EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Stan21 on June 17, 2024, 09:49:42 pm
-
Experimenting with producing HV with ignition coil controlled by thyristor.
The goal is to get sliding corona discharge on dielectric film (that is, "Kirlian" effect), and I want to see how it depends on discharge frequency.
I read there is almost no changes in how corona looks with frequencies from hundreds Hz to tens of KHz, but I want to see it myself.
As I understand, these coils are designed to work on frequencies up to several KHz, not more.
By my understanding, when the frequency on primary increases, at some value there starts a saturation effect, that is, the spark energy at the secondary stops growing and primary coil energy just dissipates heating the iron core of the coil. On the other hand, with increasing frequency the coil resistance grows and because of that the current also drops. This is why for high frequencies ferrite cores with lower inductance are used.
Hope this is more or less correct understanding.
So is there a way to calculate a maximum effective frequency for such a transformer, or is there a data on this in some specs, or from someone's experience?
So far I reached 500Hz, works fine.
Again, I read that 5KHz looks like the highest reasonable frequency when the spark is still generated - is that about right?
-
A way to look at it is: "what was the coil designed for?"
Single-coil igniton systems are uncommon today, but if we go back a few years, let's presume the following use case:
4-cylinder 2-stroke engine.
RPM: 8000 max.
8000/60 = 133.3 revolutions/second. Multiply by two (a 4-cylinder fires twice per revolution)
Maximum frequency for the coil is 267 Hz.
You can think of 6- or 8-cylinder engines, but they either rev lower, or would have two coils.
I'd say that 500 Hz is about the limit of what can be achieved.
The above applies to induction-discharge ignition, which is coil plus breaker.
For your application, a better solution is probably capacitor-discharge ignition. It's lower power, but much faster.
You'll find it on motorbikes, lawnmowers etc. (Brigg&Stratton comes to mind). Cannibalizing an old mower should be easy.
-
That's exactly what I use - 1 uF capacitor charged through 680R resistor, and discharged to primary through FLC10 chip (thyristor+zener+resistor plus protection diode).
As per your consideration about frequency I disagree. What is the engine max rpm and what is the coil max frequency - are not directly related. Coil maximum frequency is mostly defined by core saturation point. I am trying to say that even though the engine breaker hardly provides more then 300 Hz, the coil can definitely work fine on higher frequencies.
Well, anyhow I suppose the answer to my question would be - 2/3 KHz, maybe 4 but below 5.
-
I wonder why I bother.
A question without any useful information at all (ignition type, voltages, schematic/sketch and, and).
From a user that already knows the answer, but needs to be scratched on the back.
Great.
-
So far I reached 500Hz, works fine.
Again, I read that 5KHz looks like the highest reasonable frequency when the spark is still generated - is that about right?
An eight cylinder 4-stroke engine running at 7,200RPM requires 480 sparks per second (480Hz). Any racer knows that a single coil setup will be highly prone to misfiring if 480 sparks per second is demanded from them.
This highly suggests that coil energy above 1kHz will be lacking.
Of course, misfiring is not just missing sparks but extra sparks generated in other cylinders, due to capacitive coupling of ignition leads.
-
May be of some interest to the OP
https://www.eevblog.com/forum/rf-microwave/fly-swatter-ham-radio-transmitter/msg5246334/#msg5246334 (https://www.eevblog.com/forum/rf-microwave/fly-swatter-ham-radio-transmitter/msg5246334/#msg5246334)
-
You use lower inductances at higher frequencies. Typically inductors are pulsed with a voltage, the current in the inductor rises at a rate that is inversely proportional to the inductance. You don't want the current to go too high before you turn the pulse off, so for lower frequencies where your pulse will be longer a higher value is used. Yes this adds electrical resistance. It's all a trade off, more inductance usually makes things more stable but is more expensive and lossy. Less inductance is cheaper and harder to work with, one "trick" is to increase the frequency, this means that the behavior is now more like a larger inductor at lower frequencies without the cost, of course each frequency excites parasitic properties of the components differently and at higher frequencies the effects of parasitic properties are more noticeable.
Now with an inductor with a core you have the magnetic properties of the core. Every time the current/magnetic field changes direction there is a loss, a bit like switching losses in MOSFET's. So whatever the thing you are using is don't just expect to crank things up and it just work betterer.
What makes you think the frequency will change something? Do you have a diagram of what you are making?
-
An ignition coil from a CDI system, something like a 125cc 2-stoke racing motorcycle, will be able to handle much higher frequencies than an old coil designed for points.
The CDI coil needs to be able to generate a pulse very quickly and is generally of a much lower inductance than the older points type as the energy comes mostly from the capacitor, not only the coil.
So if we look at a high performance 125cc race bike (KTM 125, YZ 125) that will run around 11,000 rpm, it will produce a spark about 180 times a second. However, it goes from zero to hero and back again in about 20 degrees of rotation. So, we should be able to run it at about 18 times that frequency .. or about 3,240 Hz.
This is all theory .. YMMV. I'd also reduce the current through it dramatically as it normally lives a ~5% or less duty cycle life, and believe me, they get HOT doing that at 11,000 RPM!
Have fun!
-
For Kirlian photography..?
From past experience, you don't need anything more than 10kV & a few uA AC, preferably a high frequency transformer/static generator.. (dissect an old printer?)
Played with that ~40 years ago (red room or no illumination) ..metal/foil plate AC electrode/plate under a 2-3mm glass sheet, photo paper or film on top & press your finger or WHY on that.
Watch the blue aurora for a few seconds (30?)..then power off & develop the paper/film
Was hit & miss ..was kinda cool when it turned out.
G'Luck! :-+
-
CDI with 1uF cap is good for 200Hz, charge/discharge times will come to issue some point.
Maybe 0.47uF is better to 500Hz.
I have build few cdi ignitions to old Suzuki mopeds.
-
The spark energy and high voltage (higher than the turns ratio would suggest) comes from the rapid collapse of the stored magnetic field. The stronger the stored flux and the quicker it is released the higher the spark voltage or available spark energy. The old Kettering points and condensor system usually showed a spark resonance frequency of 10KHz and higher suggesting the system could be pumped up to that frequency. Some of the early experimental M.S.D. (multi spark discharge) systems did exactly that by pumping energy into the coil around the resonant frequency and operating the coil and included capacitances as a tank circuit. Your statement "The coil saturates at higher frequencies" is incorrect. The fall off of spark power at higher R.P.M. is because the coil hasn't had a long enough time to build up a strong magnetic flux. That was one of the reasons for the old dual point distributors. You could get a much longer 'dwell' without burning the points. The old 6 volt Kettering systems protected the coil from stalled or low R.P.M. burnout by the very resistance of the coil. The 12 volt Kettering systems actually used a 6 volt coil with a ballast resistor and that resistor was often shorted out applying 12 volts to the 6 volt coil during cranking. That is what the 'extra' terminal on the starter solenoid was for.
-
Yup some flawed thinking here.
Modern ignition systems can run to 15k RPM which is a leisurely pace for modern electronics.
Hell, even the slowest SMPS is 10+ kHz = 60k/minute.
Coil/inductor saturation is easily confirmed with a current probe and something to be avoided for best inductor/transformer performance.
-
15krpm is only 250Hz
-
Just to note, some cars have one ignition coil for each spark plug.
For example, one of the Volvo 1998 models (might be 1999 or so though) has one coil per cylinder which means it has 5 coils.
Because the use of these things is so varied you would have to know the car it came from, but because the driver circuits can also be so varied, it is probably better to test it once it is built up on the bench.
Many years ago, myself and some others built a fly trap using an old ignition coil but I do not remember how fast we switched it, this was more than 40 years ago.
-
Years ago I thought of using a car ignition coil as a single ended output transformer for a valve amplifier and just yesterday I looked it up and sure enough several people have already done it. Most distributor-type coils are 100:1 and newer ones will easily tolerate 6 amps in the primary, though not continuously. 3 amps would probably be okay. That means for a 100:1 coil 30mA in the HV side would be just fine. Many have a resistance of about 6000 ohms so you are going to drop 180V. Starting to look like a loser of a project already but practicality is not the name of the game here. The point is to see if it can be done and how good / bad it works.
First off is frequency response. I had a pair of identical 65:1 coils that have 3500 ohm resistance HV side and pumped a signal from an amplifier into the LV winding and fed the HV output into the HV winding of the second coil and them put the scope across the LV output of the second coil. Loaded it with a 10 ohm resistor.
Started of at 1kHz with an output of 8V p/p then measured in steps to 20kHz. Then started to go from 1kHz down but coil started to run out of reactance and pull to much current so I changed it to 0.8V p/p for frequencies under 1kHz. To express it in dB I just got the base 10 log of the voltage and then multiplied it by 10 to make the graph look nice. I hope that is kind of right. The response is of two coils cascaded, not one by itself. X axis is dB.
-
So, all this talk of ignition coils brings to mind an old farm hand trick with a coil from a model T (they had four). The ignition coil has a buzzer like an old doorbell had. When the coil was energized it produced a fat healthy spark at the buzzer frequency. It was possible to remove the buzzer armature and replace it with a modified hacksaw blade with a weight on the far end. when the coil was suspended (usually from a barn beam) and the weighted blade hanging downward you got a fence charger that could knock animals to their knees. Remember some of the old units had a reputation and nick name of 'weedburners'. Better have secure grounds and proper insulation or you end up with a barn burner also.
-
I have seen natural Kettering spark systems show the spark oscillation up to 10KHz (looks like a dampened oscillation) and I am glad your frequency sweep covers this range. Of course the circuit is acting like a resonant tank circuit and the coil may have some peaks and dips across a wide frequency range if driven as a simple step up transformer. I find most iron core inductors have a 'happy zone' which I conclude is a point of self resonance where the efficiency seems to get an artificial boost.
-
To run that GM transformer at 40kHz required high voltage on the primary and then let the field collapse by not clamping the BEMF. Duty cycle was low enough that temperature wasn't a concern. As shown, it worked but that was pushing the transformer.