V8 Ignition

[stevehuckss396]

Steve, I cannot find the $58 kit on Roys website.
Cheapest I can find for a kit is $80.
Can you post a link to the page?
Kel

http://www.cncengines.com/ic.html

About half way down there is the "Basic single spark CDI" for 50 bucks and a bit farther down is the hall sensor for 8 bucks.

The sensor is seperate because the module can be triggered with points also.

 

[fltenwheeler]

Steve, I cannot find the $58 kit on Roys website.

Cheapest I can find for a kit is $80.

Can you post a link to the page?

Kel

Single spark CDI module
Spark coil is soldered to power board includes 1 foot spark and ground wire.
Add hall sensor or 2 wire connector and spark plug end for a complete CDI
Does not include a hall sensor, magnet, or 2 wire connector for use with engine points. Please order these items separately.
$50.00

 

[kcmillin]

Thanks guys, How did I miss that. it makes perfect sense now.

Kel

 

[jpeter]

Here's one that works. It'll spin my v8 to 7000 rpm. Been solid for months.
I found the tim-6 troublesome. I designed this one around the IGBT ignition switch. If anybody builds one don't leave out the 20k 1watt resistor in the HT wire. If the hall effect switch is to be in the distributor, I use it inside a 1 inch dia. distributor, use one able to take high ambient temperatures. I use 125 degrees C one, 150 would be better. All the parts come from digikey. I put the sensor and zeners in the distributor and everything else on a circuit board taped to the side of the coil. I cut the circuit board traces with a dremel tool cutoff wheel. Simple as it is I put a lot of time into the development. Had a lot of trouble with EMI and heat. Works now though.

View attachment transistor ignition.pdf

 

[gbritnell]

Hi Jim,
I gave a look at your schematic. For someone that is familiar with electronics it probably makes a lot of sense but for the uninitiated could you elaborate a little more like what kind of coil did you buy or use. The auto parts store isn't very helpful if you don't have a number or model. Next is the component between the signal side of the Hall sensor and the leg of the transistor. Then you have what looks like another voltage source VCC applied to the signal side of the Hall. Would it also be possible to give a list of the component numbers you use. I have a DigiKey catalog and when you start looking for an individual component without a number the options are mind numbing.
I would sure like to give it a try but I'm afraid I can't build one from your information.
Thanks for your help,
gbritnell

 

[gbritnell]

Gentlemen,
Here is a link to another ignition system. I don't know if anyone has tried it but the author, Paul Elsmore, seems to know what he's talking about.
http://hnm.110mb.com/i.html
gbritnell

 

[jpeter]

All good questions. All the voltages are 12 volts. Vcc is connected directly to the 12v battery +. In the distributor is the hall sensor, 2 zener diodes for EMI protection, and the 2 resistors that go with them. I use an aileron extension (3-wire cable) from the distributor to the electronic unit. One wire carries 12 volts, another carries the sensor output and the third is a ground. I devise a small circuit board which I generally mount under the distributor such that the sensor sticks up through the base of the distributor housing into the distributor. I use a shutter to switch the sensor. I've tried using multiple magnets but on small cams the magnetic fields interfear with each other making switching, at least for me, difficult. On large cams I have trouble with dwell timing. The JH website suggests multiple magnets. The igbt is a robust chip. I've never ruined one. They do get hot so I generally fasten them to the circuit board and apply a dab of heat sink grease. Not so for the hall sensor. They can get zapped by EMI and ambient heat so when you buy them, buy a supply. Regarding the coil; for a coil I use a standard old style automotive coil with a built in resistor. I've attached a sketch of a distributor design I use. It works well and the top of the shutter works well too for supergluing the rotor to, which I make from g10 to which I glue a strip of brass.
[ame]http://www.youtube.com/watch?v=B9zt3SF_Flc[/ame]

View attachment distributor.pdf

 

[Admiral_dk]

Hi Jpeter

Looking at you ignition I'm sure that you're having the diode in the IGBT gate connected the wrong way - it's supposed to help with turning the IGBT off as fast as possible ...!

I'm also surprised that you haven't got a strong buffer between the hall-effect sensor and the IGBT - I'm sure that the big heat you're experience in the IGBT is caused by the slow turn on and off times. If you place the diode correctly I'm sure you'll get a stronger spark, because you turn the IGBT faster off - it can be done with two transistors.

 

[jpeter]

You may be right regarding the direction of the igbt gate diode but but then again you may not be. It sparks like hell and I'm not changing it. That portion of the circuit came directly from the IRC data sheet. I added the hall effect sensor and the zeners for EMI protection. Otherwise it's exactly as the manufacture suggested. Google for the data sheet and you'll find they recommend a diode acrossed the coil too. I couldn't get it to spark with that diode in place so I left it off. You know, I never actually understood the purpose of that diode in the gate circuit. I thought maybe it protected the switch or something. It was in the datasheet so I left it there. Come to think of it I did add the 20k pull up resistor and I'm sure the sensor pulls the gate down fast enough to spark the coil with or without the diode. Maybe its not necessary at all. Anyway, the circuit works really well, its about as simple a circuit as there is to build and the parts are really cheap. And, unlike the transistors in the tmi-6 circuit the igbt is really robust.

 

[Admiral_dk]

Well I do understand the concept - don't fix it, if it ain't broke ... ;D

But I'm also interested in improving things, so I went looking for an application note on the IRF homepage - none, googled it - again none .... So I looked at the datasheet and I'm almost certain that you're referring to figure 15 - fine, BUT it's NOT how to use it, but how to measure the device ....

I do not have a Hall effect sensor in my electronic simulation tool right now, so I can only show you how the modified circuit will look like .... that is if I succeed in attach the pdf .... ???

Your 20K resistor is replaced with R2 1K on my schematic, then follows the two transistors and another resistor and the IRFGH14C40L as you use, and the rest is just another way of drawing the same as your circuit. But it should be a much improved circuit .... but then again - it might not perform any better in your application -


View attachment Ignition.pdf

 

[Lakc]

As I look at my tenative order, I see I spec'd an IRG41BC5UD, but mostly for the reason that digikey has them in stock, and eagle has it in my library. I can be notoriously lazy at times, and have not done much hard research into yet, but this part has an integral supression diode already. This is mostly going to be an experiment for me with IGBT's, as I have not used them before. For any ignition system, you cant have enough zener diodes and buffer capacitors, so sprinkle them liberally. This is not a circuit you want to Muntz, or you will make hall effect and microcontroller suppliers rich.

Years of fixing broken ignition systems has given me the itch to try my own hand at designing them. I could just be suffering from an overinflated designers ego, but if I try and fail, I will learn something, so its never a total loss.

Somewhere along the line I need to figgure out how to make ltspice do what I want it to do. Testing anything but the simplest circuits is beyond me. So much to do and so little time....

There are several decent looking, open source, capacitive discharge systems out there. Thats not the direction I want to take, but it is an eloquent solution to a scaling problem if you dont want to use a full size coil. If you only get 1 tenth of the inductance to play with, you just hit the primary side with 40x the voltage.

 

[jpeter]

Hey, I know just enough about electronics to be dangerous. I also know that igbt gates have a lot of capacitence, thus need current or time to charge. I haven't done the math or maybe I don't know how but I was surprised I could charge the gate to 7 or 8 volts quickly enough through that 20k pull up resistor. When I found could clamp the igbt with it I stuck with it cuz I didn't want to drain to much current through the sensor during the pull down period. Possibly the igbt would run cooler with a smaller pull up resistor, clamp it tighter that is.
I haven't yet looked at your circuit but I'm gonna. That electronics stuff fasinates me.
What I'd really like to find is a source for a small 12v coil. Hiding that big clunker in the base is tough on small engines. I have a couple of engines on which I use motorcycle coils. They work too and are a lot smaller, retro ones of course from old kettering ign systems.
Thanks for the circuit. I'll check it out. Say, I've had those igbt's around for a long time. Maybe they aren't still available.

 

[Admiral_dk]

Hey, I know just enough about electronics to be dangerous.

;D Been there and done it too, especially when younger - Errare humanum est ;D

How did you measure the rise time on the gate ? I mean this can (normally) only be done with a oscilloscope, so a rise time of 1 uS or less with a 20K will impress me rather much.

IRF are certainly making them right now and I wouldn't be surprised to find them in a very high percentage of current automobiles and motorcycles - since I haven't found ANY similar product from a competitor, except those how make them under licence (those names vary a bit, but all contain 14C40L in the product number). They are made to be inside those spark plugcaps that integrate the ignition coil inside too. Find one of those and you found tiny ingnitioncoils - I will not promise that they come cheap, but then again - one for each cylinder on a modern engine ought to bring the price down.

I'm sure that you could improve your circuit quite a bit by just changing your 1K to 10-50R, the 20K down to 1K and change the direction of the diode. A capacitor with a value between 47nF and 470nF 630V either across the primary or across the IGBT (from ignition coil to ground), but NOT both - is very likely to raise the secondary voltage quite a bit too.

 

[jpeter]

Hey Admiral, I'll assume your circuit works well but what's better about a circuit that adds 2 bipolar transistors along with two more resistors and stuff to get to the same place. I've got no burning desire to complicate the thing. Disregarding the switching circuit, I've got the igbt, 2 resistors and a diode. Had I chose to use a hall effect switch with a built in pull up resistor I could have cut back the parts list even more. I can't think of a more simple circuit to get the job done.
The igbt from Digikey cost 3.39 in single unit quantities (just looked it up); the sensor cost 1.25. I'm doing TIM for under 5.00. Tough to beat that. I'm still gonna pull out that gate diode and see what happens. Anyone want to make a prediction?

 

[gbritnell]

Gentlemen, as I had mentioned in my previous posting, some of you are more skilled at electronics than the rest of us and my question at that time was, could you please compile a parts list with the part numbers for us so we could also try your system. I would be thrilled to have a $5.00 ignition for my engines as I'm now up to about $275.00 for the 4 that I have.
Thanks,
gbritnell

 

[jpeter]

Which ever circuit you use it's going to cost you around $5.00. Bipolar transistors are less than a quarter a piece so Admiral's circuit is cheap to build too. Admiral, I've been having a lot of fun bantering with you but I'm convenced you've got a depth of knowledge about the subject; you know what you're talking about. Be reminded guys you're going to have to buy a resistor type ignition coil from the autoparts store for about $20 and spend a couple of bucks on incidentals. With shipping and running around and after a certain amount of frustration you'll have a working TIM for $35 or so. If you can buy one all ready to go for $50 than going that way might be a good choice too. I like to make my own stuff. I don't want to put words in Admirals mouth but I think he does too, like to make his own stuff.

Regarding the question about the rise time: I didn't worry about it cuz I figured the time to charge the coil wasn't too important to the function so long as it got charged in time. Right? Also, I was concerned about the health of my o-scope getting it around that much voltage. Rise time would be interesting though to know. I outta check that out.

 

[Lakc]

Regarding the question about the rise time: I didn't worry about it cuz I figured the time to charge the coil wasn't too important to the function so long as it got charged in time. Right? Also, I was concerned about the health of my o-scope getting it around that much voltage. Rise time would be interesting though to know. I outta check that out.

The magnetic flux built up in the coil primary is dependant on the rate of charge, so its rather important.

Yes, its not a good thing to get your scope too close to the hv side, use an inductive pickup or make friends with a mechanic and use an automotive scope (with an inductive pickup).

George, dont worry, your on my short list of beta testers

 

[jpeter]

Hey, I thought the strength of the flux field was determined by the strength of current flow in the inductor. I didn't think time played a role. I agree if the igbt doesn't clamp up in time then full primary current won't be reached. As long as full current flow was reached before the coil fires I thought that would be fast enough. Am I wrong?

 

[Lakc]

Hey, I thought the strength of the flux field was determined by the strength of current flow in the inductor.

I think that may only be partially correct for our application. If you check out here http://en.wikipedia.org/wiki/Magnetic_field

Under the Faraday's law section, it mentions

In terms of magnetic field lines, the electrical voltage that is generated in an electrically conducting coil is proportional to the time rate of change of the number of magnetic field lines that pass completely through the loop. I didn't think time played a role.

And since its voltage that were looking for here... But I wholly admit there is a bunch of stuff on those pages I do not understand, and I wouldnt touch most of the math without Marv K at my shoulder.

I agree if the igbt doesn't clamp up in time then full primary current won't be reached. As long as full current flow was reached before the coil fires I thought that would be fast enough. Am I wrong?

I am not sure, but as velocity is usually the squared component in many energy type reactions, my gut feeling says speed counts. That could well be completely irrelevent too.

We need another button aside from the "report to moderator" one, one which says "paging Marv we need math help in this thread" Rof}

 

[Kermit]

The number of 'turns' of the coil is a primary factor in the inductance of the coil.

Increased inductance will reduce the 'speed' at which maximum flux density can be reached.(slows down response of the circuit)
Reducing the inductance will reduce the power of the spark, but increases the speed at which one can get to the next spark in the cycle

increasing the drive voltage will help, but then, you have the problem of more easily damaging other circuit components(transistor devices)

It is a balancing act, with higher speed motors needing, understandably, more expensive and intricate circuit designs.



Keeping it simple