Pele

iTrader: (28)

How do people get a 36-1 wheel on a rotary?

I'm planning on putting EDIS on an originally NonEDIS engine. (It's actually not even a rotary. It is, however a Mazda engine... B2600 truck engine, G6 family, I4 @ 2.6L)

Mounting the sensor doesn't seem like it'd be too hard, but I don't think I can machine the pulley accurately enough that TDC on the wheel matches TDC on the pulley, nor can I get it perfectly centered.

I don't want clearance between the sensor and wheel to change as it rotates, and I don't wanna throw off my pulley's balance.

ronbros3

HI! just came to this thread , exactly what does a Mega squirt do for an ignition system. what are the pluses. RON in DAYTONA

Jeff20B

iTrader: (13)

The beginning of the ignition event is controlled very precisely. It also can remove typical distributor jitter inaccuracies and the cap and rotor. The result is more power, and a more accurate spark.

muythaibxr

I guess this is the official place to post rotary ignition stuff with megasquirt, so I'll post here...

I'm currently testing leading ignition only with a stock CAS and stock ignitor/coilpack for leading...

It works and runs great. I had to build my own dual lm1815 circuit (which renns designed) and wire it up to the megasquirt, as well as add a pullup resistor on D17 (an LED). The only problem right now is figuring out a good value for the pullup. I originally started with 1KOhm, which was too small, and the current draw was causing everything on the board that ran on +5v to get too little voltage, so sensors were reading wrong, and there was a lot of noise on the TPS... So I put in a 2.2k resistor, which works well for everything but idle, at idle there seems to be a slight miss every so often... SO sometime in the next few days I'm going to put in a 1.5K resistor. If that doesn't work the way I want, I'll keep swapping resistors until I find one that works...

There's a chance that in order to get the engine to run right all the time AND get the sensors to work right all the time, I might have to upgrade the stock voltage regulator on the MS to an lm2940T, which puts out 1A of current instead of 500 mA...

Anyway, more info as it happens... I'll post a good value for the resistor and whether or not I needed to get the voltage regulator both here and on the megasquirt forum when I figure out what works well.

Jeff20B

iTrader: (13)

I once accidentally tried two pull-up resistors at the same time, and the results were the same as with one. I installed some 1k resistors on my LEDs (one each) and then another at the G input of the ignitor. The first resistor pulls it up to 5V and the second up to 12V. This was the first time I had ever hooked something like this up, and removed the 12V one as soon as I realized my mistake. No damage was done and no change in performance was noticed. This was a few months ago when I was bench testing everything. Do you think adding a 2nd pull-up resistor would do you any good? It probably wouldn't hurt anything to try.

muythaibxr

Well, I don't think it'll hurt anything, but since the stock computer uses a +5v square wave, and it's not my car that I'm working on, I'd rather not take the chance. I think I'm on the right track with different values for the resistors since 2.2k got rid of my sensor weirdnesses, but made the idle a little bad... Like I said, I'll just try various values until I get something that works, and if I can't find something that works in all situations, I'll upgrade the voltage regulator and start over again with a 1.5k resistor.

Question:

when you did your tests, were things like the TPS and sensors connected? One of the things I noticed that renns suggested is that something is pulling too much current, once he mentioned that I immediately thought it might be the ignitor, since that's the only thing that's changed that would pull more current from the +5v voltage regulator.

Also, which type of ignitor/coil were you testing with? I'm using the stock 2nd gen (FC) leading ignitor/coil, which may be different from what you're using...

I don't remember where I said this, but I tested the resistance from the ignition advance pin to ground of the old HEI module that worked fine with 1k pullups, and the resistance was around 5.1K, but when I tested the same thing on the FC ignitor, the resistance from the IGt input to ground was only 1K... which means it probably pulls a lot more current.

So anyway, I think I'm barking up the right tree here, but I'll let you know what happens

Jeff20B

iTrader: (13)

I did my tests on the stimulator and didn't notice any weirdnesses on the TPS or other inputs. I was using J-109 ignitors with 1k pull-up resistors and diamond coils. The total current consumption of the MS was somewhere around 110mA with all three LEDs on. When the MS was recieving an icoming trigger signal, the flashing LEDs dropped the current down to 90 something. My voltage reg would get warm to the touch, but that's it.

I tried the 2nd gen coil with ignitor assembly and it produced some really nice sparks. All I did was replace the J-109 and Diamond coil with the 2nd gen stuff in the circuit. There weren't any problems with the sensors or the voltage reg. Again, this was on the stimulator. I've gotten as far as test cranking the MS on the engine with three J-109s and three Diamond coils, and saw a steady cranking RPM in MegaTune. The sensors looked fine during cranking.

muythaibxr

yeah I saw steady cranking, and all that, and when the sensor values were low they were fine, but whenever I stepped on the gas, the tps spiked a lot quicker than it should, and when the engine warmed up, the temps were too high, all of which indicate low voltage from the voltage regulator. After switching to the 2.2k resistors, the problem went away, but the idle started missing every once in a while..

Hopefully tomorrow or saturday I'll be able to test with the 1.5k resistor... I'm pretty confident that the 1.5k will work well.

I don't want to try the 12v signal just because the stock ecu uses a +5v signal. If I get a chance, I'm also going to scope the IGt signal from a stock rx7.... turns out the car I was going to scope tomorrow won't be at my house anymore, so I have to wait till someone else can come over.

muythaibxr

OK after some more testing and tuning, it seems 1.5K is still a little low, 2.2K works well, as does 4.7k. I'm sticking with 4.7K.

Kouta

I have a question about this...
So you guys have found a way to controll the leading sparks, but the trailing is being more difficult (to put it simply).
So right now this system can't be used as a complete standalone? It has to be used in conjunction with the stock ecu for ignition? Or is that just for advanced features?
Basically what I am asking is can i get this and tune in my car without having to go the route of buying a fcd, safc, o2 reader (was planning on having the ms read that), and the likes.
I am fully willing to learn about this etc... just asking before going too far with my idea of ruling the world, that is the world of my car :p

muythaibxr

you still need a wideband o2 controller if you want to go that route, otherwise you're stuck with narrow band.

Other than that, it's full standalone... megasquirt will control leading spark, and fuel, and several other things without the help of a stock ECU. The only catch is that you have to build and install a couple of extra circuits in the megasquirt.

I have been working on code for trailing... I think the author of the msns-extra firmware is also working on it, but I felt like giving it a shot anyway.

muythaibxr

I just tested rotary trailing split on the scope... and it seems to work correctly... I still have a few more tests I want to do but it's working correctly for 2nd gen style leading and trailing on the oscilloscope.

muythaibxr

Alright, I think I have the last of the problems worked out of rotary trailing on the megasquirt... and I've tested them on the scope.

I also added support for FD trailing as well... and it looks good on the scope.

Once it's tested in a car, I'll probably create a separate thread with all info for getting it to work.

badge988

ok Ive been watching this forum for a while now and getting a better understanding of whats going on before I implement the MS-II into my sons FC turbo however Im still perplexed by the lead/trail timing split as well as rotorside firing. As it is right now Im experimenting with the BASIC stamp to use as a variable timing delay from lead to trail but how should I be timing the actual events across the RPM and powerband of the engine? I know I need to have more time delay between L to T at idle but how about under boost when the flame moves faster?
My other question concerns the wasted spark used for the leading side, both plugs fire at the same time correct? If this is the case and the trailing plugs fire at say 3 degrees later why could'nt the trailing plugs both be fired in the same manner being only a few degrees apart? I dont see anything on the stock CAS that would give the ECM a clear idea of rotor-1 firing to get the timing sequence. I know my engine ran on only the trailing plugs once as my son neglected to put the lead plug wires back on the coil and it ran with an impressive spark jumping the towers and to his fingers... Kid has fingers of steel saying it didnt hurt at all and just let them jump. This to me indicates both plugs fire in wasted spark so I dont quite understand why the trail side would have to be on seperate coil packs being only a slightly different timing delay.

Just looking for ideas before I commit the soldering iron and wrenches, once I get started with a project I dont give up until its right...

Thanks in advance
988

88IntegraLS

Great to hear, and thanks for all the effort! My megasquirt is about half assembled right now.

pmrobert

The trailings can't be waste fired due to their physical location higher up. The waste fired rotor's apex is between the T&L plugs when the other rotor is at TDC. The L waste spark goes into a very late exhaust stroke, a T waste spark goes into late intake/early compression due to being on the other side of the apex seal. It's not pretty when that happens, BTDT.

-Mike

badge988

Well that totally sucks, I guess I didn't account for that tight of a timing event. I wonder if I adapted your idea of the tooth shaving of the stock CAS and used the top wheel with the tooth cut off to index a #1 timing event to start a flip/flop in my BASIC stamp controller for controlling the trailing spark and use the lower reluctor with 2 teeth for actual ignition event control.

My timing idea: 1 tooth to index the stamp controller telling it #1-T is ready for fire and lead has already fired and countdown started for trail fire, Next lead fire without index assumes #2 trail countdown and fires.
I could do this with just a simple flip/flop however Im trying to simulate an actual variable timing delay with respect to engine speed making the delay shorter as speed increases. Ideas thoughts?

One thing I do want to note that Im sure many seasoned techs here know already is that the stock amplifier in the leading and trailing ignitors are polarity sensitive on the trigger lines. The input must be high to fire the coils and inverting polarity from an external source will destroy the input as I have found out the hard way on the bench. The amplifier is also very sensitive with an input impedance of around 600 ohms, +1.2 volts will trigger the amplifier.

988

pmrobert

988, the CAS is already set up to index #1 as you wish - the CAS runs at 1/2 speed so the two tooth trigger pops every 360 of crank - on #1 TDC. That said, this is what I found regarding trailing ignition in an NA street ported 12A:

Mike Robert spent some time on a dyno last year and was able to test what happened in the case of stock 'spilt' spark timing, trailing only, leading and trailing firing simultaneously (no split) timing and leading only timing.

The results were:

* Leading and trailing spark in a roughly stock timing split configuration gave ~164 RWHP @ 8500.
* Trailing spark only only ran very badly (as expected) and power was way down, EGT was high enough to make the exhaust medium red; that run was aborted.
* Leading and trailing spark firing simultaneously were no different form stock split until after 6500 rpm or so; even then the difference was almost negligible.
* Running leading only spark produced about the same power as well, within error margins.

All of these tests were run with 26 degrees leading advance max on 87 octane pump gas.

End quote...

Trailing is really only important for emissions and maybe preventing idle misfires - which also goes to emissions. A good read for lots of these things might be:
http://smrmicro.com/re-ky.pdf

I thought of using a similar solution (though with a BasicX, supposed to be faster than a Stamp but I really don't know) but did the math and found that it was pretty easy to get into a processor/code bound condition. Consider: 6000 rpm = 100 revs/sec = 10ms. 2 ignition events per rev = 5 ms. Dwell might be 2 ms. 2ms should work - http://smrmicro.com/v-web/gallery/Ma...oject/DSCN3408 This is the easy part. Each degree of crank is ~28us, I may be wrong, I'm doing these calcs in my head and from memory. Anyway, you can see things get very tight and the interpreted BASIC Atmel solutions may be a touch slow. I forget the exact specs on the BX but it was a nogo timewise. There's no doubt in my mind that an Atmel coded in asm would be up to the job. There's also the necessity of conditioning the VR signals into TTL compatible pulses. I'm not being pessimistic, just trying to share info and would like to hear any ideas you may have.

-Mike

muythaibxr

pmrobert beat me to it.

badge988

I might just forget about the trailing spark all the way around. You are right about the stamp being a bit slow however I do use them for making radio data interfaces at 1200 baud.

My only concern with leaving the trail spark out is a misfire under boost. Being FI I wont have much fuel in the intake manifold however even a small sneeze can and has destroyed TO-4 compresser turbines with fantastic results. Knowing my kid is going to be driving the car and boys being boys he will find oppertunities to really let 'er have it. Im trying to do damage control more than anything.


So I might just forego the trailing spark for a later time if I feel its needed

988

pmrobert

988, I agree with your plan of action and reasoning; you want to keep that intake nice and clean re true backfires.

OT, these little BASIC Atmel chip sets are amazing for serial passthrough stuff, among other things. I use one of the BXs for realtime display of Megasquirt/engine parms talking 9600 in from MS/ 19200 out to a VFD. Took a couple of days to get the code right, but much easier than asm and much more convenient than keeping a laptop connected all the time.

-Mike

cafcwest

in