Going to sequential fuel- normal to require a buttload of retuning?
 

Installed MS3 last month and have been running in batch/wasted without MS3X, until now. Got the MS3X in and wired up the injectors for sequential.

Car started up and runs but the wideband is basically pegged lean on the VE table that was tuned great before. I was under the impression that any retuning would be minor. In some cells it looks like I will have to almost double the previous VE values. I will post my settings for y'all to look at. I guess redoing the tune from scratch isn't so bad these days with VEA live, but I don't want to have to do it twice in case I'm missing something stupid. Injectors are RX-7 550s by the way.

Attachment 187787

Attachment 187788

I'm using 360* everywhere for injector timing. I found a table posted by Brain with more advance but that didn't help and probably made it run worse.

Still running wasted spark through the main board LED outputs but being this is the first sequential anything I've done I will post these just in case.

Attachment 187789

Hmm... Does it run smooth? Sure you don't have a wiring problem? You really should be pulling a little fuel instead of adding fuel. I'm sure Brain will have input since he's done this, I started w/ sequential.

Hopefully the wiring is good. I only have 4 wires coming out of the MS3X:
Inj A, B, C, D on pins 19, 16, 13, 10 wired to 2U, 2Y, 2Z, 2V

I had trouble tuning idle because I had the smaller idle VE table enabled, so it didn't matter what I put in the main VE table. Now that I realize that, I can work on getting it to idle smoothly tomorrow. It ran rough, but no rougher than I'd expect with an AFR north of 19:1. I ran VEA live and had it autotune some non-idle cells (holding light throttle to free rev to 2500 rpm or so) and it ran a bunch smoother as the autotuner richened it up, though not silky butter smooth. I'll wiggle the wires around tomorrow with the car running to make sure there's not a loose pin or something, but the fact that it needs that much more fuel has me doubting my tunerstudio settings.

you may want to run some of the grounds on the MS3X board.

I see that the MS3X has some grounds back to the MS3 card via the ribbon cables but I was wondering if that's sufficient or not. Guess I'll add the 5 ground wires out the MS3X and run them to the lug by the throttle body where the rest of the ECU grounds are, and see if that makes a difference.

yeah, I can help this weekend ;) ;) but sounds like you wired the injectors and setup the software correctly.

I remember having to alter my req_fuel a little when I made the switch, IIRC I went from 5.5ms to 6.5ms. been a while.

When I was running batch I had reqfuel at 5.5 or 5.6, I forget. Pretty sure I have always told the built-in calculator 1839 cc displacement and 550 cc injectors and a 13:1 target AFR. When I redid the calculator after setting it for sequential, it spit out a reqfuel of 6.3, and even using that new value it was still extremely lean.

When I get home I'm gonna load my old batch msq and try the calculator again, to make sure I've been using the same factors to calculate reqfuel, and by corollary, that I'm not losing my mind.

Those wires in the ribbon cable are like 24 gauge or something... no where near what you want for injector current, definitely need those grounds in there.

I saw you said you have injection timing at 360 everywhere. This link might help, its for the diypnp though. It suggests that cranking timing be 360 and fixed injection timing be at -350.

http://www.diyautotune.com/diypnp/docs/sequential.html

370 = -350

I idle at 375° which finishes squirting 10° before my intake valve opens, which is 365° which is 5° before top dead center.

http://www.boostedmiata.com/gallery2...ion_timing.JPG

MS3 has a 12x12 injector timing table, not just the crank and run settings like the older versions have. I made a version based off a table Brain posted a few months ago that I think he was/is running without issue, and tried that versus 360* everywhere which is the default MS3 table and not much seemed to change. I don't think injector timing is the issue; it's very likely the absence of those ground wires. I'm going to try to bail out of work a little earlier today to pick up some wiring supplies and get er done. Hopefully.

EDIT: I basically took that table and scaled the axes to better match the rows & columns of my other tables, not that it's really necessary though. Once I get the car running right I will play around and see if I can optimize the timing any better. Brain, you spend much time tuning that?

Adding the MS3X ground wires totally fixed it. Loaded up my old batch VE table and drove around for a while last night and VEA live pulled a little bit of fuel from a bunch of cells. Needs more tuning time to get the VE table more solid but it's running great. I'm using an injector timing table similar to the one above and will play around with that later to see if it can be improved any.

The DIYPNP base map for my '95 seems to have cranking timing set to 350 and injection timing -340. Would I be better off with the settings on the site, Brain's settings, the ones I have, or does it not make a difference?

the "rule-of-thumb" is 10* before intake valve opens, when using end-of-squirt. the '94-95 opens at 5° BTDC as well.

What pins did you add additional ground wires on?

First I got the fuel map tuned pretty well using a version of the injector timing table posted above. You can see in the 3000-3500 rpm columns the timing is 390*. I then dropped it to 380* in my 3200 rpm column then did some highway driving while varying the load around that rpm with auotuner running. It pulled about 1 point on the VE table at 88 kPa and 100 kPa and made no changes in the lower MAP rows. Not much improvement, but it's something. I guess as long as you're in the ballpark the gains won't be massive, but 1-2% is still something.

If you have autotune set on "hard" or "very hard" the changes will be very slow to happen though. For the idle cells, it may be best to do a scatter plot with injector timing versus AFR (while leaving the VE table constant) and run the timing that shows the richest AFR. If you were going to blindly start from scratch, 375-380* would probably be a good range. I guess that would be equivalent to -340 to -345* if you're using negative values? Also keep in mind that I can only speak to the realm of 100 kPa and under for now.

All of it, son.

The ground pins on MS3X are 2, 3, 8, 12 and 17, and I added them all. I used 18 ga wire on the connector pins, then joined all 5 of them into a little ground bus bar that I bought at HD and stashed under the passenger seat, then it's a single 12 ga wire from there to the ground point by the throttle body. I previously had just the grounds coming out the main board and that was fail.