Tests of CDI Ignition Modules

[Stefan-K]

Gentlemen
I wonder what this thread is about. It started good as a thread about testing CDI devices. And Now it drifted away to a philosophical thread about the natural wonders of the Universe and the mysteries of Electronic in specific. I thought it could be of interest but Now im out.

 

[ozzie46]

Gentlemen
I wonder what this thread is about. It started good as a thread about testing CDI devices. And Now it drifted away to a philosophical thread about the natural wonders of the Universe and the mysteries of Electronic in specific. I thought it could be of interest but Now im out.

Me too.

 

[Vietti]

I've seen much larger deviations in other threads. If you look at the original objective of this thread it was to test and compare CDIs and the Sage-Gedde system.

I still think the resolution of this comparison is very worthwhile for the hobby. I hope it continues.

 

[Zeb]

Peter, you may be interested in the table I put together for a working paper on magnetos about 15 years ago. That working paper and some others are available on my website www.dkgsite.com. This stuff is scattered all over in textbooks and I found it useful to have it all in one place for easy reference.

This was a good read.
Don't give up Don. I'm trying to wrap my head around ignition still. It would be good to see the CDI experiment completed.

For others wanting to give up, there's a handy function where you can hover over a username and select "Ignore". If used on me I would not be offended.

 

[DKGrimm]

Me too.

Gentlemen
I wonder what this thread is about. It started good as a thread about testing CDI devices. And Now it drifted away to a philosophical thread about the natural wonders of the Universe and the mysteries of Electronic in specific. I thought it could be of interest but Now im out.

Absolutely right, and it's mostly my fault. I started the thread and published some test results to start with in message #1 with all the CDI modules I could get my hands on. Then I stopped to make a Cage-Gedde coil driver to test coils. That worked like a charm but I don't really have a bunch of readily available coils to test other than the car coil test I reported as an attachment to message #79.

The detour started, I think, when the question came up about why CDIs produce such short sparks, while inductor coils give longer sparks. Somewhere along the line I read that most automobiles today use inductor coils because the longer spark can ignite lean mixtures more effectively. That question leads into the dark woods of electromagnetic theory, where no one - me included - can say he has not gotten lost from time to time.

If you can stand one more short detour, I just completed a brief analysis of what looked like a Goldilocks solution of firing a CDI through a series coil to stretch the current out in time. It does work -- in theory -- but I can't figure out a way to design a practical circuit to do it.

Next step for me is to find coils available to engine builders that might be good candidates for testing and evaluating. If anybody can help me find some of those, I'd be grateful.

Don

 

[dsage]

I'm still hanging in here. Keep up the good work Don. I'm still trying NOT to use my coil driver(s) (as good as they are) because (as I said at the beginning) using the large coil bugs me when it's approaching the size of the engine and is a distraction from the appearance on the model in most cases. I think the short(er) sparks are fine for a model as long as the circuit to produce them is compact and can be hidden - and can be improved upon through your work Don.
Just my preference.
FWIW a cheap solution CDI for single cylinder engines appears to be what I posted a while back (Posts 129 and 56) It's a distraction from this thread so I didn't elaborate on it nor post a schematic for it.
I'm looking at other solutions as well.

 

[peterl95124]

DK, I infer from your post that you think the RC time constant of the capacitor and coil are what affects the burn time, but it's not ringing because it dies in less than one cycle from of the series resistance of the spark. I'm inferring this because I think you said you were putting an extra inductor in series with the coil which alters the total inductance (when connected in series the inductances add). or am I misunderstanding you ?

 

[DKGrimm]

The whole concept is to dump all of the capacitor energy into the coil and the spark and then not allow ringing to put energy back into it. The capacitor output would be clamped to ground after it discharges. My analysis is only a part of a piecewise linear analysis that applies only during the duration of the spark itself. Ringing before and after involves a much more complex model.

 

[dsage]

Don:
FWIW - If you study some of the better CDI circuits, one end of the capacitor is connected to the coil (in series). The other end is connected to the juction of a half bridge driver. Initially the driver pulls the capacitor up to the the HV supply and the current travels through it and the coil causing a spark . This leaves the capacitor charged. Then the driver immediately reverses and the end of the capacitor is pulled to ground. The capacitor discharges with the current passing through the coil again to produce a second spark right after the first. Sort of like a controlled ringing. Or maybe like a class A amplifier where the output leg swings both positive and negative on each cycle. This doubles (or is it 4x) the power per cycle.
NOT to be confused with a multi-strike ignition which just fires the coil twice sequentially at some repetition rate.
To overcome the issue with short sparks many CDI boxes (MSD for instance) use a system of multiple sparks per trigger (maybe 8 or so at low rpm) to imitate one long spark. The issue with this is that the multiple spark repitition rate can not be sustained at much over idle rpm's.
A bit of a kludge IMHO.

 

[peterl95124]

You might want to consider theory around a fly-back transformer. When a core becomes saturated the magnetic field is distorted around the gap in a ferromagnetic material. This distortion will collapse yielding an additional energy. You will find a flyback transformer on CRTs. I just removed one last night to experiment with. An extra winding is put around the secondary to feed back this energy into the primary. This will boost the voltage output over a normally wound transformer. I wanted the ferromagnetic material more then I wanted the winding and circuits. I went down this rabbit hole because i am working on a magnetic loop antenna and wanted to know more about magnetic circuits. and ran across fly back transformer that used this property. It is a very real phenomena.

Oh No Mr Bill!, here we go again!, folks that don't appreciate the mysteries of the universe
might want to skip over this post!

the energy in a magnetic field per unit volume is
E = 1/2 B^2 / Mu,
B is the magnetic field (flux per unit cross sectional area),
Mu is the permeability of the medium

the permeability of iron is roughly 3 x 10^-3 Henries (times length divided by area)
the permeability of air is roughly 1 x 10^-6 Henries (times length divided by area)

so the energy in an air gap is larger per unit volume than in iron,
but the total energy depend on the total volume of each (intuitively, it takes more
energy to create a magnetic field in air than in iron, so there is more energy stored
in air than in iron given the same field strength and material volume, though in a transformer,
or generator, or alternator, these are never the same volume).

air gaps are typically very thin, compared to their cross section, they contain very little
volume, but this is measurable and the contribution to total energy is calculable, and
hence the division between energy in the iron core and the air gap is calculable, though
as DK points out this division ratio isn't really important.

but the air gap geometry, typically being thin with a large surface area, means the fringe
lines of magnetic field that bulge out around the gap are negligible, IE its not the distortion
of the field that matters.

it is my belief (I could be wrong here, might need some myth busting !) is that the reason
for the air gap in a flyback transformer is to avoid saturation of the core, so that its response
to the applied ampere-turns is more linear which makes the oscillator easier to design.
An additional reason is to get more energy per unit volume, reducing cost of materials (this
belief is backed up by google searches).


now back to CDI vs LDI, if DK needs to increase the inductance of the coil in order to
lengthen the spark's burn time in CDI, then an air gap in the coil is the wrong direction to
be looking since an air gap decreases the inductance of a coil (which can be compensated
for with more turns of wire, but then that multiplies how many turns needed in the secondary
so becomes a cost-of-materials, and overall size and weight of coil, issue).

I have played with flyback transformers, but have never designed the oscillator from scratch,
so I guess I need to do that some day to add to my knowledge of the mysteries of the universe.

 

[peterl95124]

The whole concept is to dump all of the capacitor energy into the coil and the spark and then not allow ringing to put energy back into it. The capacitor output would be clamped to ground after it discharges. My analysis is only a part of a piecewise linear analysis that applies only during the duration of the spark itself. Ringing before and after involves a much more complex model.

Nice !!! keep us posted.

 

[IC-man]

I'm out too. I want to make things not talk about it

 

[DKGrimm]

Oh No Mr Bill!, here we go again!, folks that don't appreciate the mysteries of the universe
might want to skip over this post!

the energy in a magnetic field per unit volume is
E = 1/2 B^2 / Mu,
B is the magnetic field (flux per unit cross sectional area),
Mu is the permeability of the medium

the permeability of iron is roughly 3 x 10^-3 Henries (times length divided by area)
the permeability of air is roughly 1 x 10^-6 Henries (times length divided by area)

so the energy in an air gap is larger per unit volume than in iron,
but the total energy depend on the total volume of each (intuitively, it takes more
energy to create a magnetic field in air than in iron, so there is more energy stored
in air than in iron given the same field strength and material volume, though in a transformer,
or generator, or alternator, these are never the same volume).

air gaps are typically very thin, compared to their cross section, they contain very little
volume, but this is measurable and the contribution to total energy is calculable, and
hence the division between energy in the iron core and the air gap is calculable, though
as DK points out this division ratio isn't really important.

but the air gap geometry, typically being thin with a large surface area, means the fringe
lines of magnetic field that bulge out around the gap are negligible, IE its not the distortion
of the field that matters.

it is my belief (I could be wrong here, might need some myth busting !) is that the reason
for the air gap in a flyback transformer is to avoid saturation of the core, so that its response
to the applied ampere-turns is more linear which makes the oscillator easier to design.
An additional reason is to get more energy per unit volume, reducing cost of materials (this
belief is backed up by google searches).


now back to CDI vs LDI, if DK needs to increase the inductance of the coil in order to
lengthen the spark's burn time in CDI, then an air gap in the coil is the wrong direction to
be looking since an air gap decreases the inductance of a coil (which can be compensated
for with more turns of wire, but then that multiplies how many turns needed in the secondary
so becomes a cost-of-materials, and overall size and weight of coil, issue).

I have played with flyback transformers, but have never designed the oscillator from scratch,
so I guess I need to do that some day to add to my knowledge of the mysteries of the universe.

OK, Peter, if I understand all you wrote here, we agree.

 

[DKGrimm]

Then the driver immediately reverses and the end of the capacitor is pulled to ground. The capacitor discharges with the current passing through the coil again to produce a second spark right after the first. Sort of like a controlled ringing.

Here's a good example from my test of an early S/S module that I run on my Otto-Langen engine.

Don

 

[dsage]

Thanks Don.
I'm confused by the voltage spike at about 55us or so without any corresponding current?
Would that be because the conduction (plasma/arc - whatever) extinguished before it came along so nothing for current to pass through?
Interesting.

 

[DKGrimm]

Thanks Don.
I'm confused by the voltage spike at about 55us or so without any corresponding current?
Would that be because the conduction (plasma/arc - whatever) extinguished before it came along so nothing for current to pass through?
Interesting.

I'll vote for your answer, but I don't really understand it myself. It's interesting to note that the ringing has two distinct frequencies superimposed. Also, note current scale. That's 3/4 of an amp peak spark plug current! That's more than an order of magnitude above what you see from a coil. That's consistent with the low voltage across the plug during the arc: more current => more plasma ionization => less resistance.

 

[dsage]

I'm not sure how it all works but it's worth considering that you are starting with 250v ?? on the primary not 12v as in a normal coil (is that your order of magnitude ?) and if the small S/S HV coil has the same turns ratio as a car coil then you would expect a higher VOLTAGE on the output. I can't speak to current because I don't know the output resistance (??) of the S/S coil. But a car coil is maybe 2500 ohms? The S/S coil?? Might be worth measuring.

In explanation of the voltage spike (maybe). It is common for a waveform to jump up in voltage when the mixture is lean in a cylinder and the arc extinguishes becasue of the lean condition leaving residual energy in the coil. Since you are operating just in air I wonder if a smaller sparkplug gap would sustain the arc longer and reduce/eliminatethe spike.

 

[HMEL]

Oh No Mr Bill!, here we go again!, folks that don't appreciate the mysteries of the universe
might want to skip over this post!

the energy in a magnetic field per unit volume is
E = 1/2 B^2 / Mu,
B is the magnetic field (flux per unit cross sectional area),
Mu is the permeability of the medium

the permeability of iron is roughly 3 x 10^-3 Henries (times length divided by area)
the permeability of air is roughly 1 x 10^-6 Henries (times length divided by area)

so the energy in an air gap is larger per unit volume than in iron,
but the total energy depend on the total volume of each (intuitively, it takes more
energy to create a magnetic field in air than in iron, so there is more energy stored
in air than in iron given the same field strength and material volume, though in a transformer,
or generator, or alternator, these are never the same volume).

air gaps are typically very thin, compared to their cross section, they contain very little
volume, but this is measurable and the contribution to total energy is calculable, and
hence the division between energy in the iron core and the air gap is calculable, though
as DK points out this division ratio isn't really important.

but the air gap geometry, typically being thin with a large surface area, means the fringe
lines of magnetic field that bulge out around the gap are negligible, IE its not the distortion
of the field that matters.

it is my belief (I could be wrong here, might need some myth busting !) is that the reason
for the air gap in a flyback transformer is to avoid saturation of the core, so that its response
to the applied ampere-turns is more linear which makes the oscillator easier to design.
An additional reason is to get more energy per unit volume, reducing cost of materials (this
belief is backed up by google searches).


now back to CDI vs LDI, if DK needs to increase the inductance of the coil in order to
lengthen the spark's burn time in CDI, then an air gap in the coil is the wrong direction to
be looking since an air gap decreases the inductance of a coil (which can be compensated
for with more turns of wire, but then that multiplies how many turns needed in the secondary
so becomes a cost-of-materials, and overall size and weight of coil, issue).

I have played with flyback transformers, but have never designed the oscillator from scratch,
so I guess I need to do that some day to add to my knowledge of the mysteries of the universe.

You can derive mathematically that the magnetic energy is stored in the air gap and only there. The air gap can consist of a voids in the material or a created one. There is a difference in a saturated core and an unsaturated core. There are two types of transformer designs and they both have different purposes. So I am attaching some lecture notes you will or should find interesting. It talks briefly about this subject. Some references talk about distortion of the field and the volume available is not just in the physical area from the volume removed. If you want the high voltage to the spark plug it worth while looking at the right transformers.

 

[Zeb]

Here's a good example from my test of an early S/S module that I run on my Otto-Langen engine.

View attachment 144252
Don

Pardon my ignorance. I'm noticing potential (across the plug gap?) after the 50 microsecond mark but no current flow. Is this indicating the energy (ExI) is dumping off somewhere other than the plug gap, such as through heat, coil, or other losses?

 

[propclock]

Just my 1.414 cents worth. I have used S/S cdi units and they work well, my experience is
the sparks tend to go their own way inside a small distributor. I am not sure if it is rise time ? just personal experience. I have a LOT of engines. I agree with Sage that the ignition must be small. I use the Sage-Gedde system almost exclusively. The ford 508 coil is ~5$ I make
the electronics on a SMD board ~same diameter of the coil, if the mounting lug is cut off
the total package is small. SMD pcboard mounted to back of coil. The 508 coil
has very low primary resistance so the current limiting is based on the ESR of the batteries.
I use AA NIMH and add a capacitor on the Sage-Gedde power supply after the supply diode.
This is needed because the battery voltage drops to ~2-3 volts when the IGBT is on.
I also use reed switches exclusively. No current draw when off and hard to kill.
This system has been used on hit and miss up to blown V8's at ~8krpm.
For light weight there is nothing I know of that is equivalent to the S/S or Rcxcel
But I don't fly my engines. SMD electronics is not for everyone I know, but not difficult
IMO. I think the coil availability/supply question is the biggest one.
I am still fascinated about the CDI/Kettering results. Thanks Don. I will post pictures
if any interest.