MSefi.com read only forum (Jan.15,2018)

I was testing in injector test mode I entered 100 pulses of 60mS with off time of 60mS and now it is "stuck"

it has the injector running just straight "on" and I can't get it out of test mode. even after setting the # of pulses to 0 andburning or not burning that value and setting to normal mode and burning or not burning and cycling power or not cycling power.

I was concerned because the instructions were vague and my concerns have been validated

My truck is stuck in my shop till I remedy this situation so if some could get back to me pretty quick that would be awesome

Vicoor,

That's strange. When the test is complete, test mode then sits there waiting for either a reset command from MegaTune or change the test option and reset to resume normal mode. So resetting to normal mode and then cycling the power should get you back to normal operation.

Since that doesn't work for you, you may need to do something more drastic. If this was mine, I would re-load the code using the serial monitor (aka. "bootloader") and re-load your MSQ. The instructions to load the code are here: http://www.megamanual.com/ms2/code.htm

Lance.

Thanks Lance for getting back to so quickly.

So can I just bootload the new s19 file over my older one, I think it is v2.886?

will I need to do anything else?

Right now I'm at work, and the laptop I have for tuning has no internet or USB, luckily I can burn a CD with the files I need to get them on the laptop.

Will I be able to load my saved MSQ or will I need to at all?

Viccor,

You will need to load your MSQ after burning the code (otherwise you will have the defaults for all settings).

The other thing I can think you might try before doing that is disconnecting the serial comms cable when cycling the power (if it was mine I would also disconnect the DB37 to make sure no power is being supplied at all). It is possible that some serial cables supply 5V to the board and keep it alive, preventing it from re-booting.

But reloading the code and MSQ ought to be fool-proof (it always is for me).

Lance.

I'll go try the serial cable, so I should set the number of pulses to 0 the set it to normal mode and then burn or not?

then disconnect the serial cable and cycle power?

I loaded the v2.0 downloader to the laptop and it won't run, says there is a missing .dll

it's an old toshiba tecra that had Win95 and now has Win98

Is this expected?

If I can't get it with this laptop, I have to have my wife bring the other one.

all I was trying to do was get some real world injector info for the parameters for MS3

I see that the s19 is a newer version, think it will behave better in test mode or should I can this as a test method?

I got it running, thanks for the help.

I had to burn the new s19 file and reload my msq.

For some reason I also had to go back and identify the LC-1 wide band because it wasn't reading it at all. Didn't expect that.

So should I expect that the downloader won't work on the old laptop?

And is the Injector Test Mode that unstable that I should stay away from it?

I really wanted to contribute to the info for the injector parameters.

Vicoor,

I'm glad you got it runnng. I seem to recall this happening before (a long time ago). I am also seeing it on my bench. If I recall correctly, the work around was to:

- Set the test mode to normal (and hit tab to leave that field) AFTER the test has finished,
- burn that value to the ECU,
- Save an MSQ,
- Cycle the power,
- Reload the MSQ you just saved,
- 'send and burn' values when prompted.

That should work and the normal operation of the MegaSquirt should resume. I haven't tried it yet, but will shortly. <edit>I tried it, and this seems to work for me</edit>

BTW, you may have to calibrate your TPS as well, but I'm not sure about that...

Lance.

When I loaded the MSQ after the reflash it was in test mode. I set the values from 100/60/60 to 0/6/60 set it to normal and cycled power it came right out.

Is it likely to be a glitch in the previous s19 file?

Vicoor,

I am not sure (without digging into the details of the code). I did see this behaviour on my bench, so I don't think there was something wrong with your S19 file.

What I do suspect is that the procedure outlined above fixes things because the test mode setting is part of the MSQ, and it is stored by both MegaSquirt and MegaTune. However, it seems that sometimes MegaSquirt doesn't get the parameter set correctly.

By saving the right value ("normal") to the MSQ, then reloading the MSQ, ALL of the values are re-written, and among these is the injector test mode. That should fix things most of the time, without losing your pre-existing parameter settings.

Lance.

Thanks again for all your help.

I guess in the future when I am utilizing functions I don't usually mess with, I'll make sure I have my wife's laptop so I can fix it

I was requested by DIYAutoTune tech support to post this.

I'd noticed that as soon as I turn the key to energize the megasquirt that both banks of injectors click (fire) every 2-4 seconds even with the PWM current limit set down to 25 and voltage correction turned to 0.00. The first LED indicating the injectors are triggering is also flashing. I did do a reboot after making changes to it as well. I am not using resistors, just PWM.

Oddly, my injectors were stuck in test mode. When I went back the next day upon powering on it prompted for controller or laptop revision, controller indicated test mode. I turned test mode off and after recycling power to the megasquirt it prompted once again and that the controller was set to test mode.

After unsuccessfully turning test mode off I decided to reflash my Megasquirt again with 2.891 and the injector clicking went away.

I understand this may have been an issue in previous revisions, it may still hold true to this revision then. The version causing the clicking was already 2.891; I simply re-flashed over the existing hoping it would take.

My Megasquirt 2 is utilizing an EDIS ignition and #90 low impendence injectors.

I fixed injector test mode in code 3.4x for the sequencer, and there is a change that is not in 2.89 code that could possibly cause a problem, but it would be freakish if it happened. What can easily happen is a burn doesn't go through putting the processor in a bad state, or the required sequence to get out of Test Mode is not followed.

Anyway, attached is a new code version (2.90) and corresponding ini file. The only changes from 2.891 were the Inj Test Mode fix and 2 that affect only MAF configurations.The ini file requires the use of TS - it will not work with MT. The only reason is a table generator program was added to allow loading new maf tables and MT won't recognize this program.

Lastly, I attached an instruction file for using Injector Test Mode.