GN400 DIY CDI

[guest1873]

The other day I managed to get a home built CDI up and running on my GN400. Its essentially the analog CDI circuit at http://www.transmic.net/en/analog.htm.

Anyway, I attach some scope shots. The 1st show both pickups at tickover, the 2nd at higer RPM. The 3rd is the 10 deg BTDC pulse and the capacitor discharge at low rpm and the last at higher rpm. You can see (and feel!) the change in advance.

Pleased with this so I'm going further and making a PIC controlled CDI. I'll be writing my own program so will let you know how it goes.

Attachment Expired

[guest1873]

Forgot to mention that the reason I'm doing this is because the stock CDI doesn't advance any more.

Attachment Expired

[Moto63]

Building your own cdi!!!!  Not that much of a plonker then.   

[guest1873]

Building your own cdi!!!!  Not that much of a plonker then.   

The wife begs to differ! ;-)

[guest1873]

Took the bike for its 1st run today – analog CDI fitted. Works great and the bike is now a breath of fresh air compared to what it was. I've set the CDI to switch to full advance at about 3000 RPM. You do notice a sudden increase in power at this point especially in the low gears but you soon get used to it.

I've attached a couple more oscilloscope shots with labels to show what’s going on. The orange trace  here is the 35 deg BTDC pick up pulse and the scope is triggering off this. So at low speed the spark occurs at the 10 deg BTDC pulse (not shown). Then at 3000 RPM the spark occurs at the 35 deg BTDC.

This DIY CDI has cost me about £5 in parts.

Attachment Expired

[Propellor]

Damn, I'm going to have to read up on CDI stuff now, so I have half a chance of understanding what the heck the wiggly lines mean.


Well done. That's impresive.

[guest1873]

More wiggly lines!  I borrowed a 4 channel scope so we have:

Orange 10 deg BTDC from low speed pulser
Blue     35 deg BTDC from high speed pulser
Purple  Capacitor voltage
Green   Exciter coil output.

This snapshot is taken at low speed and we are going forward in time from left to right.

So the 1st thing to happen is the output of the exciter coil (Green trace) rises briefly to nearly 300V. This voltage is rectified and used to charge the capacitor (Purple trace off the scale initially). Then the 35 deg BTDC comes in (blue), this is ignored by the electronics at low speed. Lastly the 10 deg BTDC (Orange) pulse comes in and as soon as this voltage goes more positive than about 2 volts or so the SCR fires and shorts the capacitor out collapsing the voltage across it (Purple trace) This represents the spark or ignition time. At higher speed the 35 deg pulse isn't ignored and triggers the SCR thus causing the spark to occur earlier.

On my GN I've now adjusted the trigger point so that full advance occurs at about 2200 RPM and the surge in power is less noticeable. I'm very happy with the bike as it is so its probably not worth playing around with more complicated ignition control but we shall see!

Attachment Expired

[Propellor]

So there are always two individual "mechanically" triggered pulses, one at 10 deg btdc, the other at 35 deg btdc? And the electronic gizmo decides which one to allow through, depending on rpm? If I've got that bit right, then what happens to whichever pulse isn't allowed through?

Cheers

[guest1873]

Yes 2 trigger pulses for each revolution of the crank.  The 2 pulses are identical for any given speed but importantly they increase in amplitude as the rpm goes up. It takes a few volts positive to turn on the SCR thus causing the spark and at low speed only the 10 deg pulse is high enough to trigger the SCR because there is a variable resistor in series with the 35 deg pulse. This reduces the voltage from it so that at low rpm its not high enough to trigger the SCR. BUT at some point it will be? In my case I have adjusted the resistor so the point is 2200 RPM.

So, to recap, at low RPM the 10 deg pulse fires the SCR. As revs increase at some point the 35 deg pulse will fire the SCR?

Not sure if this helps having re-read it lol

Edit: So both pulses 'get through', the 10 deg pulse is always high enough to trigger the SCR but at 2200 and above the 35 deg pulse beats it to it!

Edit No 2: next time I get the scope, instead of  looking at the trigger pulses themselves I'll clip a probe onto the Gate of the SCR - This will show them to have different amplitude at this point in the circuit.

[Propellor]

........

.......Edit: So both pulses 'get through', the 10 deg pulse is always high enough to trigger the SCR but at 2200 and above the 35 deg pulse beats it to it!
.......

Ah. Gotcha. I twigged with that edit.

Thanks for taking the time to explain.

[guest564]

So the GN has two VR sensors? I do recall it was rather an odd system. If you are not triggering on the zero crossing then you will get additional advance due to the faster rise time at higher RPM. How many volts do you have on the cap when you are triggering it? with an IGBT/inductive system you usually get around 450V flyback on the primary. Did you measure the GN to see what the stock primary voltage is?
Its a good few years since I did a CDi ignition system though I've done a few inductive systems - last one was an ion-sensing set up for a diesel application.

[guest1406]

I am partial to a GN, ill be watching your progress. Are you doing any other mods to it?

Mackenzie