Idle & Startup Strategies, Min Injector Pulsewidths

[markocosic]

Idle & Startup Strategies, Min Injector Pulsewidths

Relevant posts:
http://fsae.com/eve/forums/a/tpc/f/125607348/m/71210980331/p/1
http://fsae.com/eve/forums/a/tpc/f/125607348/m/81360673...056067355#9056067355 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/8136067355?r=9056067355#9056067355)
http://fsae.com/eve/forums/a/tpc/f/125607348/m/87660859...776017903#3776017903 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/8766085903?r=3776017903#3776017903)
http://fsae.com/eve/forums/a/tpc/f/125607348/m/35910386...10427331#65910427331 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/35910386331?r=65910427331#65910427331)

Idle control strategy - no sketchy electrical practices, no vacuum leaks and building a "hole" into the map seem to be what most reccommend and it fits with OEM practice. The electrical guy (Hi Dave!) know what he's doing, but I have a couple of other questions: Vacuum leaks - how long to stall your engines from ~2500rpm if you seal the intake? Holes in the maps - thanks to Frank for posting how many rpms theirs covers, but what do you do about startup conditions and these holes? (do you bypass them somehow to try and get the thing to rev past then settle or are they left in place and it revs up to them and settles?)

Startup strategy - what happens at startup? Anybody willing to share starter-motor cranking rpm and battery terminal voltage during cranking for a 600cc/4cyl? Do you fuel continually as you would if the engine were 'running' at 250rpm, or do you alternately fire bucketloads of fuel then nothing into the engine or priming pulse then ordinary pulsewidths in an attempt to get it to fire? Is there a rpm threshold from startup-running maps or is it something else, such as a time delay? (I've seen all these on production cars and they seem to work, but fsae machinery seems to do everything from burst into life to cranks for 2 mintues then go...) What's the difference between a cold start and a hot restart and what are you using as the threshold for deciding - coolant temp/intake air temp?

R6 Injector Pulsewidths - stock yamaha injectors on an MY05 engine, run either batch-fired as a four or batch fired as two pairs. Anybody willing to share experience with minimum injector pulsewidths and idle (at what rpm) pulsewidths?

[markocosic]

And this to be added to the definitely relevant:

http://fsae.com/eve/forums/a/tpc/f/125607348/m/19910425931

500-750rpm? Crikey! Similar story elsewhere?

[KU_Racing]

if you are trying to set your idle speed at 250 RPM, you are going to have serious trouble- shoot for 700-900 RPM. fueling should not be that difficult.. fuel the engine for close to stoich operation and then use timing to change the idle point.

I think the F4i cranks at about 600 RPM.

[Poe]

All our F4i's crank right at 300 rpm...

[Pete M]

Our F4 tends to crank at about 350 rpm, but this depends on battery voltage. A fairly flat battery will spin it at about 250 rpm.

Ignition timing is a relatively easy way to force your engine to idle at a lower rpm than it normally would. A word of advice if you do this though is to build a wall of higher timing below your idle rpm. The idea is that if the engine starts to drop from idle, power will increase and it'll rev back up to the idle rpm. There are other methods of controlling idle, depending on the capabilities of your ECU. My understanding is that any form of idle control valve is illegal though.

The motec's strategy for starting is to fire a single larger pulse of fuel on the first firing, and then fire smaller pulses from then on. The amount of fuel it injects is a mix of the base fuel map at that site, engine temp compensation and cranking compensation. This will be different for different ECUs, but the general idea is pretty common. Basically, the idea is that you tune the base map until it starts hot nicely. Then you tune the engine temp comp so that it runs relatively smoothly at cold engine temperatures. Then you mess with the cranking compensation to make it start quick cold. This is a bit iterative when you do it the first time. The important thing though is to make sure you tune the right table. Ie, don't tune the base fuel map if the ET comp table is throwing in extra fuel because the engine is cold. The biggest mistake some teams make here is tuning their base fuel map for a cold start and then not having it start well hot, such as during the driver change of the enduro.

[markocosic]

Originally posted by Pete M:
Our F4 tends to crank at about 350 rpm, but this depends on battery voltage. A fairly flat battery will spin it at about 250 rpm.



Cheers both for those numbers


base fuel map at that site, engine temp compensation and cranking compensation <snip> the biggest mistake some teams make here is tuning their base fuel map for a cold start and then not having it start well hot, such as during the driver change of the enduro

Agreed. What temperature do people consider to be fully hot/where to you usually end the engine temp compensation? 'As hot as it ever gets' it easy, but how far down do you then bring that threshold to ensure that an engine that got hot then sat for a minute or two but is still essentially hot doesn't get cold-start enrichments? (or - roughly where should we find our cold-start mapping ending - 60C, 70C?)

[Dallas Blake]

Our temperature based startup enrichments linearize to zero at 80 degrees and our warm up enrichments zero out at 60 degrees

[Pete M]

We have some start enrichment (limited duration after start) at all engine temps, but more at colder temps and it decreases with time. Our engine temp comp (always applies) ends about 50-60 degrees C. We haven't really spent as much time on this as we should though. It starts quickly every time and is reasonably strong when cold, so we've haven't paid much attention to it recently.

[Dan Nauts]

As far as idle strategy, I've tried the ignition timing "hole" as well as a hole with fuel pulsewidth,... I've done well to get either (or both) to work, but never as perfect as I'd like.

I did, however notice something very interesting while first working on idle stability; with the very low throttle plate angle required to get the restricted motor to idle, coming off-throttle and back to the idle position, there was enough deflection in our thottle plate to block off air flow past the plate. An easy check for this is to pull your filter, get to a stable idle (warmed up, higher RPM or otherwise), and lightly push on the throttle plate in the "closing" direction. If there's a noticible drop in idle RPM, you have compliance issues.

A definite work-around (that worked great for us) was to put not so much of a "hole" at idle, but more of a flat-spot,... minimizing timing and fuel to the point where the motor would load-up if idling long enough, and bump up the idle RPM (1500 or so) to get a little more throttle plate angle. Result: No more stalling at 0% throttle from high revs, and a very stable idle due to no fuel and timing changes over a small RPM/TPS range (800-1600RPM & 0-7%).

This year I'm trying a much smaller throttle plate in a much, much higher quality throttle-body, and so far compliance doesn't seem to be an issue. We're also trying an idle bypass around the throttle plate that's either going to use a valve or an orifice to meter idle flow (zero chance of compliance trouble).

Cold start strategy is easy for us,... we treat everthing as a cold start. Granted, there's more cold-start compensation during the first few seconds of startup at lower temps than at higher temps.

Everthing less than 1000 RPM is a starting attempt and gets a 3.0ms injector priming pulse and a 180-150% bump in fuel from 0-80 degF. NOTE: the "1000 RPM" is also key to how I've solved the idle stability problem,... if the idle falls below 1000 RPM at 0% throttle, the ECU thinks there's a startup attempt and will send a 3.0ms pulse to the injectors, and run 150% more fuel for a couple seconds, "bumping" the RPM back up to just above idle RPM, and gently stabilizing.

Aside from cold-start adjustment, water-temp compensation is linear from about 190% at 0 degF to 100% of the fuel-map at 160 degF.

FWIW, I'm disturbed that I can tell you what temperature our engine is idling at just by listening to it,... I've spent way too much time on cold-starts.

[Dan Nauts]

Oh yeah, pulsewidth is about 1.44ms on a 19lb/hr injector at idle, 1.56ms on a 14lb/hr.

[markocosic]

Thanks very much Dan!

That throttle body compliance could be a well sneaky one without the heads up. I think we'll be ok on that front - the spun alloy Jenvey air-horns/bellmouths/velocity stacks are fairly naff as far as tolerance goes, but their throttle bodies are very nicely made indeed, if on the heavy side.

Flatspot above and a high 'starting' threshold/priming pulse noted - we'll see what strategy works best. Numbers on the water temperature compensations useful!

What fuels (density/volume) are you running for those pulsewdiths BTW?

[Chris Boyden]

TWe're also trying an idle bypass around the throttle plate that's either going to use a valve or an orifice to meter idle flow (zero chance of compliance trouble).


The idle bypass is a great idea. If you look at the F4i throttle body, it has a mildly complicated needle valve idle bypass assembly that is controlled by some sort of temperature controlled (coolant) servo. I have always found that the engine needs more air at cold temps to balance the additional fuel at cold temps. If you run a small diameter tube to the cockpit with a small valve on the end, the driver could close it when the engine is warm (man in the loop).
Think poor mans idle air control. A little better control could take the form of a normally closed solenoid valve that opens at cold temps. Even better is idle air control.

I figure if the engineers at Honda resorted to air control vs. timing/fuel walls, dips, etc..
then it couldn't be too far off.

I always found that cracking the throttle when the engine was cold to be much more reliable and easier than screwing with the map from here till kingdom come.

[markocosic]

With reference to the 29/04/07 post here:
http://fsae.com/eve/forums?a=tpc&s=763607348&f=12560734...10053041#33710053041 (http://fsae.com/eve/forums?a=tpc&s=763607348&f=125607348&m=76310620041&r=33710053041#33710053041)

We'd second the idle bypass screw idea, and Yamaha apear to have a really oddball solution to idle-control/emissions-control...

[Chris Boyden]

an old 37mm keihn slide carb had a nice little
idle bypass valve with detents. It even had a pull out choke mechanism built into the valve assembly(not needed, but would close off the bypass). Pretty slick.. I was too worried about the throttle sticking open to use it, though. I did cut the bowl off and prep it for use as a slide throttle. How many teams use a slide carb or something besides a butterfly?

Hell, know that I think about, I should have just ripped off that coolant servo and have that control the idle, and once the engine is warm, it'll close the idle circuit down.