CDI ignition - gross vs nett spark energy

[bmcdaniel]

Hmm, really?
My old Delta Mk 10 didn't do any of that, and it would fire an oil soaked plug gapped at 0.120" running a sooty rich mixture all day long without missing a beat.
I suspect that most reasonable mixtures, once ignited, will burn until consumed.
Guess I'll have to keep throwing fresh matches on my BBQ starter fluid to keep it lit too ...

I built one of those Deltas back in the 70's, too. I probably still have it tucked away somewhere. Your right, it wasn't multispark but I believe the later iterations of CDI's with multiple sparks burned cleaner at low rpm. That was the whole purpose of developing multiple spark ignitions, igniting low rpm lean cruise mixtures.
And yes, if the charcoal bits in your BBQ are too far away from each other (like fuel molecules in a lean mixture) you do have to relight the mixture to consume it all.

[JohnP]

I built a Tiger SST kit.

[SwedCharger-67]

...Well guess what? Some CDI ignitions it seems have the same problem regarding advertised vs actual spark energy.
Did a little test tonight on a red 3-letter CDI with a "6" suffix and the results were interesting. Previously I thought the coil introduced very high losses but this may not be the case. It seems that of all the energy pumped into the coil primary, fully 50% gets bounced back out again into the battery feed wires and doesn't go toward making a spark. Look at the first pic. Yellow is the (simulated) plug voltage while arcing (~1600V). Blue is the current through the plug - 370mA for about 140uS. Purple is the amount of energy (mJ) as the arc progresses. it finishes at 40.8mJ - that's not very much! Where did it all go?

Thanks for very interesting evaluation Circlotron! The results are disappointing indeed, and your numbers seem very reasonable, you even see the little bump on the input when the energy is reflected back to the input side.
Shows that it's not easy to choose an ignition system...if you want an efficient one...and trusting the marketing material seems to be of limited value. It puts a heavy load on the tinkerer who has to evaluate all potential purchases on his lab bench!

[Circlotron]

you even see the little bump on the input when the energy is reflected back to the input side.
Shows that it's not easy to choose an ignition system...if you want an efficient one...and trusting the marketing material seems to be of limited value. It puts a heavy load on the tinkerer who has to evaluate all potential purchases on his lab bench!

The duration of the energy dumping out of the primary back to the battery supply is the whole time the purple trace is descending. The attached pic is slightly clearer than the previous one. See the yellow primary voltage trace (100 volts per div) is slightly above the line for 120uS after the initial poke from the capacitor. This is when and for how long it is pushing juice back into the battery instead of passing through the coil and making spark.

Prim Volts, amps, mJ 2.png

Cars are full of specialised systems e.g. auto trans, diffs, suspension, camshafts, carburettors, exhausts etc etc and ignitions are no different in this respect. Almost anyone can get these things to more or less work, but to really make them sing requires a specialist.

[SwedCharger-67]

...Cars are full of specialised systems e.g. auto trans, diffs, suspension, camshafts, carburettors, exhausts etc etc and ignitions are no different in this respect. Almost anyone can get these things to more or less work, but to really make them sing requires a specialist.

So very true! Many times it's difficult to, in the product specifications, find even the most basic data needed for an evaluation prior to purchase.
Back to the topic...what could be done to minimize the reflected energy?

[JohnP]

Series diode in the output to the coil?

[Circlotron]

^^No, diode is from coil positive to battery positive, cathode to battery.

[clshore]

you even see the little bump on the input when the energy is reflected back to the input side.
Shows that it's not easy to choose an ignition system...if you want an efficient one...and trusting the marketing material seems to be of limited value. It puts a heavy load on the tinkerer who has to evaluate all potential purchases on his lab bench!

The duration of the energy dumping out of the primary back to the battery supply is the whole time the purple trace is descending. The attached pic is slightly clearer than the previous one. See the yellow primary voltage trace (100 volts per div) is slightly above the line for 120uS after the initial poke from the capacitor. This is when and for how long it is pushing juice back into the battery instead of passing through the coil and making spark.

Prim Volts, amps, mJ 2.png

Cars are full of specialised systems e.g. auto trans, diffs, suspension, camshafts, carburettors, exhausts etc etc and ignitions are no different in this respect. Almost anyone can get these things to more or less work, but to really make them sing requires a specialist.

You are still not getting it.
Find a schematic for a CDI ignition, and try to understand how it works.
The DC-DC converter generates 600 V DC from 12 VDC
It CANNOT magically reflect power backwards from the CDI circuitry into the battery.

It just don't work that way!

If I take hot toast and shove it back into the toaster, it DOES NOT reflect energy back to the power company as it cools off.
If I take a torch and heat my brake rotors red hot, and then step on the brakes, it doesn't suck the heat out and make the car move.
If I take a battery charger and hook it up to a fully charged 12V battery, I will NOT get 120V AC out of the wall plug.
If I pump hot CO2 and water vapor into the exhaust pipe of a motor, I will not get Gasoline and Air back out of the Carb.

You follow?

[Circlotron]

You are still not getting it.
Find a schematic for a CDI ignition, and try to understand how it works.
The DC-DC converter generates 600 V DC from 12 VDC
It CANNOT magically reflect power backwards from the CDI circuitry into the battery.

It just don't work that way!

Have a look at the attached diagram of the relevant part of a MSD6.
Just before the spark, the LHS of C15 is at about +450V and the RHS is at zero volts.
At the moment of spark SCR1 turns on pulling the LHS of C15 to ground and so the RHS of C15 goes to -450V with respect to ground, taking COIL_PLUS orange wire down with it. As the capacitor progressively discharges, COIL_PLUS rises back toward ground and would in fact overshoot to about +250 or +300v if it were not stopped by D7 which has it's top end tied to battery. With a 14V battery supply the COIL_PLUS cannot go any higher than about +15v and so it sits there dumping out the remainder of it's energy back to the battery. Stick a current clamp and a scope on the wire and have a look for yourself.

[Circlotron]

Here's a scope pic of the current in the thick red batttery wire to the MSD6. Peaks at 34 amps BTW but the average current is much lower. Anyway, when the spark first starts the red battery wire current is zero as per the perfectly horizontal part of the trace. About 130uS after being triggered the dump capacitor voltage reaches zero and the coil primary current is at a maximum, the battery supply current goes negative by almost 20 amps. The coil positive wire is pulling up against the battery momentarily, trying to make it go higher than it's normal 14 volts. This is where/when the circuit is pushing unused energy back into the battery.

[JohnP]

Yes, my added diode doesn't do anything.

How about a Gate Turn Off thyristor?

It's like an SCR, but you can turn it on AND off, without waiting for the current to fall to zero.

http://en.wikipedia.org/wiki/Gate_turn-off_thyristor

[maxc]

You are still not getting it.
Find a schematic for a CDI ignition, and try to understand how it works.
The DC-DC converter generates 600 V DC from 12 VDC
It CANNOT magically reflect power backwards from the CDI circuitry into the battery.

It just don't work that way!

Have a look at the attached diagram of the relevant part of a MSD6.
Just before the spark, the LHS of C15 is at about +450V and the RHS is at zero volts.
At the moment of spark SCR1 turns on pulling the LHS of C15 to ground and so the RHS of C15 goes to -450V with respect to ground, taking COIL_PLUS orange wire down with it. As the capacitor progressively discharges, COIL_PLUS rises back toward ground and would in fact overshoot to about +250 or +300v if it were not stopped by D7 which has it's top end tied to battery. With a 14V battery supply the COIL_PLUS cannot go any higher than about +15v and so it sits there dumping out the remainder of it's energy back to the battery. Stick a current clamp and a scope on the wire and have a look for yourself.

When the scr1 is pulled too ground that's a waste of power too ground if the tranformer is on its positive cycle?

[JohnP]

It's a pulse transformer. One pulse only, Vasily.

[maxc]

It's a pulse transformer. One pulse only, Vasily.

I was pulseing 12 volts back into the 12 volt side of an old 50 amp computer transformer. It was direct driveing a 300 watt 12kv neon sign transformer. Works good. But the back emf pulse needs control. The neon sign transformer is my primary coil. My secondary coil is just as simple.
Too many voltage drops in cdi ignition lose power. Old points ignition snap the coil better.
@Circlotron do a current scope pic with a resister or zeners instead of a ignition coil.

[Circlotron]

@Circlotron do a current scope pic with a resister or zeners instead of a ignition coil.

No need. A 10 ohm resistor would give you a pulse output peaking at 45 amps that decays exponentially, decreasing by 63% in 10us as per a normal RC curve. A zener across the output would blow up itself or the box if the zener voltage is less than 450v because there is nothing to limit the current.