Poor idle, injector fault?

[hillmanjames]

Help!

Recently i have put a 1.8 MX5 engine into my Hillman hunter, the engine is using a ford coil pack with the dual coil output mod to the megasquirt for ignition, MS2 on the V3 board and the latest firmware, and WAS running the standard 254cc high impedance injectors. 3 days ago i fitted a GT17 turbo along with an intercooler and moved the inlet air temp sensor so it is just in front of the throttle body. I ran about with the on the stock injectors until my new injectors turned up. The new injectors are from an Evo 7 and flow 560cc and measure 3ohms.

Upon fitting and adjusting the required fuel and PWM values i started her up and it ran like crap! upon further investigating injector bank 1 wasn't firing at all, regardless of engine speed and temperature and which injectors that i moved the plugs to. I pulled the spark plugs and they were dry ( on the offending cylinders) . I went through the injector output testing stated on DIY Autotunes website and that revealed that there were no issues with it, and also tested each plug on each injector using the output test in tunerstudio which shown that every one was working correctly I then checked the wiring to the injectors and found no issues. Following that i installed some 8ohm resistors inline with each injector and it ran ( after adjusting the settings accordingly) but badly. At idle injector bank 1 seems to be firing intermittently and bank 2 seems to be ok. If you raise the RPM above around 1500rpm it runs smooth as butter and drives great, but as soon as you fall back to idling its missing again. I have also tried reloading the firmware (which updated it) and using a tune made from scratch which shown no improvement.

I pulled the plugs after letting it idle for a while and they read as so, starting from the front of the engine;
1 light brown
2 sooty black
3 sooty black
4 light brown

I have cleaned the plugs and this has made no difference. The plugs also have about 100 miles on them.

So in brief:
Driving is great, no issues, runs smooth and without hick-ups
Idling: rough (set to around 900 to 1000rpm to stop it stalling after blipping the throttle) and you can hear popping from the exhaust.

My MSQ is attatched below.

Thanks alot in advance for any help!

James

[elutionsdesign]

Datalog of this offensive behavior?

[hillmanjames]

Will do. I have changed a few settings ( nothing drastic) and will be taking the car out shortly . Ill datalog it whilst driving and do a separate log when i get back with it just idling.

James

[jsmcortina]

Stop !!!!

The clear difference between injector banks is probably an indication of an electrical problem on the board. You must find the problem and fix it before damaging your engine.

First off, I'd try wiring all four injectors to one channel (the leaner one) and confirm that they now work as expected.

IF that makes them the same, I would suggest that you thoroughly inspect the soldering on the small transistors Q10 and Q13.

Until the problem is solved, drive gently and avoid boost.

James

[hillmanjames]

Thanks James, ill pull it apart tomorrow and do some more inspection.

I have attached a Datalog from starting the engine cold and letting it idle.

James

[hillmanjames]

This morning i have pulled the megasquirt from the car and gone over the soldered joints checking Q10 and 13 along with the rest of the board and found no issue. One of the legs on Q13 looked close to another leg, but was not touching so i re-soldered this and put it back in the car, however the car ran exactly the same as before.

Following that i joined all 4 injectors together so that they can be ran of each set of drivers at at time. There was some difference between the two. Bank 1 was noticeably leaner Idling at around 16afr and 1200 rpm. It was very smooth and did not miss fire.
Bank 2 was noticeably richer, running at around 12-10afr and 900 ish rpm.

Between testing each bank the only modification was swapping the wiring over between the two, everything else was left the same.
I logged each start up and have attached below along with the msq that i was using at the time.

for the meantime, would i be able to just use the bank 1 drivers to get the car up and running as i am without transport atm!

Thanks for the help so far, its doing my nut in!

James

[jsmcortina]

The one time I had a problem like this, it was due to a soldering error (on my behalf) and caused a "flyback failure" on one channel which resulted in additional fuel on that channel. The the rich bank is likely to be the faulty one.

The Megasquirt can easily drive 4 hi-z injectors from one bank, so no harm in wiring them up that way while you figure this out.

James

[hillmanjames]

After a bit of searching and following some tests in this thread :http://www.msextra.com/forums/viewtopic ... 34&t=51873 I measured the voltage across R31 and R35 which gave me the following results (with the original wiring set so that bank1 is driving cyl 1+4 and bank 2 driving cyl2+3) :


Megasquirt ON engine NOT running:
R31 = 11.81V
R35= 0.00V

Engine Idling:
R31= 10.95v
R35= 0.20V


I don't know if this is right or wrong, but i would guess wrong as the other fellow had both his at 0.4 so im guessing somewhere along the circuit that R31 is on there is a problem.


James

[jsmcortina]

Try measuring the voltage to ground at R31 and R35.

These get grounded during the injector flyback clamp phase. If they are grounded all the time, the flyback will be stuck on causing a lengthened injector pulse.

James

[hillmanjames]

Hi, i have measured them again with the engine running.

When looking at the board with the writing at the bottom and the DB37 on your right the resistors measured as such:

R31 Left leg 0.8V Right leg 11.8

R35 Left leg 11.8 Right leg 11.8

all measurements were done to the same ground point

Thanks, James

[hillmanjames]

Does anyone have any thoughts on this? I'm well and truly stuck now! If any of the experts would prefer me to send it to them that can be arranged?

James

[jsmcortina]

It looks like something is pulling R31 to ground.

Are you running low-z injectors? If not, just de-solder R31.
You can heat each leg until the solder melts and lever up a bit, then repeat on the other side. You don't even need to fully remove it, just disconnect one leg.

Then run the car again and I'm hoping that both banks will be the same.

I could assess and repair the Megasquirt, but I'm unavailable next week.

James

[Matt Cramer]

Try replacing Q10.

[hillmanjames]

I'll give that a go when I am Back in the country jmscortina. I am running low-z injectors, but I am using in line resistors so the ecu is just set for high-z injectors.

Thanks,

James

[hillmanjames]

Hello everyone


This evening i tried removing R31 as suggested and that made no improvements. I also removed R35 and tried it, but again, no change to the behaviour of each bank.
Following that i replaced Q10 and Q13 as suggested and that gave no improvements. (I replaced them both so i knew that the new ones worked correctly)

To summarise where i am at now, The car is running solely on the bank 1 drivers. This gives me a nice smooth idle (15afr) and drive. If i switch to the bank 2 drivers (again driving all 4 injectors) it runs rich (9ish afr) and is generally rougher. If i run it properly with a bank for 2 cylinders its all over the place and idles poorly regardless of the tuning I do!

Thanks for the help so far

James

[Matt Cramer]

Are you able to get an oscilloscope and put it on Bank 2?

[hillmanjames]

Unfortunately not Matt. The only tools I have are a soldering iron and a multimeter.

[jsmcortina]

I could do a 1-2 day turnaround fault-find and fix if that was of interest.

James

[hillmanjames]

James I have Pm'd you and will take you up on that offer!

Thanks,

James

[Nigep]

Did you find the fault with this ?