V-twin front cyl runs much hotter than rear--- Why?

[Norzilla]

thxs 24c I have seen this document, but it is for the early model engine 1999-2003 , my engine is 2004 and although similar in basic layout it is different config.
the 2004 RSV1000R Factory motor is not twin spark, does not use CDI coils and has a 36-2 wheel not 12-1, in std form it also uses a different ecu and maps etc.

There are no inspection windows for timing so checking this way is not an option unfortunately.

[Norzilla]

I am thinking after a night sleep that i should have the trigger offset set to -57.5 deg not -60 that I have been using, this is due to the TDC being 2.5 deg offset on the trailling edge of the tooth not the centre of the tooth. I suppect that I can leave the Oddfire angle at 60 deg as it will be working from the new -57.5 deg datum.

effectively I think what i have done is tweeked the timing as if i were using the timing wizard function in Tunerstudio to align the hypothetical timing light etc.

[grippo]

Can just a 2.5 deg difference in timing break a starter ?

Anyway I checked out your msq on the bench with a 36-2 simulated wheel input and everything looks ok, as best I can tell from your timing diagram. I am now testing with the 3.740 code, but before I give it to you I would like you to read the auto trigger document at http://www.megamanual.com/seq/autotrig.htm . Then tell me which of the first 3 Figures represents your case and what number would you use for the trigger offset, noting that with auto trigger this is always a + number and doesn't have the same meaning as with present code. Also I assumed that with your engine, the 2nd cylinder, the one tied to IGN2, fires 60 deg earlier(NOT later)than it would if this were an even-fire engine.

[24c]

I sort of knew about the ECU differences, as the one in the docs is Nippon Denso and uses the same connectors as my Yamaha. I didn't think the coils were CDI though, as they were switched using the negative side. I was more interested in their SD/Alpha N set up, but back on topic, you are not alone in getting the front cylinder to run hotter. I heard today through the racing grapevine that some KTMs suffered the same problem using Motec units. In fact one team lost three engines to weak running front cylinders. Motec denied it was their problem, yet stock ECU kitted bikes were OK, but when fitted with Motecs set up by themselves they died.

BTW, in this race series you had to use Motecs, as it was the control ECU.

[Norzilla]

Can just a 2.5 deg difference in timing break a starter ?

Anyway I checked out your msq on the bench with a 36-2 simulated wheel input and everything looks ok, as best I can tell from your timing diagram. I am now testing with the 3.740 code, but before I give it to you I would like you to read the auto trigger document at http://www.megamanual.com/seq/autotrig.htm . Then tell me which of the first 3 Figures represents your case and what number would you use for the trigger offset, noting that with auto trigger this is always a + number and doesn't have the same meaning as with present code. Also I assumed that with your engine, the 2nd cylinder, the one tied to IGN2, fires 60 deg earlier(NOT later)than it would if this were an even-fire engine.

Re the starter it was the starter sprag clutch that failed, though many kick backs during the tuning process, just the one a couple of days ago was the final straw that the sprag failed.

From reading the info on autotriggers , my engine is represented by figure 3 , the counter clockwise rotating crank, the offset is +57.5 ( 5 full teeth + 3/4 to get to the trailing edge of tooth 31 which is 7.5deg + 50deg). in addition, as is the must for microsquirt, i have the cam trigger falling in the gap from the missing 2 teeth. I also have the 2 cyl (IGN2) firing 60deg earlier than if it were an even fire engine.

[Bernard Fife]

Can just a 2.5 deg difference in timing break a starter ?

Norzilla,

As another, widely separated data point on this, my 406 relatively high compression small block Chevy would break the big-block style starter nose by kicking back if the cranking timing was even 1 degree more than 10 degrees BTDC. The starter had a brace on the back, and the spacing was very carefully set to at the spec. Sometimes a new starter would break on the first start, sometimes on the tenth. Back when ignition control was new to MS, I did a fair bit of experimenting, and went through a dozen starters (all broken the same way - the aluminum casting snapped in half).

I know better than to use more than 10 degrees of cranking advance now (usually I use zero, just to be sure), but an aftermarket gear-reduction mini starter works much better, spins much faster, and has given no problems at all (it seems much more robust, and I have it held in place in all three starter bolts locations).

So I am not surprised by your sprag issues, this seems about in line with my experiences.

Lance.

[grippo]

One other possibility for the starting problem is that the cranking tolerance is set to 150%. If this is what it takes to get started, then that's ok, but if not I would reduce it.

You got exactly what I got for the auto trigger setup, so that further confirms to me that your setup is correct. I ran your msq on the 3.740 code and get the same spark and fuel outputs, but there is now better control of these. I also converted your msq for the new code.

However, I have been talking to someone else with a similar 2 cylinder, odd-fire setup and he identified a problem with this configuration, and that is that the map sampling, which takes place at the same point in time on both cylinders, can give quite different results in the odd-fire case because if for example you sampled at 0 and 360 deg rel TDCC1 for cyls 1 and 2, then cyl 2 would really be getting its map 60 deg later than for cyl 1 because of the 60 deg adv odd-fire angle. In this case it would be better to just use the one sample per 720 deg rather than 2 samples. You will get more consistent map readings, and it won't hurt accel if you weight on tpsdot. If this still doesn't solve the problem, you can individually tune cyl 1 and 2 fuel pw and spark advance with the 3.7x code. So I would like to make this change, which will only affect odd-fire engine map sampling, and then give you that version. This will take a day or 2 to make the change and test it.

Map needs to be sampled at or near BDC of the intake stroke, that is where the pressure in cylinder and manifold is the same and the cylinder has a full air charge. If it isn't, you can compensate by tuning VE. But unless you have a VE table for each cylinder (imagine tuning that), it is important to sample map at a consistent point for all cylinders, even if its not at BDCI.

[Norzilla]

Thanks for the feedback on the cranking pulse tol. I will test the down scaling of the number and see the impact.

I look foward to testing the new code when you are ready.

Today I did some testing on the impact of the timing of the injection timing. Previously I had set the timing at 26% as at the time it appeared to produce the smoothest idle, that is when I was using -60 deg in the TDCC1 offset, now this appears to be about 22% now i am using -57.5 deg

The other observation is that by changing injection timing , changes the idle AFR on Cyl2 (rear) through a range from about 11.8:1 to 14:1, when going from 10% to about 40%. By below 10% the engine dies, as it does above about 50%... I am now thinking that the inj timing can have a much larger affect on the resultant AFR than first thought. This may actually be the source of the different exhaust temp. I think now that the new code could dramatically improve or at least equalise the effect of inj timing.

(NB only have WBo2 on the rear cylinder)

Below is the cam timing of my engine maybe you can calculate the ideal ranges for inj timing for the cyl1 & cyl2 for the new code?

Basically the cam timing is as follows;

Intake cam
Opens on the intake stroke 25° before TDC
Closes on the intake stroke 61° after BDC

Exhaust Cam
Opens on the Exhaust stroke 64° before BDC
Closes on the Exhaust stroke 15° after TDC

Thanks
Norzilla

[24c]

I looked at the new stuff re the later code, and I must admit I found the bit about injecting on a closed valve and avoiding the overlap period very interesting with regards to sequential timing, plus it was a different viewpoint from what I heard elsewhere. As always something new to take in every day, and also thanks to Lance for his post too

PS AFAIK The rear cylinder generally (kart inline engines) runs hotter, so you have to compensate the fueling/ignition likewise.

[Norzilla]

Grippo, I tried reducing the cranking pulse tolerance and once I start going below about 140% starting become unreliable. So I am going to leave it at 150%

Today I did some more testing around the Inj timing % these are the results:

Inj Time % AFR rear cyl Temp MAP
13% 12.1 68 Engine dies
15% 12.2 68 engine just runs but very rough
17% 12.3 64-65kpa
19% 12.5
21% 12.7
23% 12.9
25% 13.2
27% 13.4
29% 13.5 73 60-62 kpa
31% 13.5 73 60-62 kpa
33% 13.4
35% 13.2
37% 13.0
39% 12.8
41% 12.6
43% 12.4 68 64-65kpa
45% 12.2 68 engine on cusp of rough running
47% 12.0 68 Engine dies

As a result of the above testing I have moved the Inj timing on my engine to 30 which is the point with the lowest MAP, and highest AFR which actually matches the Target AFR in my AFR map for Cyl 2 ( rear cyl ), My view is that 30% is giving the most efficient burn of the fixed fueling at idle, ie giving the leanest AFR or most complete burn given the circumstances. A postive side effect of this Inj timing position is the lean out stumble i had off idle was reduced. Subsequent to that development I have seperated the fueling for front and rear cyl and have richened up the front cyl but a VE of 5 this has furhter stablised the MAP fluctuations (pulsing) at idle and created a smoother idle and not zero hesitation off idle.

I have yet to measure the exhaust temp at the exhaust port for this setting. TBA

Another interesting observation that I have made between running the engine today and yesterday is that the tune that i made at the end of the day yesterday was giving much leaner reading on the WBo2 than the same setting gave at the same clt temp after warm up this morning. Yesterday idle AFR was 13.5-13.6 this morning 11.9-12.1 The only thing i can put this variation down to is the increase fuel temp after sustained stationary running of the engine yesterday where there is a heat soak affect on the fuel, ie the Ali fuel tank allows the fuel to heat up substantially, maybe 20-30deg C above the temp experienced this morning. I think that the fuel density ( lighter hot fuel) is having a big affect on my engine. Has anyone else experienced this hot fuel effect?