My MGB runs - sort of!

[pbsch]

Hey Guys I was active on the site a couple of years ago, but went to the garage and never came back! I got it running a couple of weeks ago and actually took it out for a drive today.

As a result of that drive, I have a few questions, but first a recap of my situation...

3.4L from 1994 Camaro, 7227 ECU from a lumina, running CPS(engine block not pulley), CTS, IAT, O2, MAP, TPS, IAC, knock and speed sensors -

While rebuilding, the intake and exhaust runners were cleaned up for casting flash and sharp edges smoothed, not a proper port and polish, just a clean up. The exhaust is a set of headers (tube manifolds more appropriately) with 1.5" primaries that merge into a 2.5" system downstream. The intake - 3" K&N cold air.

Spec'd out the TPS, MAP, CTS today - they are all spot on. Injectors were all cleaned and serviced - all perform nicely on the test bench. Fuel pressure is up around 45 psi (at idle haven't tested at speed, but suspect it is OK).

The issue - engine starts nice, idles nice and pulls really nice up to about 30-40% throttle and then it stumbles. I assumed MAP or TPS, but since they are both in spec, I'm a bit of a loss. Is this engine modded enough that it needs a re-map?

[pbsch]

follow up - just reading john(forced fierbirds) article about the 7727 ECM swap for 93-95 3.4s. Was it ever verified that the knock sensors were different and would not cross over? Could that be part of my problem?

[Superdave]

the KS should match the ECM, you could be getting false knock but i can't see it stumbling that bad.


Do you have a way to datalog it?

[pbsch]

Not yet - going to the JY tomorrow after work to pick up a ALDL port. Pulled the wires on connector B for the ALDL. There's a guy selling the ALDL-Serial-USB wire on ebay for pretty cheap, so I thought I'd pick that up and then download a logger.

http://www.ebay.com/itm/200766615811 - altho at $30 it's not a super-great deal...

[Superdave]

That's not too bad, they cost around $15 to make + time.

Tunerpro RT is the best datalogger, there are a few files you'll need to get started but it's easy.

[pbsch]

So I've been sitting in my garage staring at my car for the past two months - ok not really, I finished the exhaust system and did some more testing

Compression test 200-205 in each cyl - happy about that
Plugs (new when this car was started) - black and sooty -rich fuel at some point - def rich at idle, cold
Fuel Pressure 45-50 at any RPM in any gear - seems good
I got my ALDL cable and installed tunerpro - couldn't get them to talk. Really haven't been able to figure out which files I need and can't get anything to acknowledge anything else.
Just read that it only works on XP systems - my laptop is windows 7 - going to try and ressurect an older XP laptop and see if i have any luck with that.
At this point I'd be willing to pay someone who is very good with the Tunerpro system to help me set it up and get connected. Anyone interested? PM me!
If anyone knows of a good tuning shop in the Knoxville, TN area who might be able to help with the diagnosis, I'm really at that point. It runs too good at the low throttle positions to be too far off.
HELP!

Thanks
Paul

[pbsch]

So, now I'm starting to wonder if I got some off the wall ECM. It has an AUSJ - bcc code which I can't even find in the tuner forums. I wonder if I'd be better off finding an ARPK or ARPL chipped ECM?

[pbsch]

Robertissar convinced me that the GP1 and burn2 from moates would be the best and most flexible solution. He was nice enough to provide a bin that he thought would work and so far he's been right. Was able to burn the bin to the chip, install and start the car up. No more flat spot on throttle. Cant wait for the rain to stop so I can test it on the street. Skeered to try and test it on wet pavement - it was strong before and sounds much stronger now. thanks Robert!

[pbsch]

OK all you tunerpro experts - I couldn't get my laptop and the ECM to talk using the $88 bin, so I used a $a1 bin and based my prom from an ausj bin. I brought over the main spark table from the 3.4L iron head bin from Robertisaar) and a main VE table from the BAWX. I got the logger running and the ads file I have has a pretty cool total trim table that I can use to adjust all my ves with. Everything seems real nice below about 50% throttle and /or 50kpa, but once the throttle gets opened up, It still falls flat popping and sputtering. I decided to follow my suspicion that it was timing or knock related. So, I logged knock counts and knock timing retard and this is what I found. As throttle opening and rpm start to come up, knock retartd shoots way up - like 20 degrees even though knock count seems to only be a few <6 anyways - does this account for my hole in my high throlttle opening?

[pbsch]

[pbsch]

So, the big spiky green line with the stairstep coming off to the right is the timing retard. The blue line at the bottom is the knock counter - you'll notice it doesn't appear to be moving at all which is why I'm wondering about the retard kicking in. You'll see that on the best pull, I was only able to get 3500 rpm and 80% TPS before i had to back out of it. Any ideas?

I'm wondering if it has to do with some other limiter related to the auto trans portion of this bin? Is there a good MT bin for $a1? I can easily move the VE and Main timing tables easily enough and start over.

[TGP37]

I'm not gonna inject much cause OBD1 isn't my field of interest but.......does the PCM know when it is in drive vs p/n? Could be the PCM thinks it is in park and the rev limiter could be set at 4000 RPM.

In which case, if that is the issue, grounding the right wires will tell the PCM it is in drive. It shouldn't prevent starting as I believe that is a different switch, but again I'm not an OBD1 guy.

[pbsch]

Thanks! It's worth a shot. I'll try it tonight and report back.

[pbsch]

No luck - back to thinking about the possibility of it being knock sensor related... For a quick proof of concept is there a way yo make the ecm think there's a knock sensor when there really isn't? I seem to recall it not being as simple as an inline resistor. How about removing the sensor from the block and wounting it in a grounded non-engine location?

[robertisaar]

3900 ohm inline resistor(in place of the knock sensor) would work.

removing the sensor and grounding it would also work.

you may be able to just disable the knock retard function by changing the maximum knock retard to 0*.