Tests of CDI Ignition Modules

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
“You can see this clearly in websites from folks that make them, especially ferrite cores where its easy to grind in a gap, they state a zero gap inductance and then less and less inductance as they grind in more more and more gap.”

To a layman like me, this could make perfect sense if you look at it from the perspective that rather than citing the increased in air gap as the culprit for the loss of Inductance, perhaps they should be looking at loss of mass of the core as the culprit. Just saying!

I understand that post on a topic can get off course at times on forums. To me DKGrimm topic piqued my interest as I and probably some others. Scale engines builders are always looking for improvements in the current technology. I believe that what DKGrimm is after.

Branching off into the Origins of the Universe, so to speak, really has gone a little off course. If you have something that works, show us and show us why and how it works. Theory and conjecture really aren’t adding much to the conversation.

I am not condemning or condoning anyone, getting a little off topic is just the nature of the beast. There comes a time when it distracts more than it helps!
 
“You can see this clearly in websites from folks that make them, especially ferrite cores where its easy to grind in a gap, they state a zero gap inductance and then less and less inductance as they grind in more more and more gap.”

To a layman like me, this could make perfect sense if you look at it from the perspective that rather than citing the increased in air gap as the culprit for the loss of Inductance, perhaps they should be looking at loss of mass of the core as the culprit. Just saying!

I understand that post on a topic can get off course at times on forums. To me DKGrimm topic piqued my interest as I and probably some others. Scale engines builders are always looking for improvements in the current technology. I believe that what DKGrimm is after.

Branching off into the Origins of the Universe, so to speak, really has gone a little off course. If you have something that works, show us and show us why and how it works. Theory and conjecture really aren’t adding much to the conversation.

I am not condemning or condoning anyone, getting a little off topic is just the nature of the beast. There comes a time when it distracts more than it helps!

DK is doing an empirical study and sharing the results, as well as using a storage scope to view wave forms to try to understand what's going on, as well as trying to use that understanding to improve what can be done with CDI. He's not just summarizing what works, he's investigating, and sharing, what doesn't work so he can create an understanding of what will work so he can design something that does work. Sharing what doesn't work is part of the process, and all of this understanding takes theory. DK is currently trying to solve the "why are CDI sparks of such short duration" Mystery of the Universe, by using observation and theory, which will only succeed if the theory is correct. When incorrect "facts" and "theories" gets stated on a public forum it can create a mythology that propagates to other forums (I read it over on HMEM...), which is harmful for people trying to understand things. And its the people that post incorrect information that are taking things off topic, the people correcting them are just doing their public duty.
 
“You can see this clearly in websites from folks that make them, especially ferrite cores where its easy to grind in a gap, they state a zero gap inductance and then less and less inductance as they grind in more more and more gap.”

To a layman like me, this could make perfect sense if you look at it from the perspective that rather than citing the increased in air gap as the culprit for the loss of Inductance, perhaps they should be looking at loss of mass of the core as the culprit. Just saying!

I understand that post on a topic can get off course at times on forums. To me DKGrimm topic piqued my interest as I and probably some others. Scale engines builders are always looking for improvements in the current technology. I believe that what DKGrimm is after.

Branching off into the Origins of the Universe, so to speak, really has gone a little off course. If you have something that works, show us and show us why and how it works. Theory and conjecture really aren’t adding much to the conversation.

I am not condemning or condoning anyone, getting a little off topic is just the nature of the beast. There comes a time when it distracts more than it helps!

DK is doing an empirical study and sharing the results, as well as using a storage scope to view wave forms to try to understand what's going on, as well as trying to use that understanding to improve what can be done with CDI. He's not just summarizing what works, he's investigating, and sharing, what doesn't work so he can create an understanding of what will work so he can design something that does work. Sharing what doesn't work is part of the process, and all of this understanding takes theory. DK is currently trying to solve the "why are CDI sparks of such short duration" Mystery of the Universe, by using observation and theory, which will only succeed if the theory is correct. When incorrect "facts" and "theories" gets stated on a public forum it can create a mythology that propagates to other forums (I read it over on HMEM...), which is harmful for people trying to understand things. And its the people that post incorrect information that are taking things off topic, the people correcting them are just doing their public duty.

meanwhile, I'm getting back to my machining, hope to post some progress soon...!...
 
Wow, the subject of inductors and transformers from one misread PDF is not good. I don't know where you guys got your knowledge from or what kind of equipment you have both worked on. But you both missed and are missing some information. That paper "LECTURE 25 Basic Magnetic Material Information and Relation to Electrical Properties" is as poorly written as my attempt to explain why a CDI has such as short spark.

HMEL and Peter you both missed something very simple that was very quickly stated and blown by. Frequency, they mentioned frequency and based all the flowing material on that but, failed to express they were working with frequency when comparing air gap inductors and transformers. I have worked with and repaired power supplies, switching power supplies, and RADARs that worked in the Giga Hertz frequencies. I'm going to do this as simple as I can. At lower frequencies such as 50-60 Hz an iron core will store more energy than an air gap inductor. It the iron core will transfer more energy than an air core at low frequencies, period. A laminate multi-plate iron core will out store a solid iron core or an air gap because there is air between the plates, so it's a hybrid. laminate multi-plate iron core transformers will also out perform the others up to a certain frequency, period. Any worthwhile manufacturer will provide the frequency that was used to calculate the inductance of an inductor or transformer coil. laminate multi-plate iron core inductors and transformers do have a problem as the frequency gets higher and that is with the Eddy currents being generated in the solid core and laminate multi-plate iron core. I know from the equipment that I have repaired that as frequency goes up I would find more and more air-core inductors and transformers. So yah at higher frequencies the air-cored stuff stored energy better than the iron core stuff at higher frequencies. The iron core stuff at high frequencies would not release the energy fast enough. But that damn paper failed to explain this properly. But, everything they talked about was based on frequency. The writer of that paper was very concerned about coil saturation. Do we care about whether an ignition coil gets saturated, no not really. Actual we do, I know from racing that dual point ignitions were very popular and I've even seen a picture of one that had 39 sets of points to saturate the ignition coil as much as possible. Do we want to saturate an inductor or transformer when working with AC signals? NO.

That paper mentions things like:
*a. Faraday’s Law: Open Loop of Wire in a time varying flux
*Magnetic devices form the throbbing heart of switch mode power
supplies.

*The time varying magnetic field in the core itself will lead to
core losses and heat generation.
*That is the wire current will tend to further flow
to the outside of the wire, as compared to that from simple skin
effects, and cause increased resistance at high frequencies above
the skin effect values by factors of 10-1000.
*Normal I2R losses will not be sufficient for estimating
wire losses of high frequency currents due to skin effects inside
the wire which force high frequency current to flow on the outside
surface of the wire and not flow at all in the center of the wire.
1. Inductors: A single Cu wire wound around a magnetic core.
The purpose of an inductor is to store electrical energy.
Storage will best be done in air, not in magnetic material as
we show below. Hence, inductors have air gaps purposefully
placed in their cores

And it goes on and on, mostly talking about switching power supplies and Not ignition coils.

When I had my ignition coils made they asked me questions like; what frequency range, what voltage in, what voltage out, how big of finished product, how much current flow in, where I wanted the pins. They came back with a design they thought would work using an existing design for my approval.

Now why does a CDI ignition have such a short spark. CDI ignitions use a high voltage capacitor and a low resistance primary coil. When we charge a capacitor it stores energy. If we remove the charging wires it the capacitor stays charged. If we charge an ignition primary coil and remove the charging wires the magnetic field will collapse and we loose all the energy. So it does not store energy. To create the magnetic field we need voltage. When we connect say a battery to the coil, the coil looks like a dead short and we have current flow through the coil. This creates the magnetic field we want. To get a bigger magnetic field we need we need more voltage. With more voltage we get more current flow and a bigger magnetic field. Now the capacitors I use to store a charge have an ESR of just 8 milli-Ohms (0.47uf), my ignition coils have a primary resistance of 80 milli-ohms. Now forget reluctance because we are only deal with one dis-charge/charge cycle. As for inductance that only comes into play when the capacitor discharges into the coil. We also are not going to get the SCR internal resistance into the picture.

Now remember I was talking about time constants? Okay, I'm not going to do all the calculations, I'm keeping it simple. But, using 12 volts and unlimited current, it takes just 0.000000166267 seconds to dump all it's energy into the coil, not taking the coils resistance or reluctance or inductance into consideration. The coil on the other hand needs 0.005527500 seconds to except all that energy, so quite a bit slower. Now when the voltage drop across the coil becomes more than the voltage from the capacitor the field will collapse and I get a spark. So at what point (time) does the voltage allow the field to collapse? I sure as hell can't use a coil that takes 5.5 milli-seconds to charge or discharge. It turns out it's right about between T1 & T2 or about 0.001871250 seconds. For the RPM of my engines meant that I needed a charge and discharge time of less than 2 msec. I'll let you figure out what the RPM would be. Now with a LDI or inductive ignition we want to get into saturation or very close to it, that way we get as much power out of the coil as we can, dwell time baby or dual points. Because with a CDI we can only get the coil charged at best to about 74.85% not taking coil resistance or inductance into account we will get a shorter spark. The actual energy depends on the coil and capacitor design and it will most likely be closer to 50%. So we go from a time constant of about T1 & T2 (CDI) to T1 to T5 or 1.5 time constants to 5 time constants for discharging/spark time. As I mentioned before CDI and inductive ignition coils are Not built the same and one needs to remember this. You can't or shouldn't use a CDI ignition coil on an inductive ignition, it won't last. But, you can go the other way around, use an inductive coil on a CDI but, with reduced spark strength.

Cheers
Ray
 
74 Sprint:

OMG - REALLY - you wrote: "I'm not going to do all the calculations"
and then you come up with "it takes just 0.000000166267 seconds..."

Did you really need to write twelve decimal places?
I can't even see the screen well enough to move the decimal over to make sense of this.
How about just writing 0.2us (point two micro seconds) and be done with it.
 
OMG - REALLY - you wrote: "I'm not going to do all the calculations"
...
I can't even see the screen well enough to move the decimal over to make sense of this...

That's ok, he also said "To get a bigger magnetic field we need we need more voltage", and a number of other things that are likely to confuse novice readers who don't understand what he really meant.

I generally really enjoy reading what Ray writes, but I'm just going to pretend that one didn't happen.

Will
 
Everything should be as simple as possible, but not simpler. --attributed to Albert Einstein
Don, I'm curious to understand how you took your digital scope data loaded into Excel, and computed the Joules given a voltage and current waveform. Reference the first table, in your report in post #1. I'm guessing you are treating them as sine waves, and computing some sort of RMS averaging, and took the product of the voltage and current, and divide by the sample time, and add them up to result in an integrated watts/s to calculate the joules. I would be interested to see inside your .xls/.xlsx file. Any tricks you had to do to make the calculation deterministic/repeatable? That high frequency noise in the spark plug waveform, did you figure out a way to filter those spikes from the data? I didn't see any other post asking about this.
.
The many posts on the magnetics for the transformer design has introduced things I have no experience with, so that discussion has been interesting.
 
Don, I'm curious to understand how you took your digital scope data loaded into Excel, and computed the Joules given a voltage and current waveform.
Good questions. No, I’m trying not to make any kind of assumptions or approximations. I have a dual channel, 1GHz sampling scope that takes 2500 samples of the trace on the scope. It can capture a full 2500 samples of a 2.5 microsecond event. It also has a built-in USB port to download a full set of samples the the computer. The test setup has the base of the plug grounded through a 10 ohm sampling resistor so I can record both plug voltage and current simultaneously for any single spark event.
Reference the first table, in your report in post #1. I'm guessing you are treating them as sine waves, and computing some sort of RMS averaging, and took the product of the voltage and current, and divide by the sample time, and add them up to result in an integrated watts/s to calculate the joules.
I import the 2500 samples of voltage and current, and – exactly as you say – calculate the product (watts) at each sample. Then multiply by the time interval to get joules per sample. Then it’s just a matter of adding them all up.
I would be interested to see inside your .xls/.xlsx file. Any tricks you had to do to make the calculation deterministic/repeatable? That high frequency noise in the spark plug waveform, did you figure out a way to filter those spikes from the data? I didn't see any other post asking about this.
The sample intervals are so short that I can capture even the energy during the spikes. I don’t do any filtering.
The many posts on the magnetics for the transformer design has introduced things I have no experience with, so that discussion has been interesting.
I’ve been working with electromagnetic stuff for sixty odd years on and off, but I don’t ever get tired of learning new things about it.
 
74 Sprint:

OMG - REALLY - you wrote: "I'm not going to do all the calculations"
and then you come up with "it takes just 0.000000166267 seconds..."

Did you really need to write twelve decimal places?
I can't even see the screen well enough to move the decimal over to make sense of this.
How about just writing 0.2us (point two micro seconds) and be done with it.
LOL, I just punched in the numbers and the Excel spreadsheet did the rest. Yah I guess I could have whittled it down some, I wasn't sure which way to go on that one. My bad.

Ray
 
That's ok, he also said "To get a bigger magnetic field we need we need more voltage", and a number of other things that are likely to confuse novice readers who don't understand what he really meant.

I generally really enjoy reading what Ray writes, but I'm just going to pretend that one didn't happen.

Will
You know when I went to electronics engineering school they had the habit of making us go into so much detail it was nuts. When I got out into the field I started to forget everything that wasn't related to the stuff I wasn't working on. But it is somewhere buried in my brain, use it or lose it. Anyway, they even taught us how to make microchips and I'm thinking to my self 'if the damn thing is blown then I'm just going to change it out, I'm not going to make a chip.' So why even teach us this, my head is already full of crap I'll never use.

As for Don's work, I really prefer practical over theory. Yes I can do Spice models but, Spice models only get you close. I have found that I can do all the calculations and Spice models I want but, prototype the circuit and see what you really get, it won't be the same as the theory and calcs. What I'm trying to say is 'do we really need to do all the calculations? or can we just see what Don comes up with and be happy.'

Ray
 
You know when I went to electronics engineering school they had the habit of making us go into so much detail it was nuts. When I got out into the field I started to forget everything that wasn't related to the stuff I wasn't working on. But it is somewhere buried in my brain, use it or lose it. Anyway, they even taught us how to make microchips and I'm thinking to my self 'if the damn thing is blown then I'm just going to change it out, I'm not going to make a chip.' So why even teach us this, my head is already full of crap I'll never use.

As for Don's work, I really prefer practical over theory. Yes I can do Spice models but, Spice models only get you close. I have found that I can do all the calculations and Spice models I want but, prototype the circuit and see what you really get, it won't be the same as the theory and calcs. What I'm trying to say is 'do we really need to do all the calculations? or can we just see what Don comes up with and be happy.'

Ray
Engineers are widely know for their laser focus on theory and their relative blindness to what happens in practice. Lord knows I have done my share, but a really good engineer uses both.

Don
 
HMEL, that is not the difference between an inductor and a transformer, there are inductors without air gap (EG toroidal core), and there are transformers with air gap (EG neon sign transformers).

what do you mean you haven't seen anyone measure it, every single manufacturer of ferrite core inductors and transformers measures their inductance with zero air gap and with various thicknesses of air gaps (to lower the inductance to a customer need) as part of their QA, they all get the same result, zero air gap results in largest inductance and inductance goes down progressively with size of air gap. This is an empirically proven fact, it is now up to you to understand why <<===
And specifically what units do they use, and is the same frequency and conductors the same in both cases and are you implying that faradays law disappears because there is a larger gap in the core. And as a matter of fact you can indeed couple inductors that look like transformers. So define what a inductor to you is. A general definition is normally one transforms the voltage and current and the other stores and releases energy. So lets be specific here.
 
Wow, the subject of inductors and transformers from one misread PDF is not good. I don't know where you guys got your knowledge from or what kind of equipment you have both worked on. But you both missed and are missing some information. That paper "LECTURE 25 Basic Magnetic Material Information and Relation to Electrical Properties" is as poorly written as my attempt to explain why a CDI has such as short spark.

HMEL and Peter you both missed something very simple that was very quickly stated and blown by. Frequency, they mentioned frequency and based all the flowing material on that but, failed to express they were working with frequency when comparing air gap inductors and transformers. I have worked with and repaired power supplies, switching power supplies, and RADARs that worked in the Giga Hertz frequencies. I'm going to do this as simple as I can. At lower frequencies such as 50-60 Hz an iron core will store more energy than an air gap inductor. It the iron core will transfer more energy than an air core at low frequencies, period. A laminate multi-plate iron core will out store a solid iron core or an air gap because there is air between the plates, so it's a hybrid. laminate multi-plate iron core transformers will also out perform the others up to a certain frequency, period. Any worthwhile manufacturer will provide the frequency that was used to calculate the inductance of an inductor or transformer coil. laminate multi-plate iron core inductors and transformers do have a problem as the frequency gets higher and that is with the Eddy currents being generated in the solid core and laminate multi-plate iron core. I know from the equipment that I have repaired that as frequency goes up I would find more and more air-core inductors and transformers. So yah at higher frequencies the air-cored stuff stored energy better than the iron core stuff at higher frequencies. The iron core stuff at high frequencies would not release the energy fast enough. But that damn paper failed to explain this properly. But, everything they talked about was based on frequency. The writer of that paper was very concerned about coil saturation. Do we care about whether an ignition coil gets saturated, no not really. Actual we do, I know from racing that dual point ignitions were very popular and I've even seen a picture of one that had 39 sets of points to saturate the ignition coil as much as possible. Do we want to saturate an inductor or transformer when working with AC signals? NO.

That paper mentions things like:
*a. Faraday’s Law: Open Loop of Wire in a time varying flux
*Magnetic devices form the throbbing heart of switch mode power
supplies.

*The time varying magnetic field in the core itself will lead to
core losses and heat generation.
*That is the wire current will tend to further flow
to the outside of the wire, as compared to that from simple skin
effects, and cause increased resistance at high frequencies above
the skin effect values by factors of 10-1000.
*Normal I2R losses will not be sufficient for estimating
wire losses of high frequency currents due to skin effects inside
the wire which force high frequency current to flow on the outside
surface of the wire and not flow at all in the center of the wire.
1. Inductors: A single Cu wire wound around a magnetic core.
The purpose of an inductor is to store electrical energy.
Storage will best be done in air, not in magnetic material as
we show below. Hence, inductors have air gaps purposefully
placed in their cores

And it goes on and on, mostly talking about switching power supplies and Not ignition coils.

When I had my ignition coils made they asked me questions like; what frequency range, what voltage in, what voltage out, how big of finished product, how much current flow in, where I wanted the pins. They came back with a design they thought would work using an existing design for my approval.

Now why does a CDI ignition have such a short spark. CDI ignitions use a high voltage capacitor and a low resistance primary coil. When we charge a capacitor it stores energy. If we remove the charging wires it the capacitor stays charged. If we charge an ignition primary coil and remove the charging wires the magnetic field will collapse and we loose all the energy. So it does not store energy. To create the magnetic field we need voltage. When we connect say a battery to the coil, the coil looks like a dead short and we have current flow through the coil. This creates the magnetic field we want. To get a bigger magnetic field we need we need more voltage. With more voltage we get more current flow and a bigger magnetic field. Now the capacitors I use to store a charge have an ESR of just 8 milli-Ohms (0.47uf), my ignition coils have a primary resistance of 80 milli-ohms. Now forget reluctance because we are only deal with one dis-charge/charge cycle. As for inductance that only comes into play when the capacitor discharges into the coil. We also are not going to get the SCR internal resistance into the picture.

Now remember I was talking about time constants? Okay, I'm not going to do all the calculations, I'm keeping it simple. But, using 12 volts and unlimited current, it takes just 0.000000166267 seconds to dump all it's energy into the coil, not taking the coils resistance or reluctance or inductance into consideration. The coil on the other hand needs 0.005527500 seconds to except all that energy, so quite a bit slower. Now when the voltage drop across the coil becomes more than the voltage from the capacitor the field will collapse and I get a spark. So at what point (time) does the voltage allow the field to collapse? I sure as hell can't use a coil that takes 5.5 milli-seconds to charge or discharge. It turns out it's right about between T1 & T2 or about 0.001871250 seconds. For the RPM of my engines meant that I needed a charge and discharge time of less than 2 msec. I'll let you figure out what the RPM would be. Now with a LDI or inductive ignition we want to get into saturation or very close to it, that way we get as much power out of the coil as we can, dwell time baby or dual points. Because with a CDI we can only get the coil charged at best to about 74.85% not taking coil resistance or inductance into account we will get a shorter spark. The actual energy depends on the coil and capacitor design and it will most likely be closer to 50%. So we go from a time constant of about T1 & T2 (CDI) to T1 to T5 or 1.5 time constants to 5 time constants for discharging/spark time. As I mentioned before CDI and inductive ignition coils are Not built the same and one needs to remember this. You can't or shouldn't use a CDI ignition coil on an inductive ignition, it won't last. But, you can go the other way around, use an inductive coil on a CDI but, with reduced spark strength.

Cheers
Ray
Some clarity is in order here. The paper is a compilation of notes from a class and textbook. The reason for posting it was it contains the mathematical proof of where the energy is stored. It clearly states at the top of the notes HF.(high frequency) It is series of organized notes and is much different then a published paper.

My experience with transformers of 50 to 60 Hz is in the power transmission side. Here above all you want efficiency. So it depends on the application and what you consider important. And I agree its a whole different subject.

My father used to build ignition system for stock cars out of magnetos. So the understanding of timing, points etc along with carburetion comes from experience and little from the academic world. It was a world or menthanol and mineral oil. So I understand the timing issues and the impact of capacitor constants on timing for high rpm engines.

My experience with ham radio ( extra license) gives rise to all sorts of combinations of inductors in circuits particularly in combinations with antennas, filters and other circuit elements. This includes some work in 10GHz transmitters and magnetic loop antennas.

But what I do appreciate is the detail you have submitted in your reply. It gives rise to an easier discussion of technical issues.

And for that I say thank you
HMEL
 
Apologies for the lack of progress obtaining and posting new test data. Those who know me will not be surprised to learn that I have a million excuses, but I'm kind of stuck on a problem right now. The noise bursts on the spark waveforms I'm trying to measure are generally so large that they swamp out the measurements of the voltage and current that determines the real power and energy delivered into the spark. For example, this one gives an error in energy measurement during the spike itself because the voltage and current seen in the recorded data is artificially stretched out by the recovery time of the saturated scope amplifiers.

1676567152125.png

I am using a homemade 1000:1 voltage divider combined with a 10:1 voltage divider probe on the scope to look at spark plug voltage. I decided to make a passive resistor-capacitor network add-on to the voltage divider do two things.

First is to wire directly from the voltage divider to the scope with a coaxial cable to avoid noise pickup. That means not using the 10:1 scope probe. The new network would have to do the 10:1 reduction.

Second, add low pass filtering to remove some of the wildness of the noise bursts without distorting the main part of the waveform.

I've experimented with designs for this network for days now. Given the resistance and capacitance of divider network, the coax and the scope input, the best I can come up with is this attenuation vs frequency profile. This just won't cut it. Frequencies between 100 Hz and 1 MHz would have about twice as much gain as the dc gain, while a 10 MHz component would be down about 5:1.

1676566858788.png


I'll have to find a different approach. I'm sure there is a way, but I'm not there yet. I know this is not an engineering design forum, but if anybody out there has ideas about what to try, I'd be glad to hear from you by a post or maybe better a direct message.

Meanwhile, I don't want to post erroneous results. I'm going to have to figure this out before I produce any more measurements.
 
  • Like
Reactions: Zeb
Thanks for your efforts Don. It seems to be a tough problem.
Thinking outside the box a bit - (like way outside the box).
Since you are trying to measure the energy ultimately delivered to the load (spark plug). Could absorb the delivered energy and measure it as heat. Maybe run the ignition for an extended length of time heating a resistor in an enclosed container (to avoid losses) and measure the heat energy somehow or the temperature rise of the resistor?
Yeah I know - way outside the box. But analyzing all those wacky waveforms and oscillations that may or may not even be contributing to the useful power output might be a dead end.

Another thought. Could you measure the DC power input to the system when delivering a good number of sparks per second, estimate some losses for the electronics and get close to an estimate for the delivered output. The circuit quiescent power can be measured without out any spark output. Although it probably amounts to almost nothing because once the capacitor is charged it probably stops oscillating.

Another thought. If you short the output would all the energy be absorbed by the circuit? Then maybe it can be measured as heat in the circuit or as input DC power??

Just random thoughts.
 
<snip>
I'd be glad to hear from you by a post or maybe better a direct message.

<snip>

Don;
Other's may be interested in the technical aspect of making these measurements, so I'm replying to your post.

For laughs, I asked openAI (ChatGPT) this question;
"How to filter the broadband noise while measuring spark plug voltage and current in an experimental ignition system?"
I got the following response:
"When measuring the voltage and current of a spark plug in an experimental ignition system, it is common to encounter broadband noise. This noise can interfere with the accuracy of the measurements, making it difficult to analyze the ignition system's performance. Here are some techniques for filtering the broadband noise:
1) Shielding: One way to reduce broadband noise is to shield the wires and components that connect to the ignition system. Shielding involves surrounding the wires and components with a conductive material such as copper, which blocks electromagnetic interference (EMI). Shielding can be effective in reducing noise that is picked up by wires and other conductive components.
2) Filtering: Another technique for reducing broadband noise is to use filters. Filters are electrical circuits that attenuate or remove unwanted frequencies. For example, a low-pass filter can be used to remove high-frequency noise from a signal. Filters can be implemented using passive components such as resistors, capacitors, and inductors, or with active components such as operational amplifiers.
3) Grounding: Proper grounding is critical for reducing broadband noise. All components of the ignition system should be properly grounded to minimize noise induced by ground loops. Ground loops occur when different components of a circuit are grounded at different points, causing current to flow between the grounds, which can create noise. By grounding all components of the ignition system to a common point, ground loops can be eliminated, and noise can be reduced.
4) Twisted Pair Cables: Twisted pair cables can be used to reduce electromagnetic interference (EMI) and radio frequency interference (RFI). These cables are made up of two conductors that are twisted together, which creates a balanced signal. When EMI or RFI is present, the noise is picked up by both conductors equally, and the balanced signal is maintained. This allows the noise to be cancelled out at the receiving end of the cable.
5) Ferrite Beads: Ferrite beads can be placed around cables to reduce EMI and RFI. Ferrite beads are cylindrical pieces of ferrite that are threaded over a cable. The ferrite material absorbs high-frequency noise, reducing the amount of noise that reaches the cable. This technique is particularly effective at reducing common-mode noise, which is noise that is picked up by both conductors of a cable.

These are some of the techniques that can be used to filter broadband noise while measuring spark plug voltage and current in an experimental ignition system. By using these techniques, the accuracy of the measurements can be improved, allowing for more accurate analysis of the ignition system's performance."
-------------------
Most of these are common sense, but I didn't think of using #5 Ferrite beads to filter the broadband noise from the spark current. I've used these for HIRF filtering on the IAPS, but found that printed circuit board capacitors and inductance of circuit traces made a perfect zero cost (board layer count impact) LC filter (about a 2cm trace between ground planes right at the connector entry point). However it required the ground plane of the board to be connected to a low inductance ground return (that was the hard part). Ferrite beads if I recall start kicking in around 20MHz, and their Bode plots are based on continuous wave tests, so I don't know how they would work with the broad band RF from the spark.

I also did a search on the "magic box" for "measuring spark plug energy from ignition system", and got some hits on SAE papers that cost $35 unless you have a subscription.
https://www.sae.org/publications/technical-papers/content/2000-01-0245/There are other search suggestions.
This may be of interest;
https://www.autosuccessonline.com/scoping-and-interpreting-secondary-ignition-wave-forms/This talks about a capacitive scope probe, which from reading the article is test equipment used in automotive repair. This must be some procurable piece of equipment.
 
Thanks for your efforts Don. It seems to be a tough problem.
Thinking outside the box a bit - (like way outside the box).
Since you are trying to measure the energy ultimately delivered to the load (spark plug). Could absorb the delivered energy and measure it as heat. Maybe run the ignition for an extended length of time heating a resistor in an enclosed container (to avoid losses) and measure the heat energy somehow or the temperature rise of the resistor?
Yeah I know - way outside the box. But analyzing all those wacky waveforms and oscillations that may or may not even be contributing to the useful power output might be a dead end.

Another thought. Could you measure the DC power input to the system when delivering a good number of sparks per second, estimate some losses for the electronics and get close to an estimate for the delivered output. The circuit quiescent power can be measured without out any spark output. Although it probably amounts to almost nothing because once the capacitor is charged it probably stops oscillating.

Another thought. If you short the output would all the energy be absorbed by the circuit? Then maybe it can be measured as heat in the circuit or as input DC power??

Just random thoughts.
All valid approaches, Dave. But there is no free lunch. The heat idea is one that has a long history in the RF world, especially microwave stuff. The idea of discharging into a resistor instead of into a spark plug is quite useful; I've done that with most of my earlier work with magnetos and CDI modules. Trouble is that the changing the size of the resistor changes the results, which suggests that the exact timing and amplitude of the output affects the internal losses in the coil and whatever else. I'm trying some other ideas also, and I see that "Ignator" has found some promising references in his post above. The trail has not gone cold by any means.
 
Don;
Other's may be interested in the technical aspect of making these measurements, so I'm replying to your post.

For laughs, I asked openAI (ChatGPT) this question;
"How to filter the broadband noise while measuring spark plug voltage and current in an experimental ignition system?"
I got the following response:
"When measuring the voltage and current of a spark plug in an experimental ignition system, it is common to encounter broadband noise. This noise can interfere with the accuracy of the measurements, making it difficult to analyze the ignition system's performance. Here are some techniques for filtering the broadband noise:
1) Shielding: One way to reduce broadband noise is to shield the wires and components that connect to the ignition system. Shielding involves surrounding the wires and components with a conductive material such as copper, which blocks electromagnetic interference (EMI). Shielding can be effective in reducing noise that is picked up by wires and other conductive components.
2) Filtering: Another technique for reducing broadband noise is to use filters. Filters are electrical circuits that attenuate or remove unwanted frequencies. For example, a low-pass filter can be used to remove high-frequency noise from a signal. Filters can be implemented using passive components such as resistors, capacitors, and inductors, or with active components such as operational amplifiers.
3) Grounding: Proper grounding is critical for reducing broadband noise. All components of the ignition system should be properly grounded to minimize noise induced by ground loops. Ground loops occur when different components of a circuit are grounded at different points, causing current to flow between the grounds, which can create noise. By grounding all components of the ignition system to a common point, ground loops can be eliminated, and noise can be reduced.
4) Twisted Pair Cables: Twisted pair cables can be used to reduce electromagnetic interference (EMI) and radio frequency interference (RFI). These cables are made up of two conductors that are twisted together, which creates a balanced signal. When EMI or RFI is present, the noise is picked up by both conductors equally, and the balanced signal is maintained. This allows the noise to be cancelled out at the receiving end of the cable.
5) Ferrite Beads: Ferrite beads can be placed around cables to reduce EMI and RFI. Ferrite beads are cylindrical pieces of ferrite that are threaded over a cable. The ferrite material absorbs high-frequency noise, reducing the amount of noise that reaches the cable. This technique is particularly effective at reducing common-mode noise, which is noise that is picked up by both conductors of a cable.

These are some of the techniques that can be used to filter broadband noise while measuring spark plug voltage and current in an experimental ignition system. By using these techniques, the accuracy of the measurements can be improved, allowing for more accurate analysis of the ignition system's performance."
-------------------
Most of these are common sense, but I didn't think of using #5 Ferrite beads to filter the broadband noise from the spark current. I've used these for HIRF filtering on the IAPS, but found that printed circuit board capacitors and inductance of circuit traces made a perfect zero cost (board layer count impact) LC filter (about a 2cm trace between ground planes right at the connector entry point). However it required the ground plane of the board to be connected to a low inductance ground return (that was the hard part). Ferrite beads if I recall start kicking in around 20MHz, and their Bode plots are based on continuous wave tests, so I don't know how they would work with the broad band RF from the spark.

I also did a search on the "magic box" for "measuring spark plug energy from ignition system", and got some hits on SAE papers that cost $35 unless you have a subscription.
https://www.sae.org/publications/technical-papers/content/2000-01-0245/There are other search suggestions.
This may be of interest;
https://www.autosuccessonline.com/scoping-and-interpreting-secondary-ignition-wave-forms/This talks about a capacitive scope probe, which from reading the article is test equipment used in automotive repair. This must be some procurable piece of equipment.
Good, solid suggestions. Thanks. I'll be studying your post and checking the references for a bit.

Don
 

Latest posts

Back
Top