MS2 on an LTR450

[powder]

Currently trying to get MS2 (v3.01 daughterboard) working on a v3.0 board, using MS2 code v2.687.

The engine is a 2006 Suzuki LTR-450, a single cylinder, 4-stroke engine. It uses a 17+1 wheel, meaning it has 17 normal teeth and 1 long tooth. I am taking that long tooth off to be able to run it as an 18-1 wheel. The sensor is above the rising edge of the long tooth/removed tooth at TDC.

I have a couple of questions regarding settings in the MSQ.

Do I still have to run this engine as a 2 cylinder, and simply not attach a second spark plug? The Megamanual suggests this, but that section doesn’t appear to have been updated in a long while.

As far as trigger settings, I believe I would choose:
M = 18
N = 1
Skip = ?
Delay = 0
I don’t have a good understanding of what skip teeth would be given that I am running a single. 2*M/1 = 36, but to have skip>M seems odd. Any ideas?

Thank you.

[grippo]

Skip teeth is the number of teeth between spark firngs. In your case of 1 cylinder wasted it would be 18 teeth. Delay teeth is primarily for dual spark mode (in microsquirt), so leave it as 0.

Question: are you going to run it in Dual spark mode ? if so update to 2.883j code and make the modifications (not easy) described in www.microsquirt.info If not, then there will only be one ignition output and the rpm should come out correct. If it doesn't, the ini file sacale factor for rpm can be changed.

[old guy]

Correct me if i am wrong, but I thought delay teeth determined which tooth after the missing tooth that the ignition fired?

[powder]

Question: are you going to run it in Dual spark mode ? if so update to 2.883j code and make the modifications (not easy) described in www.microsquirt.info If not, then there will only be one ignition output and the rpm should come out correct. If it doesn't, the ini file sacale factor for rpm can be changed.

No I was not planning on using dual spark mode. I was planning/hoping to run it as a single cylinder, with wasted spark, off the crank position sensor, using direct coil control via the VB921. Is this possible?




PS. I previously tried updating to the 2.883j code, and while I was able to put that code on the board, after setting the ini, Megatune insisted it was expecting 2.830, and that since 2.883j was higher, the tables would be corrupted. I was using the most recently posted 2.8x ini file, from February, so that confused me a bit and I reverted to 2.687.

Thank you.

[grippo]

Correct me if i am wrong, but I thought delay teeth determined which tooth after the missing tooth that the ignition fired?

No I was not planning on using dual spark mode. I was planning/hoping to run it as a single cylinder, with wasted spark, off the crank position sensor, using direct coil control via the VB921. Is this possible?


PS. I previously tried updating to the 2.883j code, and while I was able to put that code on the board, after setting the ini, Megatune insisted it was expecting 2.830, and that since 2.883j was higher, the tables would be corrupted. I was using the most recently posted 2.8x ini file, from February, so that confused me a bit and I reverted to 2.687.

That is what delay teeth does, but you can get the same effect with the trigger offset. In dual spark you don't have as much flexibility with trigger offset, making delay teeth sometimes necessary. But you can use it if you want -

It is possible to run your engine with MS2 in the manner you describe, but there was a problem with missed outputs when starting it. I haven't yet had time to solve this, but I don't know if you will encounter it, especially since you have a high count toothed wheel.

The ini that Lance recently released should work for all 2.883 codes. Are you sure the ini is in the correct directory ?

[powder]

Ok, I got 2.883j loaded via the new ini file, posted on Feb. 15th. I also installed the new "settings" file, to fix the "n2o" error. Megatune now shows the new menus associated with 2.8 ini and ms2 settings, etc. All good news.

SUPER-QUICK NINJA-EDIT! (Fixed - I grabbed a bad 9V battery)

[powder]

Fixed.

[powder]

Fixed.

[powder]

Edit: R52 and R56 are useful.

[powder]

Edit: Engine was way rich, yielding unpredictable results. Engine revs to 8900 today. Hope to get it to 10500 tonight, and tune load tomorrow.

[powder]

Well - the car ran. We raced on it. Car made 23hp at the wheels. Rather depressing given that many people dyno stock LTRs at mid 30s to the wheels. But not bad for a seat of the pants tune, and an intake that I jacked up. I needed a ton more plenum volume - I was off by a factor of 5x-10x. We now have an in-house dyno, so things will ramp up quickly over the next 10 months.

Several mildly serious long term issues though:

ECU was unstable - I had to turn the pots down to get it to start but then it wouldn't rev past about 7000. If I started the car, and revved it a bit, then adjusted the pots up a bit, it would rev to about 9000. We are looking at using a different filtering circuit that another university developed.

I believe that since the engine uses a stator, with magnets spinning near the crank pickup, it either has too much noise at high RPM, or the shift in signal strength from the CKP from idle to redline is too big a range for Megasquirt to deal with.

The stock computer uses differentiation to determine engine phase, then runs sequential. I haven't been on the forums recently, but last I checked MS could not do this well yet.

There are several ways to get around these issues, so far as I can tell (while retaining MS, if we so choose). We are keeping them to ourselves as other teams read these boards. Once we have successfully implemented them, we will post up.

Thank you all for your support getting this project rolling.

[Matt Cramer]

Might want to try putting a resistor in line with the VR signal - this often helps with sensors that have a higher than normal voltage. We usually use 10K or so.