what in tarnation!

[dshrek19]

Hi all, first off I'm a new member and just recently finished the "efi tuning fundamentals" course. From Manitoba Canada
sorry for the long post just wanna supply all the info I can
Ive built a 1991 Honda civic si 
with a 98 crv b20b4 stock engine 
a 60trim ebay turbo(to be replacd)
45mm turbosmart wg
2" hot side piping, 2.5" cold side 
Fmic(to be replaced)
Ebay knock off bov(to be replaced) 
skunk2 alpha tb and intake manifold
precision turbo&engine 880cc(flow tested at 840cc 2% variance) unknown dead times 
2.5" open hood exit
2" wg dump
hondata 4bar map
hondata boost control solenoid
hondata s300 v2 p06 ecu boost control components and eld resistors cut
autometer wideband wired to d14 O2 input(will not idle on d10 eld input)
innovate boost/vacuum guage
k so basically the story goes like this. sunday night the car was out running just fine for 3 plus hours on a conservative tune I've been playing with. provided here as citycruises.skl. datalogs to go along with it aswell sunday1 and sundaycruise1
so Monday I remove the ecu and take it to the local hondata dealer to have boost control components installed and the eld resistors cut. I receive the ecu on Wednesday and plumb and wire in the boost control solenoid Thursday and Friday. Saturday I go to start the car. after the wb completes its 20 second warm up(which makes the engine run differently) it goes to 16.5:1 + afr constant and makes less vacuum on all three vacuum sensors(hondata, innovate, and mechanical test guage) on the same tune. it was a slight bit cooler out that day but my compensation tables shouldn't be that far off. 
So after buying a vacuum tester and a compression tester, as I had used it pretty hard the weekend before and was unsure of the condition on this engine already having 300k km. Everything tested fine. cyl 1 181psi cyl2 183psi cyl3 184psi cly4185psi. altho when doing a boost leakdown test I can hear air passing through my crankcase and out the valve cover breather. I still tore the intake manifold, injectors, fuel rail all apart, tried going from bwr thermal intake gasket to hondata, checked all O-rings and sensor gaskets, no difference.
so now comes the interesting part, after hours of internet research and forum searching I have come across 0 answers and am resorting to creating my own post. 
hooking my wb up to d14 o2 and d10 causes the car to run differently depending on which it is connected to. when I watch my sensors on s300 manager I can see the eld voltage acting extremely erratic. as in it never goes 0 and is always jumping between .5 and 2 volts in a matter of mili seconds even with nothing connected to it. 
when I switch the ecu to closed loop, it goes back to idling normal, pulling -19 to -20 vacuum instead of -17 running back at a 13.8 afr idle and everything seems to stable out seen here in open loop vs closed loop datalog and the openloop run datalog
I'm wondering if the local shop had messed something up in the ecu, if adding boost control components, cutting eld resistors or of if pinning into a11 for the solenoid would cause open loop to not function properly or are my tunes just so far out that its mucking something up. ive uploaded a tune ive been playing with to try and correct it entitled wtf

[dshrek19]

more data

[Spunkster]

Without knowing the dead times you will have a hard time properly tuning this. You will need to get these before doing any tuning.

IF you are having trouble idling after connecting to "D10" then it sounds like it may not actually be connected to D10 and is in fact connected to a different pin. We have seen this happen to people before that swear their wiring is correct.

There is also a chance that something has been damaged in the ECU in the installation process. You should discuss this with the shop that did the work.

[dshrek19]

thanks spunkster I have contacted PTE with no result. I shall pester some more
guaranteed the pinning is correct, 5th pin bottom row looking at the back of the plug.
why when connected to d14 or d10 does it change how it runs in open loop, the wb should have no effect correct?
it also alters how the engine runs at first start when the wb goes into warm up because the voltage its supplying is incorrect. this all happens in open loop when it should not be doing anything. why is this?
also the eld voltage display on s300 varying extremely rapidly makes me believe its ecu damage
I am in contact with the mechanic who had done the modifications. he is placing blame on my tunes

[Spunkster]

The fact that something is changing the way the car runs when something is connected to the "ELD" pin makes me suspect it is in fact not connected to the ELD input, but some other input. You can search the forum and this happened fairly recently to someone else.

If you are getting strange readings with nothing connected then there could be other issues as well. All you can do is process of elimination. Try a different ECU or try your ECU in someone else's vehicle.

[dshrek19]

Unless a p06 has a different eld pin than other obd1 ecus i cant see how its the wrong pin. Ill break out the multi meter and see if that wire produces the same erratic voltage being recorded to smanager.
As i have it, the eld wire is cut running to nothin but smanager is reading very erratically between 0-2v. Ill see if i can fond another car to plug the ecu into or another ecu to try out till then im stuck
Not sure why itll run perfect on closed loop either....

[Spunkster]

https://www.hondata.com/forum/viewtopic ... 19&t=18296

[dshrek19]

Yea, read that already. Only thing i havent tried is to wire a stock 02 to d14. Which is hard as my dump tube is only ao long. Couldnt be my stock 02 wiring as im not using that at all. Its sensor to guage to ecu. Grounding on ecu ground or tstat housing made no difference with v compensations. Fuel pressureis 43.5

Plain and simple, car drove into garage. Ecu was modified and now it wont drive out.. hoping its not an ecu but what the heck happened otherwise

[dshrek19]

Here you can see i have the correct pins for d10 and d14 seperated from the conversion harness, correct?
I metered the d10 wire and on my meter if fluctuates between .01-.04v with nothing attached. My smanager is reading a very unstable voltage of .5-3.5v moving so fast its basically un readable. Should i start probing pins on the ecu to find a connector exhibiting that behavior?

[Spunkster]

It sounds like something has been done improperly inside the ECU, or it is damaged.

[dshrek19]

I have contacted hondata today, spoke with Mario. He had told me to erase the ecu and load the stockb20 non vtec tune from the hondata provided list. to set the injector dead times which I called pte and recived today, disable sensors and upload. I did this with no change, still a very unstable running condition.
while watching the laptop, hooked up engine running, I went to the sensors tab. I can see the voltage that leaks into the eld is also leaking into my o2(d14) my MAP(d17,d19), RPM signal(?), and inj cycle(?) all seem to follow the same erratic behavior. I am about to go take the ecu apart I guess and look for leaky capacitors as per marios next suggestion, or should I leave it sealed and just send it to hondata?
I don't trust the local guy anymore as we were not on great terms to begin with and maybe was a bad idea to give him my ecu.
attached is a datalog and tune of todays test

[Spunkster]

It's up to you. The form for ECU socketing and adding capacitors can be found at: https://www.hondata.com/downloads/ecu_s ... g_form.pdf