hi_im_sean

iTrader: (3)

I had the same rust issue on the pass floor, although not as bad.

12:1 NA is too rich. I found a few dyno plots showing these motors make the most power at 13.3:1 at 100kpa, not the "usual" 12.8:1.

slowcarfast

There's some wisdom that indicates the uego should share a ground with the ecu to maintain more consistency between the reading in megasquirt and the gauge. I grounded mine to the ecu ground by the throttle body. Sounds like you already took it a different direction though!

x_25

Every car I have pulled the carpet up on has had the rust there. I think it has to do with the roof likes to leak there on these?

13:1 it is then!

Yeah, I will be seeing how the reading compares and going from there. Need to finish up a few last things, button stuff up, clean up a bit and then it will be time to try and start this.

x_25

Finished the harness:



And got everything installed.



The car starts, but it is off the charts rich. Is the DIYautotune map that far off on the VE table? or is there likely a setting I missed somewhere or something. I hooked the stock ECU back up to check that everything runs correctly and it does. I will try pushing the VE table leaner later.

shuiend

iTrader: (24)

Every car is different. Its your first start and the engine is cold, of course it is going to be way off. You need to tune it.

x_25

Guess I am just . Must be a bit spoiled by my friend's FM Link. on the stock basemap the car started and ran just like it had before swapping the ecu.

The coolant was almost up to temp, so I will try scaling the VE table until I get a somewhat usable reading and work from there.

stefanst

iTrader: (1)

Is your req-fuel correct for your injectors? if you're rich across the board, using the basemap, I'd suspect req-fuel. If you're rich only at small pulse-widths, I'd suspect your injector dead-time to be off.

x_25

Didn't drive it anywhere, so I don't know what it does at load, but it keeps the meter pegged all the way up to 4k. I used the required fuel calculator and it came up with 11.4ms. Which sounds right for blue tops to me. I will hopefully get some time tomorrow evening to play with it and see. The settings I used were 1599cc engine, 235cc injectors and 14.7 for afr.

Also, when I say rich, I mean almost stalling, smoke making, rich. AFR <10:1. It smells/sounds like a race car though, which is fun.

x_25

So I drove around today watching the wideband with the stock ECU. I think the fuel control on the stock 1990 ecu can be described as "rudimentary, at best." But I am guessing you all knew that already.

I was apparently wrong about the 12:1 being richer than stock as well... I had misread something as 12.5:1 being the richest, but it was apparently the leanest a stock ecu runs. :P

Mine runs stoich and closed loop anywhere but WOT or closed throttle. Seems to have no AE and runs, WOT, 13.0:1 at 2500rpm, tapering to 10.8:1 at redline. I am now amazed I manage even 28-30mpg....

x_25

Dingdingding! But It wasn't just small pulse widths. My injector dead time ended up set to 6ms! Turns out when I imported the base tune, curve1 on the injector dead time set up stuff got "zeroed out" to all 19.2v-500%. So I thought the curve was blank (since it looked blank) and my injector dead time was set at 1.2 seconds, but it turned out it was just adding 500% all the time. A quick glance at the fuel equations window made it obvious something was messed up.

So turned out to just be a setting I overlooked. Which is what I thought it would turn out to be. Took the car out with AE and DE off, let AutoTune do it's thing and it is running nice already. Even with AE off it doesn't feel all that bad, no worse than stock.

Next step is figure out idle VE and then tune the IACV.

The learning curve for this thing feels like a brick wall, there are just sooooo many settings and things going on. But it runs!

stefanst

iTrader: (1)

Damn. Surprised it actually idled at 6ms

x_25

Actually, if I left everything alone, final injector pulse width was was 10.275ms at 1000rpm. I have a short log from last night if anyone wants to see that.

This engine seems to not mind doing things way outside of range. When my timing mark had slipped and I set timing by retarding it a bit when it would ping on hot days (I would go out on super hot days and drive up a steep hill for a while), it ended up being actually set to 22* advance base timing (once I had changed the pulley to get a mark that was in the right spot...)

But yeah, like I said, it was fuel dripping, smoke making rich. I figured a setting was waaaaay off somewhere. When I got it set right the first time the base map was actually trying to idle at 19:1 or so.

hi_im_sean

iTrader: (3)

x_25

Exactly. I spent all my down time today pouring over the MSextra manuals at work. Still feel like I don't have any clue what I am doing. At least the car is driveable.

I also learned that if you change too many things too fast, tunestudio freaks out and needs to be restarted on my laptop...

x_25

Thought so far.

I don't know why so many people have trouble calibrating the ECU and wideband so they read the same. Mine was off, but I made a custom linear calibration, set the low point to 12:1, adjusted that VE cell it was in until the wideband read 12 and then adjust the voltage for the 12:1 point on the calibration. Did the same thing at 16, checked 12 again, then checked 14 to make sure it didn't deviate in the middle. Bang on on the whole range. But I spend a good chunk of my day calibrating equiptment at work so...

This tuning thing is a lot of fun!

Data! So much data. Good thing I love looking at graphs and logs.

Tune anaylize live is amazing.

The IAC valve is mysterious in it's ways. I must learn what it knows so I can overcome it.

I had no idea the stock ecu felt so... Rough? The car runs and accelerates and eveything so much more smoothly.

The stock IAT sensor is useless. It doesn't really read the air temperature as much as it tells you the temp of the AFM.

AE is a bitch to tune with Mapdot...

aidandj

iTrader: (5)

If you have ground offset issues there's no guarantee it will always be the same. That's why using a serial input is cool

x_25

Some day I will figure that out. Probably after I make a nice ground and see how that works. It's easy enough to take a quick peek abd see if it is still close enough. I consider +/- 0.2 to be fine. Right now I am within 0.1.

If I ever want to get really fancy, I will add a narrow band to check everything against.

x_25

Preview of this weekends project.

stefanst

iTrader: (1)

Turbocharged potato canon?

hi_im_sean

iTrader: (3)

Im going to go with air filter that is being bolted to some sort of panel, like the cowl/firewall or stock air box or something.

edit- no wait, that **** is for the m45