Ignition circuit help

[stevehuckss396]

Steve:

Did you (or could you) try the dual output coil by:

Use a spark plug in free air (not in the engine)
Connect a clip lead from one HV lead to the tower of the plug
Take a second clip lead from the other HV lead to the threads of the plug. Use no grounds on the plug.

Yep, did that

 

[Lakc]

Did you want me to bring this coil to you? I can drop it off anytime this week. I have a brand new one still in the box.

Not necessary at the moment, but if you have an itch to see if that coil will spark my mk1 ignition seems to make sparks outside of the combustion chamber all day long. Your welcome anytime.

 

[Les_S]

Steve,
I hope you get the ignition going. I know you are close. I wish I had your coil to play with because that is the greatest unknown at this time for me anyways.

Sage, I think you are nuts on with all of your comments. I like your idea of an IGBT but it looks like to me that Steve made a good choice for an output transistor. I just can't tell what might be going on.

Luc, You are a very smart man and I appreciate your input. I haven't thought about a 741 op amp forever I'm sure there is a thing or two I could learn from you.

Gentlemen, I have learned much from everybody involved. I think it best that I bow out now and wish Steve the best. You are close my friend, make some sparks

 

[stevehuckss396]

Steve:

Arrange a voltmeter across the primary of the dual output coil (+/- leads), preferrably with clip leads so your hands are free.
With the magnet NOT activating the hall sensor you should measure zero volts.
Activate the hall sensor with a magnet JUST LONG ENOUGH to take a stable measurement on the meter.
What do you see?

I think you said the resistance of the coil primary was 0.8 ohms?
Did you take into consideration the resistance of your meter leads? Short your leads together and take that measurement and subtract it from your coil measurement. Why? because the meter leads might measure as much as 0.2 ohms themselves.

I don't remember. I was getting voltage one way and not the other but I cant remember if the voltage would rise on magnet presence or fall. I also checked the hall sensor output. It measured 12 volt until magnet presence and then fell to almost zero.

I just double checked. The leads shorted was 0.2 ohms and across the coil was 1.0.

 

[dsage]

Lakc:

LOL

I've been there done that. I designed a really nice CDI ignition module and after getting it to produce (as you say) "glorius sparks" I also blew it up quite gloriously. Changing the PS switching transistors in my (crammed) board layout was a real pain.

It was at that point I figured out that I would quit trying to impress myself with big long sparks and that big long sparks are a BAD thing .
I added a "protection gap" on the HV lead to ground about maybe three times the gap of a typical plug to limit the spark voltage in case the plug(s) go bad or the HV lead falls off. (You could also damage the coil with internal carbon traces between windings)
(not that you don't already know any of that)

In this case bigger (longer) is not better.

I now listen for nasty sounding and really hot looking sparks at a standard spark plug gap. Especially impressive at high frequencies.

I'd really be interested in your design. Mine did not include a uProc.
Are you using the uproc in the power supply or the input side?

You're a brave man designing anything with HV or high frequencies which includes a uProc. It can be a real nighmare chasing down noise issues. If you're having success my hat is off to you. Good board layout is a science and a must (as you've obviously figure out ).

Although working really well, I set mine aside on the bench for further development. It was a bit large for a model engine use. But much smaller than a typical MSD unit. I need to scale the PS section down a quite a bit. More power than necessary - takes up too much space.

Having said all of that, if Steve's coil is FooBar even a CDI is not going to help. Or are you figuring something else about the coil??

Sage

 

[stevehuckss396]

Having said all of that, if Steve's coil is FooBar even a CDI is not going to help.
Sage

I have another new one. It is brand new in the box. If Lakc can get it sparking, i can just terminate the ends and put the boots on. If Lakc blows it up I could care less. There's always plan C

 

[dsage]

What !!! - you have another coil
Why don't you try the new coil on your circuit. Maybe you just got a dud (understandable from Ebay me thinks)
I can't imagine how the circuit could be damaging the coils since the worst you can do is put the coil right across 12V and that's what's it's designed for. (Hmm - it is a 12v coil isn't it?)
It would have damaged the old-school coil.

If you have no HV boots, temporarily use short clip leads on the HV leads to the plug (be careful, don't leave them open circuit).

I think you're on the right track using this type of coil since you've done a lot of work to arrange for it and it looks great.

I wish I had more suggestions.


Sage

 

[RamTurret602]

Hello Everyone,

I am new here. I assume that you folks are talking about ignition for model engines? If so anyone using CDI?

 

[Lakc]

Lakc:

LOL

I've been there done that. I designed a really nice CDI ignition module and after getting it to produce (as you say) "glorius sparks" I also blew it up quite gloriously. Changing the PS switching transistors in my (crammed) board layout was a real pain.

Still on the proto board right now, only 2x the noise issues like that.

I'd really be interested in your design. Mine did not include a uProc.
Are you using the uproc in the power supply or the input side?

This version runs the whole shebang, generates the pwm for the flyback coil, measures the high voltage for cutoff and hold, watches the trigger and fires the scr. The previous version used another controller chip to handle charging the capacitors.

You're a brave man designing anything with HV or high frequencies which includes a uProc. It can be a real nighmare chasing down noise issues. If you're having success my hat is off to you. Good board layout is a science and a must (as you've obviously figure out ).

Ignorance is often equated for bravery. I came into electronics backwards, as an auto tech and then a Ham. I know enough to get myself into trouble, and still have a hard time thinking of 30khz as high frequency.

Having said all of that, if Steve's coil is FooBar even a CDI is not going to help. Or are you figuring something else about the coil??

Sage

Steve has two new coils, I dont think they both are bad. His trouble with the resistance readings make me wonder if they have an internal clamping diode. Although I have never seen that done, it does not seem out of the realm of possibility.

 

[canadianhorsepower]

Les_s: thanks for your comment

Steve: I have two more sugestion for you
#1 would be to simply replace your coil with a led and a 470 ohm resistor in series and slowly turn your motor
by hand. this would tell you how long your charge time is. the led should be off only when the magnet is in front
of your hall sensor .
#2 this is the method "if in doubt use C4"Rof} use a variac and increase the input voltage and use a 120 volt house swithc
to control your pulse. In crease the voltage more and more and look what the outcome is. I know for a **** that some miniature coil
need 200 volt on the input to work could be you case

And if plan C doesn't work.... so what their is 23 other letter in the alphabet Thm:

 

[Lakc]

Hello Everyone,

I am new here. I assume that you folks are talking about ignition for model engines? If so anyone using CDI?

Welcome aboard!
This thread has been mostly about one particular engine and application. I dont recall a lot of CDI discussion around here. I do have one under development.

 

[Lakc]

Les_s: thanks for your comment
#2 this is the method "if in doubt use C4"Rof} use a variac and increase the input
And if plan C doesn't work.... so what their is 23 other letter in the alphabet Thm:

Plan D is "if in doubt use a 3-500z"

 

[dsage]

It sounds like all of you are thinking that maybe the coil was originally designed for CDI activation. Good thinking. Canadianhorsepower may have a good idea there in so far as applying an AC waveform to the primary - even if it's from say a 12VAC transformer - if a variac is not available. I think what he's suggesting is that if you had a scope (or a meter depending on the voltage produced) on the output you could measure the output and determine the turns ratio. Treating the coil more as a transformer not as a high voltage generator just for investigation purposes. Not a bad approach since the coil does seem to be odd.
I would assume a coil designed for a high voltage input would have a low turns ratio. Of course I'd only be guessing what a typical coil turns ratio would be - maybe 100? You may have to do the same thing with the old-school coil and compare them.

Are there any identifying numbers on the coils that can be traced to a particular manufacturer. Which might lead to what they were used on and further what they were firing it with?

Probably grasping at straws with all of this but certainly it seems the coil is somehow different than a typical one.

I guess success or failure using the CDI will tell it all.


BTW - what's a 3-500Z ?

Sage

 

[Lakc]

 

[Mosey]

Duh,
Here's an idea called "I don't know nutthin".

Dump that coil and try one you know to be good. I'll let you know where to send the check later.

Mosey

 

[dsage]

Lakc:

Transmitter tube - right?
I guess if you had the plate voltage for that baby you wouldn't need the coil. Just big relay. I'm not sure how long the sparkplug contacts would last though.

I think you said you had your CDI working (sort of) on the bench with a regular coil - am I right?
Maybe you should just hook up Steve's coil in place of yours and see what you get. The CDI circuit need not be finished if it proves the point and gives some sort of spark. Especially if Steve gets diddly.


Mosey:
Apparently Steve has a couple of more coils he could try. It's possible the one he's been trying is a dud but considering that they are all from the same batch they are probably all in the same boat (good or bad). It's likely we're missing something about driving it properly.
He has tried a conventional coil and apparently it works.

Sage

 

[canadianhorsepower]

The magnets are 135 degrees apart because the combustion cycle of the engine is 270 degrees apart.

Pop - 270 - pop - 450 - pop - 270 - pop

Steve, I'm totaly lost where does that 450 comes from. I must assume that your crank is like a V-TWIN motorcycle

BTW are you trigging that hall sensor with a south pole and with ground are you using to check your spark your primaty
or your case ground

 

[Lakc]

Lakc:

Transmitter tube - right?
I guess if you had the plate voltage for that baby you wouldn't need the coil. Just big relay. I'm not sure how long the sparkplug contacts would last though.

Yes, good for a kilowatt or so. Make the inductance of the piston mistune a rf circuit and it oughtta spark on its own right.

I think you said you had your CDI working (sort of) on the bench with a regular coil - am I right?
Maybe you should just hook up Steve's coil in place of yours and see what you get. The CDI circuit need not be finished if it proves the point and gives some sort of spark. Especially if Steve gets diddly.

Indeed, if Steve wants to make the trip, the offer is open.
If all goes well, version 2 of my CDI is only a few hours from working itself. Of course, two hours of "going well" can sometimes take weeks...

 

[modelman1838]

Hi
Last week I decided to give my old Wyvern IC engine which has been gathering dust a makeover and give it a run. The original coil and buzz circuit that I had used was now used on another engine so I had to look elsewhere. Not wishing to spend much money since I would only run it once, I purchased a " Masterclass" electric gas lighter for £7-69p from ebay. I took it apart and removed the switch part and connected the microswitch which was used for the contact points in its place, attached the spark wire which was`nt even HT to the plug terminal gave the engine a flip and off she went a treat. Not a bad complete ignition circuit for
£7-69p I think.
Hugh

 

[Mosey]

Lakc:

Transmitter tube - right?
I guess if you had the plate voltage for that baby you wouldn't need the coil. Just big relay. I'm not sure how long the sparkplug contacts would last though.

I think you said you had your CDI working (sort of) on the bench with a regular coil - am I right?
Maybe you should just hook up Steve's coil in place of yours and see what you get. The CDI circuit need not be finished if it proves the point and gives some sort of spark. Especially if Steve gets diddly.


Mosey:
Apparently Steve has a couple of more coils he could try. It's possible the one he's been trying is a dud but considering that they are all from the same batch they are probably all in the same boat (good or bad). It's likely we're missing something about driving it properly.
He has tried a conventional coil and apparently it works.

Sage

I don't mean to butt in with smartypants comments, especially in an area where I know next to nothing, but what I do mean is to put a brake on the experimentation by using a conventional device. It's about the engine, and getting it to run. Unless Steve is happy putting in tons of time.
Mosey