Hey Dave,
When you replace a darlington transistor in the ignition driver how do you handle the select at final assembly trim resistors like R508, R526 (sets the peak current to the coil) and R517 (sets Vcb to let the darlington turn off faster) like they do at the factory?

Folks,

I just read this one. Wwest is correct. With a CD ignition that is designed to recover energy stored in the coil that would otherwise be lost (many did not do this however), current draw goes up with cylinder pressure. This can be simulated by increasing the size of the spark gap so that the energy demands are greater. With the CD ignition I build, current draw increases slightly with widening of the gap until the increased energy demand increases beyond the capability of the power supply at very high rpm. It's not a lot with this one; about 100mA superimposed on the normal increase in current with rpm. This ignition tends to wring the capacitor nearly dry with a few extra oscillations(not single phase like MSD, but alternating current) so increasing cylinder pressure or gap width simply means one or two fewer oscillations with the energy consumption nearly constant per spark event. This tends to cancel the effect to a certain extent, but it still exists. With some designs, I'm sure the effect is much greater. Wwest was correct on this. Fred
With an inductive ignition the current demand does not change, but the spark duration certainly does get shorter with a wide gap or higher cylinder pressure.

Dave this sounds like a Loren statement about Bosch CDI.

Bumping this back on top. No technical input in last post by dicklague

I just read this one. Wwest is correct. With a CD ignition that is designed to recover energy stored in the coil that would otherwise be lost (many did not do this however), current draw goes up with cylinder pressure. This can be simulated by increasing the size of the spark gap so that the energy demands are greater. With the CD ignition I build, current draw increases slightly with widening of the gap until the increased energy demand increases beyond the capability of the power supply at very high rpm. It's not a lot with this one; about 100mA superimposed on the normal increase in current with rpm. This ignition tends to wring the capacitor nearly dry with a few extra oscillations(not single phase like MSD, but alternating current) so increasing cylinder pressure or gap width simply means one or two fewer oscillations with the energy consumption nearly constant per spark event. This tends to cancel the effect to a certain extent, but it still exists. With some designs, I'm sure the effect is much greater. Wwest was correct on this. Fred
With an inductive ignition the current demand does not change, but the spark duration certainly does get shorter with a wide gap or higher cylinder pressure.[/QUOTE]

I use the Daytona-Sensors CDI model CD1 in my 1973 911 2.7 MFI and love it. The draw is claimed to be a less than 5 amp current draw at 8,000 RPM. Because of a highly efficient switching power supply based on US Patents 6518733 and 6636021. Specs show a 135 mJ spark energy output. It comes with a matching coil and is Mil Spec compared to MSD and others.....Very very happy with it after 7 years of use.

Highly recommended. Lots of features.

http://forums.pelicanparts.com/uploa...1466184983.gif

Yes, he is always correct on everything, all the time.;)

Unfortunately, he had little respect for others differing views and he got 'personal' very quickly. If you check your first post, he 'has a go' at you citing his usual 'lack of knowledge / understanding'.

In this particular thread, I actually tested (e.g. on a real life actual CDI box in the real world ) and could measure no difference in current draw. I did exactly what you described, opened the spark gap massively to almost 3mm, no measureable difference. Then, followed the tirade of abuse...

I think his lack of composure ultimately led to him being banned from this forum. This is a shame, since he did have some knowledge.

Respect to your dad for being the CDI pioneer. :cool:

dicklague, I'm sure the Daytona Sensors is a good system. However, it will not have 135mJ to the spark. They will be quoting the energy in the discharge capacitor when it is fully charged. Actual energy to the spark is probably less than 25mJ. Also 5 amps at 8000rpm is not particularly efficient for the quoted charge on the capacitor, assuming single spark at high rpm. The power supply in my unit approaches 80% efficiency which is extremely high for the component count, and it won't run away at high supply voltage. At 8000rpm my unit draws just under 3 amps. If I were to follow the advertising hype of other companies and quote the energy storage in the capacitor and declare that as spark energy, mine would have 100mJ. Fred

I think I missed something here Fred....What is YOUR unit? I am not sure what you are pointing at.

I have not tested the amperage needed or the output on the Daytona-Sensors. I just know that it really improved the performance of my MFI equipped 911, and it has been very reliable and from my observation very well made [and made in the USA].

Yes I quoted the claimed specs, but I do know this company as a conservative no nonsense operation.

I think I will bench test amperage and output in the near future to end all this bickering about specs.

I this a
"shootout" of spec sheets, or are we looking for the best performance? I am getting confused.

Jonny, Thanks for that. My father didn't profit from his invention but certainly lots of companies did and the lawyers on both sides (I have nothing against lawyers by the way, with my son in law school). Had a look at your website. 6 layer boards? My head is swimming already. My circuit is a single layer with a few external components connected by wires! Simpler than the Bosch with even fewer components, and universal polarity with a long duration and voltage controlled (unlike the Bosch). Anyway, it just proves there's more than one way to skin a cat! Fred

Yes, 6 layers. 3.3V microcontroller living in the same box as a 300V supply needs very careful layout for noise immunity.

We have the entire circuit including the transformer and main charging caps on a single board so we don't have the luxury of three dimensions or any peripheral wired components attached to the case for example. Remember also we have a jog wheel, a programming connector and I/O including signal conditioning / driver circuits as well.

There's a lot going on in a very small space. Most programmable ECUs are double the footprint of a standard CDI case.

Edit. Our entry for the CDI current shootout below!

Classic Retrofit CDI+ Current, approx:

2 amps @ 1000 RPM, eight sparks
3 amps @ 4000 RPM, four sparks.
4 amps @ 8000 RPM, two sparks.

~100mJ cap energy per spark.

Jonny H.,

"Dave" us very knowledgeable he actually repairs Bosch CDIs and defends them as being all you need at any chance he gets in this forum.

It is my belief that "Dave" is actually a gentleman named Loren who had been around Bosch CDI for years.

Dave/Loren is always asking for dyno comparisons of CDIs I hope you publish your results as well.

The results are coming in. We are trying to get dyno data for all CDI cars from 69 to 83 (+ the Turbos ). We hope to then publish ignition maps on our website.

The latest result from 911 Design in Montclair, CA. Independently conducted test on a rebuilt but stock ’69 S. See here:

CDI+ Dyno Results

Are you talking about the energy stored in the leakage inductance of the coil? Is that significant?

Another thread you talk about AC waveform to the spark (multiple sparks?) Is this due to the leakage inductance and the capacitor? Is it after the spark event?

I wouldn't think leakage inductance helps much and is likely wasted for the most part with Cdi. However, there is energy exchanged between the coil and the capacitor, when you consider that an inductor opposes a change in current flow because of the primary inductance. The energy comes from somewhere and it is from the discharge capacitor and the power supply which never quite shuts off, but nearly shuts off when heavily loaded. Once the initial spark is struck, the capacitor recharges in the opposite polarity and then the SCR shuts off. Another spark is struck with lower energy and lower voltage on the next discharge of the capacitor as the current swings once again back through the rectifier bridge recharging the capacitor to the same initial polarity. This is where most using that design shutdown. With my design, there are two circuit modifications that allow the power supply to continue to feed the spark on the positive going portion despite the SCR being shut off. The output still decays with every oscillation but the oscillations continue for a few cycles longer with most coils giving out progressively weaker sparks of opposing polarity until the energy is too low or the voltage is too low to form another spark. If I told you that I understood the process completely, I would be lying, but I can observe the results and make a determination of spark duration. I have tested other ignition systems that were based on my father's design (and others) and most are less efficient because of poor matching of components. The open circuit waveform with this CDi is a sine wave with a given coil, with a frequency just short of the natural open circuit coil frequency (which is also a sine wave). Most CD ignitions do not output a sine wave in the open circuit condition. I believe that matching (or nearly matching) the circuit to what the coil already wants, increases efficiency immensely. Fred

I am just trying to figure out how this works to make comparisons. In all previous discussions there were three personalities butting heads and if you didn't agree completely with any of them you were an idiot.

In the bottom CDI schematic the significant part is the primary regulated flyback that charges C5 up to about 370 Volts The thyristor triggers and applies -370 Volts to Vp of the coil model.
In the model below the coils mutual inductance is 8.5mH, and the primary resistance is 2.1 ohms and the leakage inductance is not listed anywhere. a is 1/100 so the secondary should not be significant.

Here is where I get sketchy.

When the Thyristor fires Ip quickly approaches 20 Amps as the energy from the Cap is transferred to the spark gap through the ideal transformer apparently limited by the primary leakage inductance, spark Voltage and spark impedance. The 20 Amp peak is at the zero crossing of the voltage in Jonny H's picture.

Where does the energy to charge the capacitor come from? Is it Lm or Lp? It is only 1/4 the energy originally stored on the capacitor (200 Volts) but seems like a lot based on the topology. Does the spark extinguish when the coil reaches +200 Volts? I think the cap voltage drops to zero with the first flyback pulse.

Is there a schematic of your box anywhere or a patent number I could get?
https://upload.wikimedia.org/wikiped...39/TREQCCT.jpg
http://forums.pelicanparts.com/uploa...1138499120.jpg

Rick, My brain is far less than a Cray computer. I've seen identical schematics that behave completely differently because of component value choices and layout. My suggestion, if you want answers to your questions is to build and experiment. You may not know why something works, but you will soon know what does work. I can't look at a schematic from 40 years ago and know exactly what is going on even if I knew the exact electrical parameters involved. Really, when it comes to these designs, as simple as they may seem, the only way to know the truth is by testing, testing, and even more testing. Fred

Interesting.

Memory Lane
In this schematic is the energy recovered the the current through D3 and D4 where on the CDI waveform (Johnny H's picture) the voltage drops to zero (the flyback only has a single diode) the full wave bridge on the Royer oscillator allows it to pass through? Is that a significant amount of energy?

http://forums.pelicanparts.com/uploa...1466316381.jpg

http://i804.photobucket.com/albums/y...psg7pnp7nd.png

Yes, The return path from ground is through the parallel/series paths of D1,2,3,4. The energy recovery is significant and varies with the setup of other ignition components. The return path should be as low a resistance as possible. Some designs bypass the coil on the return. I've tried that and it makes for a weak spark which should tell you there is a lot of energy left over after the initial strike. I've seen a few designs where the return was through resistors and in one circuit I saw recently, and inductor as well. That would certainly set the capacitor up properly for the next charge cycle and recover some energy, but it is a wasteful way of doing so. Fred

You haven't answered your question. You should be able to calculate that from the
the known values, i.e. the waveform period & capacitor.