Injector Test Mode operation?

[kredden]

Hi,

I just tried using Injector Test Mode and only one bank of 3 of my 6 injectors are firing. I assumed test mode should fire both banks, is this incorrect?

Reason I'm doing this is that I had just gotten my car up and running with a basic tune (starting & idling) on the MS and was driving around my local "test track" (commercial zone road loop) and after about 20 minutes the car died. When I try to restart it, it won't idle and runs like crap, basically like it is not firing one or more injectors... I did check a lot of other things and everything seems OK. I tested with the Stim and both Injector LEDs are flashing in alternating sequence. I have my injectors set up with the following:

Speed Density
Squirts/Cycle: 2
Inj Staging: Alternating
4 Stroke
6 Cylinders
Port Injection
6 Injectors

So did I find my problem or is injector test mode working as it should?

Thanks,
Kevin

[kredden]

Hi, I'm using low impedance injectors an have an issue where one bank of 3 has stopped firing. The injectors are fine as I swapped wiring harness connectors with the ones that work and they spray fine us MS test mode. I was on a test drive for about 20 minutes messing around with the auto tune feature and all was going well until the car died. I did some poking around and found this blurb in the "Injectors and Fuel Supply" section of the manual:

One sure way to know if you can't use the standard V2.2 flyback circuit is to have a flyback failure. The circuit will most often fail after some time spend at high speeds and loads, rather than immediately when you start the engine for the first time. Generally, when the flyback circuit fails, the MegaSquirt® works okay on the stim, but not on the car.

Signs of an impending flyback failure are:

MegaSquirt® will often require higher PWM% over time,
The engine may start running erratically, especially at higher speeds and loads,
the injectors may 'stick' open and flood the engine.
When the flyback fails, sometimes Q1 (the TIP32 on the bottom of the PCB) looks pretty rough, all burnt, etc. However sometimes it looks fine.

If Q1 (on the bottom of the board) does look burnt however, this is a sure sign of a flyback failure.

I had some of the symptoms:
- I did spend time some at high load and highish RPMs (~4600 range) and was just coming out of this when the car died.
- The MS works fine on the stim but not on the car.
- The PW's matched the duty cycle pretty closely for most of the run, right near the end they rose quite a bit over the duty cycle numbers, that might have been because I had some bad data still in the VETable causing it to go very lean for a short while though.
- The injector do NOT stick open
- Didn't see any burns on the PCB

Now all this is written in relation to the v2.2, can it happen on v3 as well (which is what I have). I built it with both the PWM flyback damping circuit as well as the standard flyback circuit. When I run the MS injector test mode I can hear the injectors cycling but no spray comes out just a lot of clicking. I had also swapped fuel rail & injectors with another set I had laying around and saw the same behavior. I tested the voltage at all the injector Bosch connectors and they all showed the same voltage (12.36v).

If the driver components are failing would I see the behavior where it is still activating the injectors but just not with enough durations or something so they don't actually squirt?

Any pointers would be greatly appreciated.

Thanks,
Kevin

[trakkies]

If this helps, the injectors are fed with +12v all the time the main relay is made. The MS grounds them.

If you wish to see the duration any individual injector is open for, a dwell meter will show this. Dwell meters were used for setting points on old ignition systems, but can still be found as part of an auto DVM - or self contained perhaps secondhand on Ebay. You could also use an LED or noid light which will flash at low RPM. But a dwell meter will give a reading at any speed. The dwell meter shows the time the device is on relative to the time it is off and shows this as a percentage, etc. So would show up any difference between the two banks at a steady speed.

[kredden]

Thanks for the tip, I'll see what it shows. Is it better to use an analog dwell meter for something like this or would a digital work just as well?

Kevin

[Matt Cramer]

I've just put together a tech article on how to check the injector driver circuit here:

http://www.DIYAutoTune.com/tech_article ... ooting.htm

[trakkies]

Thanks for the tip, I'll see what it shows. Is it better to use an analog dwell meter for something like this or would a digital work just as well?

Kevin

The analogue ones have electronics too, so probably no difference. You're not looking for an actual value - just a way of comparing one known good bank with the other. But it will show the opening duration changing with rpm etc, so quite a useful tool to have in the box. I've seen older ones for 50p at car boot sales in the UK.

[kredden]

Excellent timing Matt, thanks! I'll try this out tonight.

Kevin

[Bernard Fife]

kredden,

You might also check number 11 (and 6) here: http://www.megamanual.com/ms2/V3trouble.htm

Lance.

[kredden]

Yeah, I had already looked there but it looked like that was a test for if it didn't work on the Stim, I'll look closer.

Kevin

[kredden]

Matt,

I grounded pins 21 & 22 of U1 and the LED comes on as expected so does the mean my injector circuit is fine?

I did check a couple of the other steps:

First, check pins 2 and 4 on U4. If they're not getting 5 volts in the off state
I get 4.93v in off state, I assume this is good enough? (I'm using a power supply for power)

U4 is an inverting driver. If you don't see its output on pin 7 go high when pin 2 is grounded, or pin 5 go high when pin 4 is grounded, you've got a bad U4. Same if it's stuck high when it gets 5 volts on the input. Note that if you have a 9 volt battery, the points I've marked at "12V+" are going to be lower, more like 8 volts. You'll usually see somewhere around 12 volts on a power supply. If you're seeing more like 10 volts, don't panic; it's still working OK. Time to move on.
I get about 11.5v or so on pin 7 & 5 when 2 & 4 are grounded otherwise I think they were around 4.93v or so when not grounded. Also the LEDs on the Stim go on when I do this.

Check the input signal on pin 1 of Q1 and Q5. It will be a little lower than the voltage at U4, but not by much. If the voltage has disappeared, power it off and check the ohms on R15 or R20 as appropriate. Either R15 or R20 has failed, or the transistor in the current limiting circuit (Q14 or Q15) has shorted out and needs to be replaced.
I didn't do this step as I wasn't sure which is pin 1, I didn't see it labeled in the manual? Which one is 1?

bolt
---------
- -
- -
---------
| | |

I didn't do the rest due to not knowing which is pin 1. The weird thing is that the injectors that are not squirting are trying to, they aren't just dead. I can hear them clicking and spitting a little so I'm guessing that they just aren't staying grounded long enough.

Kevin

[Matt Cramer]

The one thing this check doesn't test is how much current the transistors can sink. It could be your pulse width you're using to test is less than the dead time as well.

[kredden]

The injector test PW that I'm using is 10 and the injector test off time is 60. The one bank runs fine though. Also can you tell me which is pin 1? My text drawing attempt in my previous post didn't come out too well.

Looking at the board as you mention in your test writeup, so that the pins for Q1/Q5 are facing down (and the hold down bolt facing up) is 1 on the left or the right? I'll do my picture using "."'s as spacers instead of spaces, maybe that will work better:

.....Bolt
===========
=............=
=............=
===========
.|.....|....|.
.1.....2....3.
or
.3.....2....1.
?

Thanks,
Kevin

[Matt Cramer]

It's the top one, 1-2-3.

[kredden]

Thanks.

Kevin