit: Just took it for the first drive! It pulls! Still need to work on accel enrichment, but it was rock solid. It's smoother than the factory setup by a mile, and it's still a rough tune.
Last edited by odonnell; 07-11-2015 at 10:55 PM.