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

Hi chaps

I'm finding it very difficult to get my V12 to start when hot as well as it does when cold. I have found the following.

My injector opening time is about 1.4ms.
When hot, it requires about 1.5ms during cranking to start fairly OK. When cranking pulses get to 1.6ms, it effectively floods.
This requires me to open throttle to beyond 70% to activate flood clear mode and then it starts (with a lovely roar of course .

When the engine temp increases marginally (say from 78degC to 84degC) the slope of the cranking curve makes the PWs become very close or less than the opening time such that no fuel is injected (as happened to me yesterday...in heavy traffic)

I believe the issue is that MS-2 squirts with every ignition event during cranking and on the V12 this results in too many injector events while cranking such that the injector pulses have to be only marginally more than the opening time making starting difficult. This would only be an issue on V10s and V12s I presume.

What complicates matters even more is that on the WUE page, one can only enter one decimal for the PWs buckets.

A solution might be make the number of injection events during cranking be user adjustable with default being = the number of cyls?

If this is port injection, then the number of cylinders should have nothing to do with it. But either way, I am surprised that you would be so sensitive to one or two tenths ms as far as it making the difference between won't start for lack of fuel and won't start because it floods. That sounds like the injectors are way too large. With v2.68 and up you can enter 10-point tables of prime pulse, crank pulse and after start pulsewidth as a function of engine coolant temperature. So you can enter a very nonlinear curve if that helps. Certainly if you killed the priming pulse on hot restart that should make a difference without at all affecting cold start performance.

For MS2/Extra I'm looking to make the cranking pulsewidth a %age related to ReqFuel to make new installs easier.

It should be fairly simple to include a tach divider as Philip asks for. I have to agree with your suggestion as those pulsewidths are so close to the opening time.

James

Thanks for the prompt reply!

grippo wrote:If this is port injection, then the number of cylinders should have nothing to do with it.

It is port injection but it was James that told me (http://www.msefi.com/viewtopic.php?t=20 ... ight=start) that MS-2 squirts with every tach event DURING cranking. It certainly sounds like it does listening to the injectors going "crazy" during cranking. If so, the V12 squirts 6 times per engine revolution where an 8 cyl would squirt 4 times per rev not so? If we assume an open time of 0.2ms per squirt x 12, that adds up to 2.4ms of open time per engine cycle (2 revs), add the 1.4ms opening time and were close to the idle pw.

The cranking PWs entered in WUE is therefore not related to cold and warm idle PWs. This confused me big time!

grippo wrote:That sounds like the injectors are way too large.

Could be, but these are the std original injectors for the car. Idles at about 4ms and 11ms at WOT - which seems OK to me (albeit with my extremely limited experience?)

grippo wrote:With v2.68 and up you can enter 10-point tables of prime pulse, crank pulse and after start pulsewidth...

Yes, I'm very eager to move to 2.6x. Been wanting to help beta testing but too busy enjoying the car right now, BUT the 10 point table might not help me UNLESS the figures I enter will allow me to enter 2 decimals, otherwise I can only choose between 1.4 (no start), 1.5 (starts but not quickly) and 1.6 (no start = flooding). I will need 2 digit resolution between 1.4 and 1.6. Being able to half the number of squirts (assuming this is the case) during cranking would allow greater resolution on cranking PWs?

The real solution to this is to provide 8 fuel channels and you could use 6 of them and double up 2 cylinders on each. Then you would fire only 12 injectors per crank rev. The present system, during cranking, is firing 12 injectors 6 times each per crank rev - 72 squirts.

Since we are a bit away from 8 outputs, I will provide a solution in the next software release along the lines you want. Right now the code automatically squirts both banks at every tach pulse. I will mod this to only do it if no_cylinders <= 8, since I can't imagine anyone wanting to squirt both banks.

grippo wrote:I will provide a solution in the next software release along the lines you want. Right now the code automatically squirts both banks at every tach pulse. I will mod this to only do it if no_cylinders <= 8, since I can't imagine anyone wanting to squirt both banks.

Thanks Mr Grippo!
It seems no matter what I do, I can't seem to get it to start when hot as quickly as it does when cold.

PS: I don't want to be pedantic but the 72 squirts is not the real issue is it? Is it not the fact that each cyl is being squirted 6 times per revolution that is the issue?

Philip Lochner wrote: PS: I don't want to be pedantic but the 72 squirts is not the real issue is it? Is it not the fact that each cyl is being squirted 6 times per revolution that is the issue?

I think we are saying the same thing - you get 72 squirts because each cylinder is being squirted 6x. Anyway this fix will be in code v 2.686t11 which I will release to the beta list shortly.