Ignition dwell randomly shortening

[dnb]

I have been using a spare ECU running 1.2.28 firmware to work on testing some coils.  All was going well until I noticed that sometimes (roughly 1 in 25) test spark events have a dwell much less than I expect.

I have set the dwell to 4ms (I know it's long, but there is method in the madness) and turned off all compensations for dwell.  The ECU and coil test rig are powered with a 6A bench power supply set to 13v.  I then used VEMS tune test mode to trigger spark events and recorded output on a scope.  Most of the time the dwell shows as 4ms on the scope and I see a strong spark, but on occasion it produces a weak spark produced by a short (usually around 2.6ms) dwell.

Is this expected when operating in test mode?  Or have I found an issue?  I have a data log, but since there was no running engine it's a bit boring!  The dwell gauge doesn't indicate a problem.

Any ideas?  If this is a genuine problem, it could explain the random stalling at idle problem I see from time to time.  Thanks!

[VEMS]

Hello David,

There were some known spark related problems around 1.2.27 firmware (if i remember correctly) i would recommend upgrading to 1.2.31 which is known good.

Best regards, Dave

[dnb]

The original problem persists in 1.2.31 (ie in test mode) but doesn't appear to be a feature of the ECU when running with a crank trigger.

[NOTORIOUS VR]

The original problem persists in 1.2.31 (ie in test mode) but doesn't appear to be a feature of the ECU when running with a crank trigger.

I can confirm this, as just 2 weeks ago on 1.2.34 also still has this issue (sometimes weak spark on test mode).  I don't believe it's an issue during running (my car seems fine)