Coils and CDIs

[neilbardsley]

Just incase others find it useful. I asked Fred Winterburn some questions about CDI units and coils

I asked Fred Winterburn about different coils with CDI units and turns of secondary wirings to primary in the coil.

I believe you are referring to the classic retrofit CDI. Like many other multi-spark CDIs, it does not play well with a high inductance coil. A high inductance coil (typical canister coil used for standard kettering points ignition) will store some of the energy from the first strike. This stored energy in the coil's magnetic field will produce a voltage that opposes (IE, bucks) the voltage from the second strike of the CDI rendering any spark either weak or useless. To compensate, classic retrofit has to time the second spark almost 400µS after the first spark so that the coil's opposing voltage has died down sufficiently that it will perform like a transformer again and actually deliver the capacitor's energy to the spark gap without destructive interference (IE the voltage bucking). So, they want you to use a 'CDI' coil which is basically a low inductance transformer. The CDI that I build is designed to use a high inductance coil (NOT a 'CDI coil) since the capacitor is sized appropriately such that the capacitance and inductance form a tuned circuit that oscillates with alternating negative and positive sparks until most of the energy is used up. The old MSD 6A gets around the bucking issue by waiting a full 1mS between capacitor discharges so that coil oscillations have died down (wasted as heat in the coil) and to give time to recharge the capacitor. 1mS is far too long to wait to be of any use in my opinion. Regarding higher turns ratios, I have not seen a correlation that would suggest a higher turns ratio is quicker with regard to voltage rate of rise. More likely the opposite is true. What I have seen is that cheaper coils, with cheaper iron cores, have a slower voltage rise than other inductive coils, all of which do have slower voltage rises than 'CDI' transformer coils.

Don't be tempted to put a CDI coil and pair it with your CDI or the switch will burn out as soon as it's switched to STD since so called CDI coils have very little resistance in the primary winding. Perhaps you have one that I've fused? I started doing this a while back to protect the switch (here Fred is referring specific to his 6v CDI. I have one in my 356)

Now I admitted I had the old MSD unit (in my 914) he was talking about

You could use a lower inductance coil but you might not want to.. One thing about the MSD is that it will have a very short duration spark no longer than 50µS long and then 1mS goes by before the next short duration spark. The power supply uses a large transformer and charges the capacitor in 1/4 cycle through two large power transistors in parallel. 1mS = 6 crankshaft degrees at 1000rpm if I remember correctly, so you can see how subsequent sparks (except at perhaps idle) are completely useless. Using a lower inductance and primary resistance coil could gain some extra power out of that MSD considering it really isn't that powerful. However it might come at the cost of early failure of the unit. Some impedance to tame the capacitor discharge isn't a bad thing. Also if you needed to switch back to kettering you'd need to carry a spare inductive coil.

Any CDI is better than none

Sent from my 22011119UY using Tapatalk

[Ian Comerford]

Thanks Neil. I have a very small section of my brain allocated to anything to do with electrics/electronics so this level of detail is a bit much for me. Is there an executive summary, ie what works well with CDI+ etc?

[Jonny Hart]

duplicate post

[Jonny Hart]

Our CDI+ boxes work best with the correct Bosch CDI coil - it's that simple. It wasn't by accident, chance or luck - it's by design.

If you set out to extract the best performance of any system in a motor car (electrical, mechanical or otherwise), you play to the strengths of that system. In this case, the Bosch coil is unique in its ability to charge and discharge much faster than others. Our boxes have been designed to exploit that trait to fire a second spark at 400us intervals. No other system can do that. As Fred says, there is no point in firing more sparks after 1ms - it's too 'late' to make any difference to performance.