Retarding timing on a nonadjustable CDI

I'm working for a snowmobile shop now and I'm working on a way to retard the timing on brand new snowmobiles so that they can accept more boost from the turbo kits that we put on them. The motors can handle it if we use race gas but we want to make a reliable pump gas turbo.

So here is what I'm dealing with. The CDI uses two trigger coils, one for low RPM's and on for high RPM's. I only want to retard the timing on the high side trigger. The problem is that both coils feed the same thyristor that dumps the voltage from the capacitor and fires the sparkplugs through the coils. I can move the trigger coils but only about 2 degrees and I want to pull at least ten degrees of timing at top end.

I was wondering if anyone has any ideas. I was thinking about a potentiometer with a diode acting as a half wave rectifier. The reason I want to do this is because the thyristor triggers the system to spark has a certain voltage at which it fires. The trigger coil produces a wave at which point the thyristor dumps the capacitor. As the engine spins faster the amplitude of the wave increases which triggers the thyristor sooner making an electrical timing advance. I want to reduce the amplitude of that wave effectively retarding timing. Ideas??

A picture of why I want to decrease the amplitude. The taller wave is the wave that I'm getting and I want to reduce it so that it triggers the thyristor later.

To trigger the SCR later, in your sketch, why not just use a two-resistor voltage divider. Just use high resistance resistors to minimize current or you will get a inductor delay (assuming your signal is coming from a reluctor).

Or since you are looking for a delay, just let a little current bleed through a resistor. This will delay higher RPMs more than lower RPMS.

Both plans will have the disadvantage that at low RPMs, the signal may be too low to trigger at all, but I take it that's when the other pickup will have control.

The big question is what decides which trigger to use and at what RPMs.

Here's a simple divider. The bottom line may be replaced with a ground depending on how the circuit is constructed.

.

The other option of using a diode to reduce the voltage by a fixed amount might work if you have enough amplitude at lower RPMs to work with, but then the amplitude won't reduce as much at higher RPMs. This may be what you want, depending on what you want the advance curve to be like.

Any other static-delay time will be problematic in that it will delay more with higher RPMs. That is likely the opposite of what you want.

sounds interesting,but makes my head hurt trying to understand it! :laugh:

A small trigger (gate) voltage requirement to activate ~ 1-2Vdc. You have an exact nomenclature number off the SCR?
I can look up/obtain the data sheet at work for you.

What is the amplitude of the larger (normal) trigger pulse?

Reasoning I ask, the voltage divider would attenuate the pulse height but as Loudhvx mentioned, if your already at minimum trigger voltage may not work at all....

Also, in using this concept to decrease the pulse amplitude to one SCR, this would effect both low and high side coil circuits...

If you choose to go this manner, stay away from potentiometers for this, noisy compared to fixed 1% resistor networks.

You have a schematic diagram of the CDI to coil circuits we can look at?

A small circuit with an additional delay timer & SCR circuit to fire the high end it would a better choice of control... The timer (ex 555 Timer) could be made adjustable providing you some tweaking capabilities.

Possibly a RC gate trigger delay to additional SCR for the high end would be all you need as well...

OMR

Have you been to MSD's site? Have a look into the different controls available, like these:
Adjustable timing controller: PN8680
Universal Boost Timing Master: PN5462
That way it can be tuned easily.

I can't really see the thyristor or get any detailed information from arctic cat about what the circuit looks like. I still have to do a bunch of testing about what the amplitude and voltage are at different RPM's. They don't exactly support us tweaking everything like this. Actually they are very much against it.

I really like your idea of bleeding a little current off with a resistor lou. These motors use CVT clutches so basically they stay at max revs all of the time. So I need to get retarded timing at the top end where I'm making the most boost and keep the stock timing on the low end in order to get it to rev quicker.

Could I still use a potentiometer as the resistor??

I'll look into the MSD systems. I've never seen them before.

Thanks everyone for your help, and don't feel bad plummen. Wrapping my head around this is like getting tri's mother into spandex. Like the spandex my mind can stretch but its approaching it's limits. :woohoo:

is this going to cause warranty issues down the road?

Yeah, I was afraid you were dealing with a black-box issue, that is; no real circuit details to work with.

Are you even sure the pickup is a reluctor/pickup (that is, a coil of wire in a magnetic field with an iron pointer passing near the pickup?

If it is a Hall-effect pickup, the output amplitude will not vary with the RPM.

I ask because I don't know if you've actually observed the pickup signal on a scope, or are just assuming it is a reluctance pickup triggering the thyristor, as that is the classic, old-school CDI. I would be surprised to see such an old method of driving a CDI on a new machine.

If you are sure the pickup is outputting a reluctor based signal, and the amplitude affects timing, then yes, you could make one of the resistors adjustable, or I'll draw up an even better setup. I'm not at home, so I'll have to draw it later.

As I said earlier, the diode method may also be a consideration.

The resistor divider will reduce amplitude as a ratio... output = ratio x input.
Here, as input goes up by one volt, output will go up by a fraction of one volt.

The diode method will reduce amplitude as a fixed amount of reduction...
output = input - fixed diode voltage.
Here, as the input goes up by one volt, the output will go up by one volt, but the output will start out lower by .7v, and will always be lower than input by .7 volt (until input gets less than .7v at which point output will be near 0v).

But in the end, it may take a hybrid of the two methods to really get what you want for a range of RPMs.

Yup, I'm most definitely sure that it is the reluctor type. It is an older method but it seems to work well. If it ain't broke...

Ya the warranty has long been voided on these machines. I think it was somewhere in between adding aftermarket fuel injection controllers, rising rate fuel pressure regulators, and the LCD touch screen AFR gauge.

Anything that you can do for me would be awesome Lou. I really appreciate it.

Oh, and the diode method sounds like it would be exactly what I'm looking for. Especially if it could be adjustable.

I'll try to draw up some possibilities later when I get home.

How are you accessing the pickup signals?
Are you able to get to both wires of each pickup, or are you accessing the signal somewhere else?

Also, a standard reluctor pickup signal is AC. When there is a big positive spike, there should also be a big negative spike (not necessarily the same amplitude in both directions though). If not, then there may be some internal signal processing happening.

Also, what does the rotor look like? Is it a circle with a rectangular tooth on it, or is it more like a pointer, or does it have multiple points?