Maximum Allowable Injector Pulsewidth

[spitnl]

Hi Guys,

I have been reading Four-Stroke Performance tuning by Graham Bell and came across the following:

My question is about the 16ms maximum he is talking about.
I have searched but have not been able to find anything regarding this 16ms maximum.
As I understand it you do not want to run more than 16ms even when you are way below the maximum duty cycle (normally 80%).

Is this something you have encountered/have had problems with.

I'm going to run a turbo-ed sequential injected 6 cylinder and according to some calculations I have done the injectors will be open for longer than 16ms at maximum torque even though pulse width would only be ~50%.
Should I worry about this and go with larger injectors?

Thanks in advance,

Fred

[Matt Cramer]

The "normal" MS3 code allows pulse widths up to 65 ms. You'll notice there is a "4X" mode to allow four times that pulse width. That was put there after testing on slow turning industrial engines with redlines in the 1,000 to 2,000 RPM range proved this was both necessary and workable for certain applications. It would appear Mr. Bell's ECU had a somewhat more limited range of pulse widths.

[ol boy]

We do injector tests for flow rate and will do a single 60,000 ms pulse (60seconds) without damaging the injector. The fuel flowing through the injector acts as a coolant.

Sent from my SM-G920V using Tapatalk

[spitnl]

Wow, you are quick.

Matt, I think Bell's concern is not the ECU's limited pulse width range but that running more than 16ms could cause injector failure.

Ol Boy, thanks that puts me a bit more at ease.

[billr]

The PW numbers in that "Bell" table are for a 2-stroke; with a 4-stroke you have twice as much time as shown. I, too, believe the injector on-time has little/no influence on its life. That assumes, of course, that you aren't doing something odd like driving a low-Z injector without proper P&H or dropping resistors.

[whittlebeast]

I use this formula to generate Max Allowable Duty Cycle

100*((120000/[RPM])-3)/(120000/[RPM])

See this thread for lots more info

http://www.msextra.com/forums/viewtopic ... 34&t=60224

Andy

[spitnl]

Billr, you're right, I hadn't noticed. I copied that from the 4-stroke book on googlebooks, but they are indeed for a 2 stroke.
All pulse widths in that table should be multiplied by 2.
That doesn't change his comment about 16ms though, unless that is subject to the same error.

I'll look later tonight in my copy, which is the latest version, if it's been rectified.

[ol boy]

Billr.. that's funny how the table is for 1 sqrt per crank rev while the rest of the article is for a sequential 4 stroke..

Sent from my SM-G920V using Tapatalk

[spitnl]

Andy, thanks, I looked at the link, very interesting and something else which you rarely see mentioned when talking about injectors.

But the problem here isn't running out of injector duty cycle, but running too large a pulse width, which could damage the injector.

[ol boy]

What's the date of the article? I imagine that back in the pre EV1 injector days (late 70's, early 80's) any large pulse could cause the field coil to break.

It seems to take several decades for bad engineering design and practices to get filtered from the normal opinion.

Sent from my SM-G920V using Tapatalk

[whittlebeast]

My gut feeling is there really is no max PW that you will ever hit on a running motor that does not hit my max DC first before you have a real problem. I have never hit it.

I use a minimum of 2ms PW and about 85% DC as a rule of thumb unless you get over 6500 rpm and then I use my formula for DC

Andy

[spitnl]

The book I got that from is 1998:
https://books.google.nl/books?id=OTOYHR ... 16&f=false

But the reprint from 2009 has the same error.
The earlier 1981 version doesn't have anything about fuel injection.

[ol boy]

A lot of refinement has happened in the last 20 years.

Sent from my SM-G920V using Tapatalk

[whittlebeast]

I first found the bust in conventional logic when I was Squirting a 2 stroke jetski and I was needing a pair of 800 cc injectors to support 120 HP at 8500 RPM.

Andy

[ol boy]

That's about right. 1.2 lb-hp/hr. The best we ever see in the small airplane motors is 0.8 to 0.9.

Sent from my SM-G920V using Tapatalk

[spitnl]

I just checked the latest edition(2012) of the book and it has the exact same error in the table.
So it makes you question whether the 16ms should have read 32ms.

Guys, thanks for the help
I will sleep better tonight.

[Matt Cramer]

Wow, you are quick.

Matt, I think Bell's concern is not the ECU's limited pulse width range but that running more than 16ms could cause injector failure.

Ol Boy, thanks that puts me a bit more at ease.

That slow turning industrial engine I was talking about that needed a 260 ms maximum pulse width typically runs for 3 months nonstop before getting any downtime for maintenance. They did not report any issues with the injectors or their peak and hold drivers (we were using Versafuelers for this test) failing even at those pulse widths.