A new ignition circuit

[dsage]

Just a couple of notes on your testing.
First off thanks for the results.
A conventional ignition circuit requires a high wattage ballast resistor in series with the coil. Usually a couple of ohms or so to limit the current. A full sized car coil by itself is only a few tenths of an ohm and on 12v you could have at least 24amp pulses trying to be produced by your supply. (~ 12v /0.5 ohms). Probably more. For that reason a power supply is not usually a good thing to test with.
Having said that - you can THANK your power supply for saving your coil, the driver circuit and probably your wiring. Had you been using a good battery you might have smoked it up (with no ballast resistor).
You mentioned a few posts ago that you thought your point closed time (dwell time) on your "old iron" engine was REALLY REALLY long. A full sized coil and even the small model coils never require more than 20ms to be fully saturated. Any more time energizing it than that and you're just wasting power heating the coil, the ballast resistor and the driver transistor and stressing your points (on a standard ignition). Also killing your battery prematurely. This is where the 20ms came from in the original timing values. No point in powering the coil any longer than it takes to saturate it. Your just wasting power.
So if you want my opinion you should spend some time looking at adjusting your points dwell (closed time) to 20ms (or so) at cranking / idle speed for a low revving engine. It's probably just a screw on the points set.
Yes that time will become less as the engine speeds up but a full sized coil will produce a good spark at less than 1ms dwell time.
The forgoing are all typical values and user results may vary. But this is why ignition circuits are different on different cars - because manufacturers choose values that work for their coil, expected rpm, etc. etc.
Using the ignition circuit was not intended to be "plug and play". Some pre-work needs to be done to tailor your trigger signal for the intended usage.
i.e. points adjustment, magnet size, expected rpm etc etc.

BTW. The capacitor size vs time values you came up with from your testing are pretty close to expected. As I mentioned in another post I expected 5x the capacitance would give you 5x the time. Given component tolerances that's pretty much what you got. (18ms for 1 uf and about 90ms for 5us). Give or take.

Sage

 

[willray]

Just a couple of notes on your testing.
First off thanks for the results.
A conventional ignition circuit requires a high wattage ballast resistor in series with the coil. Usually a couple of ohms or so to limit the current. A full sized car coil by itself is only a few tenths of an ohm and on 12v you could have at least 24amp pulses trying to be produced by your supply. (~ 12v /0.5 ohms). Probably more. For that reason a power supply is not usually a good thing to test with.
Having said that - you can THANK your power supply for saving your coil, the driver circuit and probably your wiring. Had you been using a good battery you might have smoked it up (with no ballast resistor).

Ah, I used a low-ohm dummy load, not a no-ohm dummy load... I'm not that much of a dummy This power supply isn't capable of enough current to do much damage regardless, but in this case it was definitely not helping.

And yes, I expected that the timeouts with different caps would end up where you predicted them to. I just wanted to confirm that it really was so, and perhaps save someone else the trouble of doing the math if they wanted to select a longer timeout.

 

[tightwadracer]

Hi
Firstly I must congratulate you on doing so much fine work and for sharing it.

I am a Technology teacher in Australia and 2 of my students are building engines which will require this ignition system, I have been building a proto.
However I was wondering if you have a PCB layout available to share as I see more of these motors will be built in the future.

 

[dsage]

How do you plan to fabricate the boards?
Model Engine Builder Magazine Issue 34 had an article with pdf files you could use for the toner transfer method of both sides of the board. Component lists, schematic and stuffing diagram was also in the article.
There was also info on how to calculate magnet size etc. for your particular engine to properly trigger the unit.

Sage

 

[tightwadracer]

I have contacted the magazine in relation to the purchase of issue 34 and have had very immediate and positive response.
We have a few options for manufacturing PCB, including toner transfer.
we may even be able to use our four axis hartford machine centre to engrave them.
When I get time i will upload some photos of our fantastic workshops.

Thanks again

 

[canadianhorsepower]

Ok so if i go this way

7) IF A HALL SENSOR IS NOT BEING USED, R2, C1 AND E6 CAN BE DELETED

I assume R6!!
what do I do with D3 ?????
and R3 goes directly to the base of Q2
thanks

 

[dsage]

Refer to the diagram in post #47
If a hall sensor is not being used you can delete R3 and C1. (but IMHO for all they're worth you might as well keep them in case you want to switch to a hall sensor).

Everything else must stay.
If your using points then they must switch to ground when closed and then they can be connected to E6 connection (Labelled HED IN which goes to R7).
For easy connection (rather than finding a ground elsewhere) you can connect the points across E6 and E7
Don't use a condenser across the points (like in a car).

Hope that helps.

Sage

 

[paulc]

Hi Guys, I'd like to give this design a shot. Is there a basic vero board layout or similar available?

 

[dsage]

EDIT:
See message 211 Below for update.

Here is a circuit board layout that you can translate to a vero board. It's double sided but you should be able to work that out on vero board. Also attached is the simpler schematic. It's electronically the same as the original but simplified to reduce the board complexity. You might want to check with the Editor of Model Engine Builder Magazine. I gave him all the circuit board files and he was investigating getting some made. I'm not sure if he followed through.
I don't have any circuit boards to offer.
A complete write up and the files to make your own circuit boards with toner transfer was featured in Model Engine Builder Magazine #34 for the original (more complicated) circuit You might want to ask the Editor about a copy of that edition.

 

[paulc]

dsage,
That is very kind of you, perfect!
I went ahead and ordered back issue #34 MEBM.
But in the mean time........
The PCB pdf you attached the top side traces, in red are understood but the bottom side of the board (blue traces) is partially obscured.Would you by any chance have the layout of the bottom side to share here too?

I will post my results here. I actually have a 7 cylinder (350cc) radial that I scratch built in 97' it uses a kettering system (12 volt TIM Bosch module off an old VW Golf) and my own hall sensor picking up sesing 7 rotating magnets, it runs well, once I get her running again I will post here. This is this engine I want to convert to this system. I would like to retain a 12 volt supply.

Paul

Here is a circuit board layout that you can translate to a vero board. It's double sided but you should be able to work that out on vero board. Also attached is the simpler schematic. It's electronically the same as the original but simplified to reduce the board complexity. You might want to check with the Editor of Model Engine Builder Magazine. I gave him all the circuit board files and he was investigating getting some made. I'm not sure if he followed through.
I don't have any circuit boards to offer.
A complete write up and the files to make your own circuit boards with toner transfer was featured in Model Engine Builder Magazine #34 for the original (more complicated) circuit You might want to ask the Editor about a copy of that edition.

 

[dsage]

Here are the top and bottom separately and the schematic. Please double check the connections and the part numbers vs the schematic. I have several versions of the board and although they are all functionally the same the board layouts and part numbers may have changed.

 

[paulc]

I think you may forget to attach your files?

Here is what I interpret the bottom (blue) traces would look like if seen from the TOP through a transparent board with the red traces on the top side deleted for clarity. But Im not overly confident.

Here are the top and bottom separately. Please double check the connections and the part numbers vs the schematic. I have a few versions of this and although they are functionally the same the part numbers may vary depending on the version.

 

[paulc]

David, This is how I interpret the tracks for both faces of the PCB. ALL seen from the TOP. Can you confirm I have the bottom face blue correct?

 

[dsage]

BTW. If you are using a 12v supply you should use a ballast resistor or a coil that has one built in. Just like a car does.
You will notice there is a bit of space left around the IGBT power transistor on the circuit board. This is to allow a small TO-220 heatsink to be fastened under the transistor. You might find you need the heatsink with high spark rates - which your 7 cylinder will have.

 

[dsage]

David, This is how I interpret the tracks for both faces of the PCB. ALL seen from the TOP. Can you confirm I have the bottom face blue correct?

I think we were typing at the same time and I had deleted the files until all was correct. See the update - post 211 and my comments in post 214.
Sorry for the confusion.

 

[paulc]

Got it! and thank you,
I read in an earlier post in this thread, that there was a component change (a resistor) if one intends to run on 12 volts over 6 volts?

Regarding heat , I intend to make a decent sized aluminum base pate that is bolted to the heavy transistor, the whole show will be potted on top of this plate, the plate will extend past the ends a little as tabs to allow for two small mounting holes. The "module" in this package, will then be mounted to a larger heat-sink via heat transfer compound on its rear metal face, then be placed in ventilated area in the prop wash.

Im ordering the parts now!

 

[dsage]

There is no need to go overboard with the heat sinking. One of the small TO-220 sinks that fit under the transistor right on the circuit board is sufficient. The transistor runs pretty cool even on a V-8 with normal rpm. The heatsink will be live. So keeping it small and isolated will avoid a lot of issues with it touching something. See attached. The 273 version fits on the board.
It's by Wakefield and if you order from Digikey it is part number 345-1022-ND.
Link here
https://www.digikey.ca/en/products/...0320?s=N4IgTCBcDaIMwBYCsBaAjABjGFA7AJiALoC+QA
Concerning component changes, there was some discussion of the circuit self-triggering when used on a large single cylinder hit-miss engine at low hand cranking speeds. In which case increasing C2 might help. But you wont have that problem with 7 cylinders. I don't recall any other issues. It works fine on all my engines as-is.
Let me know if you find otherwise.

 

[paulc]

. The heatsink will be live.

Oh.... the heat sink if connected to the ear of the transistor will be VCC +?
So the heat sink can NOT be mounted on an earthed chassis?

 

[dsage]

Correct. The transistor tab is live.
There is no need for that much heat sinking. But if you must, you could purchase a TO-220 insulating add-on with a mica washer, and insulator for the screw. The small heatsink I showed fits on the board and should be sufficient. Even more so if you can arrange for a bit of air flow.
Keep the whole circuit as compact as possible. Don't use long wires anywhere or you will likely have issues.
Attached is the connection diagram. Again - keep all the wires as short as possible. Especially keep the battery close with a direct (and short and heavy gauge) ground from the engine block direct to the battery negative.
All of the issues I have seen using any coil driver (not just this one) is from folks spreading things out. The whole coil driver board is only a couple of inches square. Keep it that way and connect it with short heavy and direct connections. (Yup I can't stress it enough - four times now)

 

[paulc]

Thanks for the advice, I'll do all that. I'm working on the circuit board print now, going to etch the board using the laser printer transfer method. This will be perfect for this job. Once I have all the bits I'll post some pics to show others how I went.

Also what are the ramifications for having longer leads, what issues does this cause?

And lastly the components you listed on the PCB are ordered with some spares. But what is the difference between the 6 volt and 12 volt set up? Is it just the ballast resistor on the 12 volt version that differs between them?

When I test the 12 volt automotive coil what should I look for and how many amps should this circuit be kept to?