# Heat

#### Jaguarjoe

Joined Apr 7, 2010
767
The attached pdf is the circuit for the ignition system in millions of GM cars and also my Jaguar. It commonly known as a 4 pin HEI module and looks like this:

http://tinyurl.com/2bjxqwx

It limits the charging current to the ignition coil to about 5 amps. The driver darlington MJ10012 has a Vce(sat) of 2 volts. When the darlington conducts, it has 5 amps flowing through it and a 2 volt drop across it. That's 10 watts of heat to dissipate. I realize that is not an insignificant amount of power, but why does this thing run so hot you can't put your fingers on it?

#### Attachments

• 110.8 KB Views: 31

#### mik3

Joined Feb 4, 2008
4,846
It should be mounted on a heatsink. If it is designed to work without a heatsink then something is wrong.

#### hgmjr

Joined Jan 28, 2005
9,029
I agree with mik3. A heatsink is need to increase the heat dissipation out of the device.

hgmjr

#### Jaguarjoe

Joined Apr 7, 2010
767
On GM products it is mounted to a substantial aluminum piece. On the Jag its mounted in a small box attached to the inlet manifold through two standoffs.

#### DigitalReaper

Joined Aug 7, 2010
70
You can actually see a tube labeled silicone in the picture which is probably thermal paste (used for mounting it on a heatsink). Do you have the module securely bolted on to where it's supposed to be mounted and did you use the paste? It uses whatever it's mounted on to dissapate heat.

#### mik3

Joined Feb 4, 2008
4,846
It seems that GM has done a better job than Jag!

#### Jaguarjoe

Joined Apr 7, 2010
767
Silicone isn't a good thermal conductor. I use berylium oxide heatsink paste.
I remounted the module off to the side of the car away from the engine heat and covered the top of it with a heatsink. It still gets hot as a pistol.
How much heat does 10 watts generate?

No doubt GM did a better job with this than Jaguar did. This car was built before Ford took them over and re-invented them. In the pre-Ford years anyone did a better job than Jaguar and Joe Lucas.

#### marshallf3

Joined Jul 26, 2010
2,358
10W is more than you think but I don't recall those old modules ever getting that hot. You might have an ignition coil drawing too much power due to a shorted turn.

Chrysler used a similar device, they actually mounted it in the air cleaner housing so some air would blow over it.

#### retched

Joined Dec 5, 2009
5,197
I would have to agree with the shorted winding..

You should really probe this and get the actual amp draw.

#### Jaguarjoe

Joined Apr 7, 2010
767
Over the years, I've had three different ignition coils in this car and there was no noticeable difference in temperature. Being current limited to 5 amps, what difference would a bad coil make anyway?
A few years ago I used a 0.1 ohm shunt resistor to see the module work on my scope. You could see the coil charge up and then sit at 5 amps just like it should.

#### marshallf3

Joined Jul 26, 2010
2,358
The driver darlington MJ10012 has a Vce(sat) of 2 volts. When the darlington conducts, it has 5 amps flowing through it and a 2 volt drop across it. That's 10 watts of heat to dissipate.
Old technology. With only four pins it couldn't be that complex of a circuit, why not make one up using a modern MOSFET?

#### retched

Joined Dec 5, 2009
5,197
Im thinking that the current limiting process or circuit is, or may be dead/dying.

I would like to see numbers on a ammeter.

Unless you have 'super sensitive' fingers, this device should not be "HOT" hot to the touch.

marshallf3 said:
Old technology. With only four pins it couldn't be that complex of a circuit, why not make one up using a modern MOSFET?
I dunno... have you had a look at the motorcycle AC regulator thread? It has 1 wire and it contains QUITE a smathering of components that has taken weeks to put together.

Is "smathering" a word?

#### Jaguarjoe

Joined Apr 7, 2010
767
I found this mosfet in about 5 minutes which would almost work:

http://www.fairchildsemi.com/ds/FC/FCPF13N60NT.pdf

I'd need a good gate driver but worst of all, the MC3334 IC which is the brains of this whole thing is no longer available

Years ago Silicon Chip magazine had an article on building one of these. I don't recall if it was BJT or FET equipped. I think it used the same IC.

#### marshallf3

Joined Jul 26, 2010
2,358
You can still buy things that will subsistute but I'll almost bet you the circuit's out there somewhere. If I recall there's just a hall efect or simple electromagnetic sensor in the distributor that operates some simple switchable biasing on the coil driver transistor.

So you haven't even tested or replaced the thing? Most any auto store will test those for free, I even have a tester for the common ones here at home.

---

You can do a lot better than that MOSFET both in amperage capability and Rdss_on for about the same price.

Try something like this:
http://www.mouser.com/ProductDetail/Fairchild-Semiconductor/FDP8860/?qs=sGAEpiMZZMtCrm2fS1SYQuPoGNV9k5Et%2b3UvskdTnnw=

#### Jaguarjoe

Joined Apr 7, 2010
767
I've been through numerous modules. Many run for 20 minutes and die. Others do better. At only $15 each, I kept looking for a cool one, but they're pretty much all alike. I looked at that mosfet you found. Its only good for 30 volts. I need 375- 600. As voltage goes up, so does Rds on. I end up with just about the same drop across it as the darlington has. #### marshallf3 Joined Jul 26, 2010 2,358 I've been through numerous modules. Many run for 20 minutes and die. Others do better. At only$15 each, I kept looking for a cool one, but they're pretty much all alike.

I looked at that mosfet you found. Its only good for 30 volts. I need 375- 600. As voltage goes up, so does Rds on. I end up with just about the same drop across it as the darlington has.
Unless there's an integral CDI hidden in the box or elsewhere you're just seeing the back EMF from the coil when the current source to it is broken. That's what a good rectifier reverse connected across the coil primary is for.