Got a current-controlled circuit that operates off of 24V, the European standard for large trucks. When I operate a 12V automotive ignition coil on it the coil hisses and squeals. Under 16V that noise stops. The only difference is that the ramp up to the same prescribed peak current happens in slightly under 1msec instead of 2msec. Repetition rate is around 10msec.
A coil designed for 24V operation does not exhibit such noise but they are usually of older technology and inferior to modern coils for sports cars.
I am clamping the primary leakage inductance spike at the usual 400V.
It sounds like[1] you're trying to emulate the classic Kettering ignition coil circuit using an ultra high speed high voltage switching transistor *without* the 100nF capacitor across the coil primary required to slow the voltage rise time sufficiently for the original CB points' expanding separation distance breakdown voltage rating to outrun the rising back EMF produced by the very rapid drop in current from your high speed switch.
In the original Kettering circuit, this 100nF capacitor (across the CB points - effectively in parallel across the coil primary via the battery) was essential to the working of the circuit to prevent the back emf energy being robbed by the arc initiated by CB points separation. This solved the problem of arcing but introduced another problem, namely that of electrical erosion from the short circuit discharge path of the
6/12/24v ignition current supply from the LSI battery on the 100nF capacitor when the points were closed ready for the next ignition cycle.
If you haven't already done so, try including a 100nF 500vac rated capacitor across the coil primary to limit the dV/dt stress.
[1] Corona discharge if you've not got the secondary protected with a
15mm air gapped spark gap load to emulate the effect of a spark plug gapped to 0.8mm operating at a compression pressure of 15 to 20 Bar.
I was testing my own version of the Wireless World CDI circuit at 200 sparks per second some four decades ago and would hear what I assumed was internal corona discharge if I made the test gap larger than 18mm or so by sliding a 3mm thick piece of perspex sheet too far into the 6mm air gap - the spark would extinguish allowing me to hear the quieter corona discharge.
I was able to burn the perspex from the heat of the sparks if I let them play on one spot for more than 5 to 10 seconds at a time. The ignition point was always on the edge of the sheet where the sparking was in the most intimate contact and able to introduce heat into a small volume not so well served by the heatsinking effect of the bulk of the perspex sheet.
Reduce the dwell. It's not that hard. Usually in old cars with points the d well was 66.6%. Some electronic is like that but some is AC. The AC ones wi ll give you more trouble. A scope will tell. A precision capacitor could do it. If it is just 66.6% dwell I could see some sort of thingy with a ring counter and a certain reset system...
But I don't know what it is getting now so I can't say shit.
Uneven magnetization of the core, perhaps? There's a skin depth for magnetization, and faster magnetization does magnetostriction of the surface, making internal stresses in the laminated steel. Changing the ballast resistance would be one way to treat it, or redesigning the core (using thinner sheets or higher resistivity steel).
Another possibility is the low-voltage side capacitor, could be the wrong value and the RLC circuit hasn't got the usual frequency of damped ringing. It's unclear how the clamping circuit affects this.
Yes, I operate it without since that's how most transistor ignitions work these days. The voltage rises to the 400V clamp level over a few microseconds which didn't seem outlandishly fast. But maybe it is for an older style coil so I shall try with a cap then. Thanks for the hint.
The secondary is always connected to a gap of 1mm or less.
There is no dwell seting in my case. The circuit charges the coil until it reaches a prescribed current, then lets go. It's timed so that this fits the situation, the coil never dwells at the peak current.
The peak primary current is regulated and always the same. There is no dwell time, the circuit lets it fire the microsecond that the prescribed peak current has been reached.
Hmm, that might make noise. However, the onset is quite sudden at around
Some coils have a diode to ensure there is no spark during the charging event. Is it possible that this diode, or the air-gap associated with it is breaking down before you go into the constant current mode?
The coil is in effect a nominal 100:1 transformer, 16V in will provide
1.6kV during the charging time and back to zero during when the current becomes "constant".
--
Mike Perkins
Video Solutions Ltd
www.videosolutions.ltd.uk
Interesting, I never heard of diodes in there. When I measured around at the coil with generator, scope and such I never had any diode effect on the secondary.
So far I haven't seen sparking during the charge, I am running this all open right now. The current also ramps up nicely as expected.
Yow! At sparkplug voltages, that's gotta be SOME diode. A gas-filled tube (pointy electrode = anode, flat electrode = cathode) would work, if the right gas mixture was used.
The gas-filled diode really DOES look like the diode schematic symbol, too.
If done with semicondutors, it's gotta be some exotic multipellet design. And, you're not likely to find a surface-mount or thin-wire-leads version of that diode. Silicon diode stacks need a LONG package, and HV corona will kill thin wires exposed to air.
There's an excellent reason for that, namely (even assuming it was a practical proposition) that using unipolar spark current is detrimental to spark plug electrode life.
The effect of asymmetric spark current produced by the old fashioned Kettering inductive energy method whether by the traditional CB with condenser or not quite so ancient high voltage transistor switch, is detrimental enough to spark plug service life without resorting to the stupidity of HV steering diodes.
Whenever I see a "Simple Transistorised Ignition" module being employed, it makes me want to spit on the so called 'designer' for persisting with such an old fashioned circuit dressed up with 'the modern technology of HV switching transistors'. The reason being that, unlike the innovative CDI circuit, it fails to address the issue of leakage in the HT circuit caused by damp ingress on HT leads or due to plug fouling interfering with the inductive discharge process which the original CB and condenser (and 'transistor switched' analogues) are so reliant upon to generate the required sparking voltage.
Using a CDI module designed to produce just one full cycle of HT (two sparks one straight after the other) reduces the asymmetric erosion effect on the spark plug's electrodes. Admittedly it doesn't completely eliminate it but it's enough to double or even triple the service life of most spark plugs in use today.
You might not think that electrical erosion on the spark plug electrodes would be the main limit on service life but, just as with CB points used by vintage vehicles, this is in fact the case.
Using a CDI module to relegate the ignition coil's function to that of 'step up transformer only' not only provides a hotter spark it eliminates the damping effect of a leaky HT circuit on the inductive back emf mechanism which the original Kettering (and transistorised variants) are so entirely reliant upon to provide the spark energy.
If you're going to 'transistorise' your spark ignition circuit, don't do it the shit way with a simple HV switching transistor, do it the right way with a CDI module. :-)
I see your point. Though when I look at the "old" spark plugs from my fairly simple vehicle (Mitsubishi Montero Sport) they look almost pristine and without erosion after 15 years of service at 70000 miles. So I kept the as spares, which I'll probably never need.
In this case CDI is out but I can't go into project details. The spark plugs do not need a long life because runtimes are short and infrequent.
It only has to block the charge-up direction where assuming a 100:1 turns ratio there won't be more than 1500V reverse voltage in a normal vehicle with a 12V battery. In the spark direction the diode conducts so no reverse voltage.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.