Incrementing rpm signal!?

grippo · Post by **grippo** » Sun Sep 23, 2007 5:36 am

I wrote a signal generator that ramps rpm from 100 to 600 rpm, up and down, over and over and I tried my msq and your exact msq from 20 sept. Both give consistent timing on the scope in the sense that it just moved a reasonable amount that I assumed matched the changes in table values. I didn't check absolute timing because that is easy to fix once we have stable timing - just move trigger offset. However, to check for consistent timing you have to do two things. First you have to scope the tach input to the processor rather than the VR sensor, because you can't tell from the sensor where the processor will put the square wave which is the real timing reference. The zero crossings on your setup are very difficult to detect because there are long periods where it stays very near zero, so it is not a good signal - you want a signal that crosses the axis at a sharp near vertical angle. The second thing you need to do is fill the spark table in the region you operate with a fixed spark angle everywhere. Then the output timing should remain fixed relative to the input timing and this will be easy to see with square waves on both top and bottom.

As far as differences in our msq, I have 6 skip teeth, 2 cylinders and so do you - this is the proper setup for your case. However, the dwell could make a difference, although I did try your msq exactly as you sent it and it worked.

The Benz Master · Post by **The Benz Master** » Sun Sep 23, 2007 8:26 pm

Well... compare the IRQ with spark output with every timing related factor/table set to zero and i still get the same thing as i last posted but with a square wave instead of a AC signal. And i still cant get spark constantly and if i dont put it at Delay teeth = 1, i cant get both sparks constantly (explosion and wasted)

And what is mapDOT ? it seems like that the first datalog i did, every time a deltaT was detected (one out of 4 or 5) it was followed by a highly negative mapdot value. Could it be the cause or the mapdot is just drived by delataT so it's normal it will go way off if deltaT is wrong/missing?

I really need the bike running this week so i'll try single tooth set-up as soon as i can.

grippo · Post by **grippo** » Mon Sep 24, 2007 6:33 am

From looking at the delta_ts in your datalogs, it's possible that you need to set the pulse tolerances higher. A 1 cylinder is a lot more erratic than a 4 cyl, but cranking speed is higher because it is so light. I can't simulate this on the bench, but that may be the reason we are getting different results. If you haven't already taken apart the engine, I would try raising the 3 pulse tolerances (default 70, 50, 25) to all be 70 and see if that works.

Mapdot only affects acceleration enrichment so that's not the problem.

The Benz Master · Post by **The Benz Master** » Mon Sep 24, 2007 9:51 am

P.S. if i pull the throttle while cranking, engine cranks faster and i stop receiving any rpm signal to MS (VR signal and IRQ is still ok tho)... what could cause this? I wonder about the Max cranking RPM value!?

I was going to take it appart tonight... i'll thry this before i go on.

thanks

grippo · Post by **grippo** » Mon Sep 24, 2007 12:41 pm

Since you have specified trigger rise, then when you are below cranking rpm, it should fire spark immediately whenever the first tooth is sensed - or second tooth if delay_tooth = 1.. Once you go above crank speed it goes to your spark table to get the advance and waits until that many degress before sparking. But the pulse tolerance is also changed as you go from below to above cranking rpm. Then when after start enrichment is over, it goes to normal pulse tolerance. cranking rpm also affects fuel in that all injections are simultaneous at a specified cranking pulsewidth when below cranking and then switches to VE table values to calculate PW when above cranking.

The Benz Master · Post by **The Benz Master** » Tue Sep 25, 2007 12:23 am

Bad news, raising pulse tollerance did not work

Good news, did not have to remove 6-1 wheel, it does not interfere with stock trigger location so it only took me 1.5 hour to set it back to stock 1 tooth.

Bad news, still the same problem. mostly get one out of 2 ignition event i should have! timing seems beter with this set-up tho.

The Benz Master · Post by **The Benz Master** » Tue Sep 25, 2007 10:47 pm

Now got a nice spark at tdc with an external CDI ignition module from a honda trail 70 !!

But I did not get a single rpm count today, even using the msq from yerterday that effectively yielded some rpm!? starting to think i have a problem on my MS2 board or something like this... it cant be!! i disconected the MSD, plugged my other module and checked if square wave was still ok at the irq and it was.

So how could i set-up MS to use the 1 tooth trigger and give me a reliable rpm so i can get my fueling going on? 1 square wave signal = one engine turn, that's not really complicated... i can set the box to any cylinder count as long as i have the right CC's to control injection. The CDI box has to run on a single tooth trigger so i cant go back to the 6 tooth 6 cylinder set-up which is the only one who was working perfectly.

I tried to fool MS into having a 36 tooth wheel.... 36-35 that is with skip teeth at 9 and cylinder count at 4 and still nothing. also tried at 4-3 , 2-1 and 1-0...

grippo · Post by **grippo** » Wed Sep 26, 2007 4:21 am

If you are going to use 1 tooth per crank rev then set it up with No_Teeth = 0 to take it out of wheel decode mode. Then set it up as 2 cylinder so the rpm will read correctly.

The Benz Master · Post by **The Benz Master** » Wed Sep 26, 2007 6:29 pm

Making progress !!!!

No_Teeth = 0

Tham.. that's what i was missing here!! thanks grippo!

Setting to 1 cyl and two stroke gives good results for rpm too.

Ok... my CDI box seems to trigger from leading edge of the crank tooth so i cant use this because the tooth is 60 degree long (or was it 45 !?? it's too long anyway) see drawing below.

So i pluged back the MSD on MS and it's beter than the CDI.. i get it to start (reves only 3 to 10 turns before dying) but spark is still like it was with the 6-1 wheel but rpm is constant, once in a while i get one with strange timing that mostly make the engine kick back. and spark is mostly only on the ignition event... wasted one is there only some times (only crank sensor so it cant tell which spark is ign and which is wasted so i guess it's just luck).

Spark does not seem to be affected by any advance/retard numbers... trigger offset or table driven!?

I tried to get the crank signal from leading to trailing and raising to falling without sucess in changing the spark output timing.

Ok, so i got a few datalogs while running (if we can consider this running ) and the only thing weird i noticed is that map never gets bellow 80...

grippo · Post by **grippo** » Thu Sep 27, 2007 7:18 am

Trigger offset for this setup should be positive, somewhere around 0-10 deg should work. Set trigger rise with a cranking rpm of about 300. You seem to be getting decent rpm, but the advances are all -20 something deg which is probably not what you want. It looks like the rpm is ok, so the output from the MSII that goes to the coil driver should be correct. In trigger rise mode, the spark table is ignored if rpm is below cranking. If you can keep it running with a single spark advance like 10 or 15 deg, then you could set crank speed high, like 1000 or 2000 rpm, so then it will stay in trigger rise mode and fire every time it gets a pulse from the crank wheel. There is no logic to go wrong in this. Changing trigger offset will change the advance in MT, but it should not affect when the spark occurs - it should be right after the wheel edge is detected. Then you can time it and set trigger offset to make MT match the timing light. Then you should be able to reduce the cranking speed input to 300 or so and the table advance should take over.

If this doesn't work in trigger rise with high crank speed and you see on the scope that the square wave input to MSII always comes in, but the spark command direct from MS II (again a square wave) never or sometimes does not follow, then I would guess that there is something wrong with the processor. I can then send you a new MS II daughterboard to test.- just email me off line with your address.

The Benz Master · Post by **The Benz Master** » Thu Sep 27, 2007 9:50 am

grippo wrote:Changing trigger offset will change the advance in MT, but it should not affect when the spark occurs - it should be right after the wheel edge is detected.

Exactly.. this is my problem... this sends the spark 60 BTDC cause my tooth begins 60 btdc and ends at tdc. So i guess i should set cranking speed to 100 so MS takes into account the triger offset to set the spark where it should? I deliberately shifted the table to have neg timing to cope with the sensor's 60deg advance cuz trigger offset was not doing anything to the spark output(neither did the table changes but this is ok according to what you are telling me here). Details like these should be included in the manual so people might have a beter idea of what's going on.

grippo wrote:Then you can time it and set trigger offset to make MT match the timing light. .

Second problem, flywheel with crank triger and timing mark is in the engine crankcase (filled with oil)... so when i remove the timing plug and try to start the bike, oil squirts about 10 foot in front of the bike

So i'll have to rely on tooth position to set timing what ever the real offset is.

Slowly making my way to sucess.... i hope tuning will be faster/ more straight forward.

going to test the cranking rpm to take account of the trigger offset and let you know how it turns out. I tought of maybe reducing my tooth's lenght to 10deg to set the timing back exactly where it should if trigger offset does not work.

P.S. trigger rise or trigger return generated the same results on my scope... but i think i know why... the square wave to IRQ looks like this

So the leading edge is maybe just 5 to 10 deg offset from the last pulses trailing edge instead of beign related to the width of the tooth.

The Benz Master · Post by **The Benz Master** » Thu Sep 27, 2007 10:27 pm

Ok, tested lowering the cranking rpm to 100... i effectively got some timing changes on the scope but it does not change as much as i do in MT. I still have weird ignition events at about 180deg and stuf like that.

i've been able to run the engine a full two seconds...lol
heres the log and msq.

So, do you think i should narrow my trigger down to 5 or 10 deg long just before tdc?? or should i try MS1 or something llike this!? I contacted a guy who did an install just like mine and said he had to switch to MS1 processor to get it to run properly!? Will MS1 have all the same capabilities to run my engine (beside pwm which i dont need on my bike anyway...) or do i need MSnS or extra code to run ignition with MS1 processor?? would keep MS2 for my car when i'm done with the bike.

thanks

grippo · Post by **grippo** » Fri Sep 28, 2007 10:42 am

I don't think this is a software failure since it is failing in 2 totally different modes with and without a wheel. To help I have to have a yes / no answer to this question:

If you set cranking rpm to 2000 with trigger rise and you have the scope on the tach input (square wave) and the other probe on the processor output for spark (there is only the ONE for Non-Dual Spark) then do the tach and spark square waves match up in the sense that for every tach you always have one and only one spark and the spark always starts immediately after one of the tach pulse edges. If it is the wrong edge you have to change one or both of the input or output polarities which remember is the polarity as seen by the processor. For example we know coil spark always occurs on grond, but the processor may have to hold its output high to trigger the coil through a driver, in which case the spark polarity would be high. Same with the tach input.

I will try your latest msq on the bench tonight - and if you have a stim you could try this on the bench also since there is now no missing tooth which can't be simulated by the present stimulator. If it doesn't work on the stim it won't work on the car, if it does then you can concentrate on what is different.

To answer your questions, any positive trigger offset should work in the non-Dual spark mode. I don't think that is the problem.

You can certainly try MS1, if nothing else it would be new hardware. However, you do need the extra code to get ignition. You could also try the MS2 extra code with the same hardware.

grippo · Post by **grippo** » Sat Sep 29, 2007 6:18 am

I tried your exact msq last night and it works perfectly on the stim. There is one output pulse for every input pulse and the output pulse occurs exactly at the correct degrees AFTER the input because you have all negative advances in the table. Negative makes it ATDC. I am sure you do not want this. You should adjust the trigger offset to make the timing come out correctly and then change the table values to positive. Forget what I told you about the trigger offset needing to be positive for non-wheel applications. It can be anything and I tried it for trigger offsets of +30 and -30 with table advance of 30 deg and everything comes out the way it should.

The Benz Master · Post by **The Benz Master** » Mon Oct 01, 2007 9:02 pm

grippo wrote: If you set cranking rpm to 2000 with trigger rise and you have the scope on the tach input (square wave) and the other probe on the processor output for spark (there is only the ONE for Non-Dual Spark) then do the tach and spark square waves match up in the sense that for every tach you always have one and only one spark and the spark always starts immediately after one of the tach pulse edges.

No, absolutely not but every "going high" are lined up with the IRQ (i might need the opposite to fire the MSD??)

Anyway, if they line up with the IRQ's square wave, this brings the spark 60 deg BTDC which is not good.... i need it only +- 10 deg BTDC so this is why i've been writing a negative ignition table to bring the spark farther towards TDC

IRQ at the bottom (MS2 pin 14) and ignition output (MS2 pin 17) on top

The Benz Master · Post by **The Benz Master** » Mon Oct 01, 2007 9:07 pm

P.S. i tought that the "going high" and "going low" options in megatune were to tune the output's polarity!? how do i change output polarity if this is not ?

thanks

grippo · Post by **grippo** » Tue Oct 02, 2007 10:25 am

The going high and low do change the polarity of the output. If you change just this, the top trace should become inverted., but it will still be off from what you want because in trigger rise mode it fires when the tach pulse comes in. The only way to change this is by mechanically moving the sensor. You can only avoid this by using the 'calculated trigger' option, but your timing will jitter quite a bit because your tach pulses are not even because of the 1 cylinder when it explodes it is much faster than the exhaust.

The Benz Master · Post by **The Benz Master** » Tue Oct 02, 2007 5:33 pm

ah... this explains quite a bit now !

but could i just build a 60deg negative spark table to cope with this?

And it still does not explain why i dont have all the output spark and i have a nice square wave input...?

Thanks

grippo · Post by **grippo** » Wed Oct 03, 2007 8:41 am

If it works and you don't care about the numbers being negative, then that is fine. As far as the missing outputs I am totally stumped. I would like you to try the exact same setup in your last scope shot, bit try it with the attached ini. It is identical to the v2.881 ini but it has an added gauge that will allow us to see if it is missing or discarding any tach pulses that may not show up on the scope. When you bring up MT, right click on a gauge you don't care about right now - like ego or map. A menu will appear and you should select the spare 3 xtra/missing pulses gauge. Keep an eye on this gauge while the engine is running. If a tach pulse is missed or an extra one comes in, the spare3 counter will increment or decrement and you will see the gauge move.

The Benz Master · Post by **The Benz Master** » Sun Oct 07, 2007 7:02 pm

Sorry Grippo, been prety busy since last week... have to do some clean up cuz i have a heated warehouse to store cars for winter ... so it's prety much time i do all this... bike will have to wait a week or 2 more

I'll give you some news about this miss/extra tach pulse gauge as soon as i get time to fidle with it

thanks