Ignition circuit help

[stevehuckss396]

Steve:

I'm pretty sure what you have is a dual output coil used for waste spark ignition. I hope I'm right on this and haven't wasted a lot of effort typing this.

I know it sounds odd but you have to get your head around the fact that the heads / block are not supplying a ground for the spark, just an electrical connection between the two plugs.

Using a CDI unit with a conventional car coil will result in excessive HV which may cause you problems on a model sized engine. 7kv is plenty on our small plugs. 60kv on the other hand will be looking for all kinds of places to jump.


Sage

No you have not wasted your effort. It is in deed a wasted spark ignition.

I am on board with how the circuit works. The engine is just acting as a jumper between the 2 ground straps on the plugs. I considered that when I was testing.

All my ignition components are full size. The plugs are NGK V-Power. I had a few spares from a hotrod that I no longer have. This engine is my old air compressor with 2.625 bore. It has 1.175 inch valves with full size plugs.

Again thanks. I am in the process of redrawing and should post it up real quick here.

 

[stevehuckss396]

Here is the new circuit ALA Dsage.

Now that it is redrawn I will need to make a new board. If I'm going to do that, should I lower the value of R3 to increase current flow in the coil primary?

Yes CHP! I did it in paint!Rof}Rof}Rof}

 

[stevehuckss396]

I have the new board layed out. I won't do anything with it until Lakc weighs in with the CDI circuit. I will most likely build and test both. If they both work I will go with the one with the most spark and keep the other for an emergency board. Nothing ever seems to go wrong until there is a crowd of people watching.

 

[Les_S]

Absolute newby to this forum and I certainly don't want to upset anybody but I beleive, after looking at your schematic, that you have your 2N3906 transistor upside down. Emitter should connect to power supply positive, not the collector.
Best regards,
Les

 

[canadianhorsepower]

Hi Steve, "the artist"
I went to Princess to get a compressor head twin like the one posted
on the other blog that you opened. I started reading comp head specs and
then eek. The MAX RPM is 1900 RPM is it the same for you or is it pure
warning because of the flywheel size What's your opinion on this matter.

Did you mount your hall inside and using the shaft for the trigger

 

[jack620]

Absolute newby to this forum and I certainly don't want to upset anybody but I beleive, after looking at your schematic, that you have your 2N3906 transistor upside down. Emitter should connect to power supply positive, not the collector.
Best regards,
Les

I think so too. But I don't think turning it the right way around will work either. Shouldn't it be an NPN transistor?

 

[Les_S]

A lot of work has gone into this design so again, I certainly don't wish to upset anybody. The circuit will work with the 3906 turned around. When the hall effect switch is turned on by the magnet, the 3906 will be forward biased which will pull the 3906 collector up to nearly 12 V which will in turn forward bias the BU941 and energize the coil building up a magnetic field. Spark will actually occur when the BU941 turns off and the field collapses.
Best regards,
Les

 

[stevehuckss396]

Hi Steve, "the artist"
I went to Princess to get a compressor head twin like the one posted
on the other blog that you opened. I started reading comp head specs and
then eek. The MAX RPM is 1900 RPM is it the same for you or is it pure
warning because of the flywheel size What's your opinion on this matter.

Did you mount your hall inside and using the shaft for the trigger

I believe the flywheel is the limiting factor. Because it is cast and real heavy it will come apart real easy. If it were billit and balanced I bet that number would increase to 6-8K

I have a disk on the tail of the camshaft that has the magnets in it. The sensor is mounted to the back of the junction box and the magnets fly in front of it. There are 2 magnets 135 degrees apart. Here is what it would look like if the jbox were see thru. You can see the circuit board and the disk behind it. The sensor gets glued to the back of the jbox behind the magnet disk.

 

[stevehuckss396]

Absolute newby to this forum and I certainly don't want to upset anybody but I beleive, after looking at your schematic, that you have your 2N3906 transistor upside down. Emitter should connect to power supply positive, not the collector.
Best regards,
Les

Are you sure? The data sheet shows all the test circuits with the collector positive.

 

[Les_S]

Hi Steve,
Let me first say that I really admire your work!

I am sure that the 3906 in your drawing is upside down but I can see how the datasheet can be confusing.

Note the waveform used for the delay and rise time test. The test signal starts @ +.5V at which time the transistor would be off. Then the test signal goes to -10.6V at which time the transistor would turn on because the emitter is setting @ 0V which is way more positive than -10.6V. Same situation with the duty cycle test. For a PNP transistor, the emitter has to be more positive than the base to turn on. For a NPN transistor the base has to be more positive than the emitter to turn on.

Hope that helps. The test circuits show a means of testing transistor operating characteristics. They are not application circuits.

Best regards and I can't wait to see your compressor run!

Les

 

[stevehuckss396]

How about this one? I also flipped it on the circuit board drawing.

 

[Les_S]

That will work. There are a lot of smart folks on this forum and I don't claim to be one of them but the electrons have had their way with me for many years so I'll be happy to help if I can.
Les

 

[stevehuckss396]

That will work. There are a lot of smart folks on this forum and I don't claim to be one of them but the electrons have had their way with me for many years so I'll be happy to help if I can.
Les

Thank you very much. My original drawing was drawn correctly and somehow I turned it around on the new one.

 

[dsage]

Hi Steve:

The circuit in post #71 is almost correct. I would suggest a couple of more changes / corrections.

1. The 0.5 ohm resistor should be in series with the coil on the collector side and the emitter should be grounded....

BUT !!

After having another look at the spec sheet. The power transistor has about a 2 volt drop across it when turned on (Vce under certain conditions). Considering that, with the 0.5 ohm resistor in the circuit the coil may only see 6 volts.
So on further consideration (and considering the fact that you will be diligent in turning off the ignition if the engine stalles etc. etc) let's remove the 0.5 ohm resistor from the circuit and ground the emitter.


2. As you mentioned the 680 ohm resistor could be reduced to say 470 ohms for a bit more drive to the transistor. There is a limit how low you can go here since the 2N3906 has a current limit as well.


It's difficult to get all the values correct just from spec sheets but I think you will be close. I would activate the circuit and take a few very quick voltage measurement. One being how much voltage is actually across the coil and across the transistor.

Did you try the coil / plug circuit testing on the bench to confirm the coil works as expected ??


Canadianhorsepower:

To clarify - What you wrote was

Quote:

"Hi Steve, I did a few test and calls today and I have a few good answers.
1 even if it works you should fire your coil thrue ground and not live.
thats an easy fix replace both transistor with the oposite ( from NPN to PNP)

I'm not sure how you read the above but to me that says you can swap the type of transistors and your good to go. Which is not correct.
If there was more involved (which there is) then you should have explained further. You must keep in mind that you are giving instructions to people that know a lot less electronics than you (or me).

This is why I try my best to explain things thoroughly and burn up a lot of space and time doing so. (sorry for that).


Sage

 

[stevehuckss396]

Ok now you went and used the B word. If the IGBT would be better what IGBT should I look at? If I do change to a IGBT what resistors will need to change and to what values? Remember I am starting from scratch so I would perfer to design this thing to the best it can be with the most current information. As long as I can fit it all into the box. I would have no objection to changing it up if the outcome will be a better design.

No I did not get a chance to do any testing. I did get some spark out of it when fooling with it before.

 

[dsage]

Steve:

I went back and looked at my IGBT spec and it has only marginally better voltage drop at high currents. So stick with what you have. (makes sense since the IGBT is also a bipolar transistor).
The only thing significantly better in terms of voltage drop woud be a MOSFET but it wouldn't have the high breakdown voltage and would (like the plain old transistor in the original TIM5 circuit) supress the primary kickback and reduce the seconday output.

I think you'll be good with what you have. (as best I can tell without testing it).

I would strongly suggest you try it out on the bench even just "dead bug style" (a term used to describe cobbling it together without a circuit board) to be sure it works before you take the time to make a circuit board. Maybe you could borrow a proto board to wire it together.

Please re-draw the circuit (again) for the benefit of us all.

Thanks

Sage

 

[canadianhorsepower]

Hi Steve:

The circuit in post #71 is almost correct. I would suggest a couple of more changes / corrections.

1. The 0.5 ohm resistor should be in series with the coil on the collector side and the emitter should be grounded....

BUT !!

After having another look at the spec sheet. The power transistor has about a 2 volt drop across it when turned on (Vce under certain conditions). Considering that, with the 0.5 ohm resistor in the circuit the coil may only see 6 volts.
So on further consideration (and considering the fact that you will be diligent in turning off the ignition if the engine stalles etc. etc) let's remove the 0.5 ohm resistor from the circuit and ground the emitter.

BTW this is another reason the IGBT transistor (I use in my circuit) would be better. It has almost no drop when properly turned on and so all the battery voltage would be available to the coil. But lets continue with your circuit.

2. As you mentioned the 680 ohm resistor could be reduced to say 470 ohms for a bit more drive to the transistor. There is a limit how low you can go here since the 2N3906 has a current limit as well.


It's difficult to get all the values correct just from spec sheets but I think you will be close. I would activate the circuit and take a few very quick voltage measurement. One being how much voltage is actually across the coil and across the transistor.

Did you try the coil / plug circuit testing on the bench to confirm the coil works as expected ??


Canadianhorsepower:

To clarify - What you wrote was

Quote:

"Hi Steve, I did a few test and calls today and I have a few good answers.
1 even if it works you should fire your coil thrue ground and not live.
thats an easy fix replace both transistor with the oposite ( from NPN to PNP)

I'm not sure how you read the above but to me that says you can swap the type of transistors and your good to go. Which is not correct.
If there was more involved (which there is) then you should have explained further. You must keep in mind that you are giving instructions to people that know a lot less electronics than you (or me).

This is why I try my best to explain things thoroughly and burn up a lot of space and time doing so. (sorry for that).


Sage

dont be sorry just don't be a smart ass or an ass hole what ever you prefer.
I NEVER MISSED LEAD SOMEONE, I WILL GIVE OPINIONS ,YOU NEED ADVICE ,THEN I WILL STEP UP AND BOW OR STAND FOR WHAT I'M TALKING ABOUT...................................YOU SHOULD TRY THIS

I did mentionned before I don't put BS I'n my response to someone to look better then someone else. I hope you would realize that all the BS you did put so far only got THIS FORUM to a big pile of ****. Steve had a borderline
working circuit "changing 1 resistor" would have work. Now he's starting almost from the start. Thank you:rant:
Hope you have enought BALL'S to help ???????? Steve now that almost no one whan't to get involve in this forum:shrug:
When I whent with my cancer surgery they had special equipement to keep your scare together . maybe you should look for one that will keep your chest smaller./............. if it does not work try one for your head >>>>>>>>>>>>>>>>>

 

[stevehuckss396]

Steve:

I went back and looked at my IGBT spec and it has only marginally better voltage drop at high currents. So stick with what you have. (makes sense since the IGBT is also a bipolar transistor).
The only thing significantly better in terms of voltage drop woud be a MOSFET but it wouldn't have the high breakdown voltage and would (like the plain old transistor in the original TIM5 circuit) supress the primary kickback and reduce the seconday output.

I think you'll be good with what you have. (as best I can tell without testing it).

I would strongly suggest you try it out on the bench even just "dead bug style" (a term used to describe cobbling it together without a circuit board) to be sure it works before you take the time to make a circuit board. Maybe you could borrow a proto board to wire it together.

Please re-draw the circuit (again) for the benefit of us all.

Thanks

Sage

Give me about 30 minutes!

 

[stevehuckss396]

This is the latest.

 

[stevehuckss396]

I noticed that the IGBT is surface mount. How is the device heat sinked? Does the board act as a heat sink? If so it might be worth going that route rather than the remote mounted transistor. Would make things easier and in the end, cleaner looking.