Delanoso's 2.9L Build

[Delanoso]

Understood about the manifold leak at WOT. But even the plots you just link show that a TPS not allowing the ECU to go inot open loon will cause a reduction in peak power - unless I'm reading those wrong. I just meant that between a TPS not making WOT and the manifold leak, the engine wouldn't have gotten near it's real peak at any RPM. Maybe part throttle above 3500? It actually felt pretty good there.

[Delanoso]

Still running stupid rich or at least it smells that way. TunerStudio shows off the charts lean AFRs or off the charts rich. The tach reading in TunerStudio is different form what shows in the dash. Is there a chance that the coil is firing on a different timing than MS thinks and therefor the fuel pulses and spark are out of sync? That would likely leave a bunch of raw fuel in the intake but the exhaust would read stupid lean, exaggerating the next cycle?

Edit: Never mind, a timing light on the engine has shown correct several times. Still don't understand why the stock gauge and TunerStudio don't match.

[ForcedFirebird]

It's batch fire, so all the injectors fire at the same time anyways.

As I mentioned in the email, when there's a lot of unburnt fuel, it can psych out the sensor and it will read lean. The logs showed dead rich, now matter what we were doing with the fueling.

How did the plugs look when you pulled them?

You can also check the injector spray patterns pretty easily with MS. With the valve cover off, you can slide the whole rail out, with the injectors attached still (if you have the long connector, you will have to remove it and plug it back in). Put some cardboard under them laying flat to see the patterns, and go in Tunerstudio: canbus/test mode>output test modes>injector channel to test: all

[Delanoso]

Plugs looked better than I expected - they were definitely black and rich but they were dry and with out any build up. Changed them out anyway.

I'm thinking I have a 5v short on the o2 sensor input to the ECU. Friday afternoon I suddenly got a reasonable AFR in TunerStudio and a very low idle. Was able to run VE Analyze long enough to get the car running and driving on Saturday morning. I think the logs show that the AFR signal just blips to max (5v) from time to time and then shakes loose. Still has idle issues but I think I found those - I have't told MS what the open and closed limits of the ICV are yet and I think the idle valve frequency is set low. I'll correct those this afternoon.

Also the strangest tach signal issue - TunerStudio shows a correct signal but the dash shows something very different. Idle shows below the scale. A light touch of the throttle send the dial upward to about 1500 or 2000 on the dash while TunerStudio shows 3500 or 4000. At that point the dial in the dash literally changes direction and goes downward at what seems a linear rate to the actual engine RPMs increasing. Obviously I've got something crossed up and sent to the dash in place of the correct RPM signal.

[Delanoso]

Got the o2 sensor sorted out and have the car running reasonably well on the wideband now. DIYautotune helped me find a TunderStudio setting to fix the tach signal. Idle is still a little quirky but it's actually better than what I had on the stock ECU - actually, everything is better than the stock ECU, even with the Megasquirt still untuned. I've run VE Analyze some but I think I've reached the end of what I can do with it. The VE table is spiky and uneven and I realize I can smooth that out by hand but I'm not certain it matters. I think the AFR targets and timing need to be adjusted by someone with real knowledge and a dyno. Put it on the same dyno as before and got what I think is promising results. Is there another 5-10% to be achieved just from tuning?

Before MS ECU:


After MS ECU:


Here's all runs from yesterday:

[2mAn]

After MS ECU
Here's all runs from yesterday:

Id aim for a realistic number around 175-180whp and anything more is a bonus. Might need more cam, but should be a strong setup as-is when properly tuned

[McGyver]

Congrats on sticking with it and getting some good numbers out of it!

Too funny

[ForcedFirebird]

That's more like it!

[2mAn]

It must be significantly better too now in everyway

[Delanoso]

I thought about that post when I saw the new dyno chart. Nailed it.

Yeah, it feels freed up. It clearly needs a bit more work but it's so much better.

[digger]

No real power to be gained with fuel except maybe 2-3k where it should be no richer than 13

as for how much more power it depends what timing is like at moment and how optimal

[Delanoso]

New gauge cluster from Bavarian Restoration!

My econometer doesn't work and neither do the economy functions in the OBC. The old gauge barely had tach and spedo readings. Greg at Bavarian Restoration says I'm missing a signal from Megasquirt. Can anyone direct me to some details on that? Not finding anything on my own that really makes sense.

[ForcedFirebird]

Had to look back at the first post to see if you have an early model. Since you do, the tach, OBC lockout, and the eco gauge all run through a white plug next to the ECU in the glove box. Black is tach, green is OBC ignition lock, yellow is the eco gauge.

Who's MS box did you buy? DIY and Whodwho's are a little different from one another, but either way the yellow wire needs signal - maybe start by checking for A/C on the yellow with car running.

[Delanoso]

I see the plug. The black wire is cut so we got that moved to the proper location - my tach works. I have no signal on the yellow wire. By A/C did you actually mean an alternating current signal?

[ForcedFirebird]

Yes, it's PWM, but you can use a DMM set to A/C to at least tell if there's any signal.