Yamaha YFZ450 hard to start

[plak424]

Hey Everyone,

So this is our team's first year using fuel injection in about 3 years. We're using a PE ecu and we've been having problems starting throughout the entire time we've had the set up. We're fairly sure that we've gotten the timing down and the fuel map set, but we can't seem to resolve the initial starting problems. We only have problems during the initial starting, after the engine warms up its much easier to start the engine.

If anyone has any suggestions for us or some experience with problems like this any input would be greatly appreciated.

Matt
University of Hartford Team Captain

[plak424]

Hey Everyone,

So this is our team's first year using fuel injection in about 3 years. We're using a PE ecu and we've been having problems starting throughout the entire time we've had the set up. We're fairly sure that we've gotten the timing down and the fuel map set, but we can't seem to resolve the initial starting problems. We only have problems during the initial starting, after the engine warms up its much easier to start the engine.

If anyone has any suggestions for us or some experience with problems like this any input would be greatly appreciated.

Matt
University of Hartford Team Captain

[coastertrav]

We just went through and messed with the cold start enrichment and ignition tables in the Haltech we use. We also use a Suzuki gsxr 600, but tuning should be similar regardless of the motor for what we're dealing with.

Get to those tables, and only manipulate them within the cell you are currently in (they are an overall fuel or ignition trim based on only coolant temp). Start with a cold motor, get it to start two or three times reliably without warming up, then let it warm up into the next cell, and do it again till all cells have been done.

[Adambomb]

The first year we used a YFZ450 we also used a PE computer, never did get it running properly. Best reason I heard was because the PE is set up for a 12-1 crank trigger batch fire setup, and with the abnormally high angular acceleration experienced by a single, with that few teeth there was huge issues turning a dt into a stable rpm. The missing tooth made it worse. I know PE has come out with a new computer, perhaps it addresses this issue. PE also has really good FSAE support (expecting a reply from the factory rep... http://fsae.com/groupee_common/emoticons/icon_wink.gif ).

[B Lewis @ PE Engine Management]

Hi Guys,

Yes, we do have a new ECU (The EDGE) that works especially well with the YFZ450 engine. We have a ton of pro guys using the ECU with that engine on the GNCC circuit where starting from a dead stop is very important. There is no reason why you shouldn't be able to get it to start very well. Send me your tuning file and I will take a look at it.

[plak424]

Thanks guys, and Brian I'll send you my tuning file later today. Also I was wondering if the ecu we got was the Beta for the edge, I think its the PE-3 beta.

Thanks,

Matt

[Mbirt]

Matt,

To cold start our Megasquirt 3 CRF450X with a 24-0 crank and 1 cam pulse setup, I have to cycle the ECU on and off several times to make the injector fire a corresponding number of priming pulses. If you've configured the PE to fire a priming pulse, it might be worth a shot.

Adam,

We found improvement going from 24-2 to 24-0+1 cam tooth. Since MS needs to receive that cam pulse between two crank teeth, we've moved to 12+1 for 2011. Using your logic, this may have been a bad move. Would you be willing to divulge what trigger scheme ISU switched to? I know I saw a picture of the team dialing in a cam position sensor.

Brian,

What trigger scheme are the GNCC guys using that works so well for them?

[B Lewis @ PE Engine Management]

The GNCC guys are all starting with a stock fuel injected quad, pitching the OEM ECU and connecting ours to the stock wire harness. The stock trigger wheel is also a 12-1 tooth pattern but it has a little engineering behind the placement of the missing tooth. It is in a location that makes it less affected by the acceleration of the crankshaft. They start very well.

[Mbirt]

Thanks, Brian! The guys on the megasquirt forums suggested I do the same, moving the missing tooth to a location of steady rotational speed. It had previously been near TDC. The trigger logs looked better after the clocking of the flywheel but I was still receiving the same "Sync Loss Reason 2" (missing tooth detected at wrong time). With this in mind in addition to my competitive benchmarking showing there were no single cylinder teams running wasted spark other than us, I added a cam position sensor and teeth where the missing teeth once were. Starting is now much improved.

Would I be correct in assuming that the EDGE samples for the lowest MAP value every two revolutions?

[B Lewis @ PE Engine Management]

Most of the stock quads with fuel injection are wasted spark. Based on our experiments with a stock YFZ450R it is one of the few sequential setups on a single but it uses the MAP sensor to identify the revolution within the cycle. We are planning to implment something similar in future software releases. I believe this is how the stock Aprilia works as well.

[Adambomb]

Mbirt:
Don't remember off hand, thinking it was in the ballpark of 32 or 36 or possibly 40ish -0, with cam sensor. We were running a Motec, and likewise with its timing scheme the cam tooth should be dead center between two crank tooth edges, which can be a real pain to get dialed in within a reasonable level of accuracy when you have a lot of teeth; i.e. if you've only got 7.5 deg. between edges then just getting it +/- 1 deg. may not be close enough. This of course becomes a problem when your method of dialing in the cam sensor involves an angle finder stuck on a breaker bar in a parking lot in Texas in July (this is also coming from the guy who used the "angle" app on one of the guys' iPhones to align the car at comp http://fsae.com/groupee_common/emoticons/icon_wink.gif ). Also if you have this setup I strongly recommend paying a bit of extra attention to what exactly it takes to adjust the cam sensor, with ours I remember we had to remove the cam cover gasket and shim the cam cover to height with precisely two U.S. dimes. Then one time a dime slipped a bit too far inward and underwent a stamping operation via one of the cam lobes. I died a little inside when I saw that. Luckily dimes are fairly soft and there didn't appear to be any real damage. I remember another ECU I heard about that used MAP for phasing, after dealing with these types of shenanigans that sounds pretty HOTTT. I have really come to hate adjustable cam sensors.

Of course I've been away for the last year, I know they've since switched ECUs, not sure what setup they're using now.

+1 for getting the phasing tooth off TDC. I know conventional wisdom would dictate it "should" be there, makes it easier to get dialed in because you can just align the stock timing marks to TDC and then install the wheel so it triggers right there and voila your offset is 0, but it does seem a whole lot less finicky when you get it away from TDC.

[B Lewis @ PE Engine Management]

My experience is also that modified singles are a little harder to get started. Once the tune is good they are still pretty reliable. Generally we find that they want a bunch of fuel to get going.

[coastertrav]

Well then, I had no idea of the physical difficulties syncing the ecu to engine rpm and stated that I was only familiar with a 4 cyl application.

I just assumed it was a simple tuning issue and tried to offer help with my best guess.

[Adambomb]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by coastertrav:
Well then, I had no idea of the physical difficulties syncing the ecu to engine rpm and stated that I was only familiar with a 4 cyl application.

I just assumed it was a simple tuning issue and tried to offer help with my best guess. </div></BLOCKQUOTE>

Yeah, that was one of the surprises we got when we switched to a single, it seems much more sensitive in starting/low-rpm conditions, I'm guessing because again, you only have one powerstroke per two revolutions, and these engines have relatively light flywheels. Can't cope with the modifications like big plenums as well. There are actually motocross guys that switch to heavier flywheels.

[Mbirt]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Adambomb:
Yeah, that was one of the surprises we got when we switched to a single, it seems much more sensitive in starting/low-rpm conditions, I'm guessing because again, you only have one powerstroke per two revolutions, and these engines have relatively light flywheels. Can't cope with the modifications like big plenums as well. There are actually motocross guys that switch to heavier flywheels. </div></BLOCKQUOTE>

Yet another reason to go with a quad thumper instead of a dirtbike one. I'm hoping the TRX's increased flywheel weight, larger starter, and lack of factory kick starter equate to added reliability in starting for us over that of the CRF.

Some single-related observations from comp: McGill showed up without any plenum volume whatsoever. My dyno testing shows this puts them in the 20-30hp range with their Rotax 450. Their 64th place in accel was followed by 8th place skidpad and 26th place auto-x finishes. Sacrificing some engine development for pre-comp test time appears to have worked here.

GFR runs only a high-mount injector and no low-mount injector. Correct me if I'm wrong here.

UAS Amberg-Weiden's big 610cc KTM thumper (downsized from 690cc) makes 63hp/46 ft-lbs (with a mighty trick variable length intake runner), yet won the economy event by .07 gal over GFR. I'm sure GFR's pace could justify the extra fuel used, but I was impressed by Amberg-Weiden's powertrain.

[B Lewis @ PE Engine Management]

I am glad to see McGill do well this year. There single seemed to start very well without running sequential. They just used the stock crank setup like on the quad.

[bob.paasch]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mbirt:
GFR runs only a high-mount injector and no low-mount injector. Correct me if I'm wrong here. </div></BLOCKQUOTE>

Our injector sits in a well built into the plenum. I'd call it a low mount injector. Delft used to run two injectors, both a high and a low, and had algorithms for transition between the two.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">UAS Amberg-Weiden's big 610cc KTM thumper (downsized from 690cc) makes 63hp/46 ft-lbs (with a mighty trick variable length intake runner), yet won the economy event by .07 gal over GFR. I'm sure GFR's pace could justify the extra fuel used, but I was impressed by Amberg-Weiden's powertrain. </div></BLOCKQUOTE>

We ended up with a kink in our sight tube after endurance, and the fuel folks overfilled our tank enough to spill over the top and down into our chassis.

With 42 HP, GFR had the fastest lap. Think about that.... http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Mbirt]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by bob.paasch:
Our injector sits in a well built into the plenum. I'd call it a low mount injector. Delft used to run two injectors, both a high and a low, and had algorithms for transition between the two. </div></BLOCKQUOTE>
But it's aligned axially with the intake runner, spraying into the bellmouth, right? This is a classic high-mount. Delft was running E85 and might've been spraying upstream for a cooling effect.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">We ended up with a kink in our sight tube after endurance, and the fuel folks overfilled our tank enough to spill over the top and down into our chassis.

With 42 HP, GFR had the fastest lap. Think about that.... http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif </div></BLOCKQUOTE>
The post-endurance refill is incredibly nerve-wracking when you're gunning for the top economy spot, isn't it? We're coming for you guys in Cali (especially if you bring the big aero package) http://fsae.com/groupee_common/emoticons/icon_smile.gif

That 42 HP better be at the wheels with the slicks aired down. I was told 48 hp at the crank by one of your engine guys. With your Hot Cams Stage 2 cam, crf450r head gasket, intake tract tuned for the 3rd ramming wave in the 9-10k rpm range, plenty of plenum volume, high-mount injector, and short exhaust head pipe length, I'd infer that you're in the 50's. Numbers aside, what GFR is doing is working.

The GFR FTD did give me ammunition against the underclassmen who want to "boost the thumper" next year. See how much it helped Penn State? And Wisconsin could have placed just as well without a turbo and its 20 extra ponies, not to mention the harrowing smokey re-entry after their driver change. I was afraid they wouldn't be allowed on course for about 30 suspenseful seconds.

[bob.paasch]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mbirt:
That 42 HP better be at the wheels with the slicks aired down. I was told 48 hp at the crank by one of your engine guys. With your Hot Cams Stage 2 cam, crf450r head gasket, intake tract tuned for the 3rd ramming wave in the 9-10k rpm range, plenty of plenum volume, high-mount injector, and short exhaust head pipe length, I'd infer that you're in the 50's. Numbers aside, what GFR is doing is working. </div></BLOCKQUOTE>

48 HP for acceleration and autocross, 42 HP for endurance. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Mbirt]

I'd call it throttle-induced EGR and put a "green" spin on it in design.