Need to dissolve epoxy potting in ignition coil.

[Circlotron]

I have a General Motors LSx ignition coil (has a potted coil driver cct inside) and I need to find out the just exactly what this circuit inside is. I have looked into what is involved in dissolving the epoxy but the chemicals needed  sound hazardous. Can anyone tell me of a company that could do this for me? Preferably Australia but United States is also okay; have a contact over there.

P.S. Don't anybody say just use a MC3334. They are not suitable.
I have been building ignitions for 37 years.

[calzap]

If you search on youtube for "Decapping ICs", you will find a post by Applied Science on how to use hot fuming nitric acid to dissolve (destroy really) epoxy.  It does involve a dangerous chemical, and he does not take adequate precautions.  The worker should wear closed shoes, long pants, gloves, goggles, and a lab coat with long sleeves, and there should be a fume exhaust system rated for corrosive gases and a clear, shatter-proof shield between the worker and the work.   Ideally a lab-grade fume hood should be used. This method will do away with most metals and plastics.  Semiconductors, gold, glass, some plastics (e.g. teflon) and ceramics should be relatively untouched.   The other relatively inexpensive method is a slow, careful grind with a small milling bit in a high-speed drill, like a Dremel or its air-driven counterpart.

Fast methods involve the use of high-power water-jet or laser cutters, both of which are very expensive.  Both can take off a finger or put out an eye faster than human reflexes can react.  Depending on what you need to know, an x-ray might do the trick and would be quick.  Best bet would be to find a x-ray tech, hospital, dentist or vet to do it.  Dental and some industrial x-ray machines have very fine resolution.

Mike in California

[eurofox]

Maybe this will help ....

[MK]

Warm Chloroform in an ultrasonic tank may help too.

[eurofox]

Warm Chloroform in an ultrasonic tank may help too.

Good night     

[Alex Eisenhut]

Maybe try the stuff they sell in hardware stores. It's a gel, it doesn't smell that strong. You just scoop it on and wait.

[SeanB]

Go to your local large fibreglass wholesaler and ask them for a gallon of epoxy remover, commonly used to remove fibreglass resin from parts. Note it is hazardous, noxious, you need 3 pairs of nitrile gloves to handle it , it dissolves paint brushes ( buy a good few Blondie brushes there, they last longer and do not drip paint all over from the fumes), and use a metal tank to hold the coil, and use outdoors in a well ventilated area with somebody to watch you in case you fall over. Drop coil in, leave overnight and in the morning you will be able to pick off a layer of soft epoxy. Soak again until done.

BTW it will NOT work after doing this, but you will have a big kit of parts. then you can identify what they were.

[katzohki]

I have experience doing this with various materials. Is the epoxy hard or soft?

If it is soft you might get away with heating it up and picking it off. The hardest of hard will require highly caustic chemicals or physical abrasion (sanding) to remove.

[Richard Crowley]

If the potting is "soft" (which seems unlikely), then is is NOT "epoxy".  It is probably hard tar or something they might  be "melted" out under high heat.
But if it is truly epoxy (or any common thermosetting polymer, you will need very harsh and extreme (and dangerous) chemicals to truly de-pot it, and it will almost certainly be desctuctive.

http://en.wikipedia.org/wiki/Thermosetting_polymer:   Polyester, Polyurethane, Epoxy, Bakelite, Melamine, rubber, etc.

http://en.wikipedia.org/wiki/Thermoplastic: Acrylic, Nylon, Polyethylene, Polypropylene, Polystyrene, PVC, Teflon, etc.

[Corporate666]

I've had some luck removing hard epoxy using MEK (Methyl Ethyl Ketone).  Soak it in the stuff for a while (days) and it will absorb the MEK and become jelly-ish. Then you can use something like a dental pick to remove the epoxy.

If the epoxy is thick, you won't be able to get it off in one shot, you'll need to soak multiple times.

There may be harsher stuff that works faster than MEK, not sure.  I've also used acetone for softer stuff.  It generally makes it crumbly and easy to remove.

You can also bake it.  Components should be fine at 400F, but very few epoxies last much above 250F.  So you could bake the coil in a toaster oven (not one you will be using for food) for a while - like a few hours - and it will allow the epoxy to be broken off carefully with a pick.

[katzohki]

Have you considered contacting the manufacturer and see if they are willing to help you out?

[atferrari]

IIRC, in his "Hardware hacker" colum, Don Lancaster mentioned potting compound and IIRC how he removed some.

All his  columns are collected in his site. Be patient; sooner or later you could find what you are looking for.

[max_torque]

Ok, potentially stupid question of the week:  why?

It's and ignition coil, it's going to be pretty obvious what is "inside" of it, and because it's a mass produced mainstream automotive part (take a guess how many are made each year!!) it'll almost certainly be all custom packages and boarding etc, so even if you get the epoxy off, you won't be able to find the "datasheet" for the devices anyway!

There are only certain ways of driving an automotive ignition coil, as i'm sure you are aware, so i can't exactly see any "magic" being done in there, but i can see every attempt at component / system "cost down" possible...........

[Andreas]

Hello,

usually something like a BOSCH BIP373 powerstage (or similar) is used within the coils.
Mainly consisting of a IGBT and some diagnostic circuits.

With best regards

Andreas

[NiHaoMike]

There's a trend towards using a two stage topology where the 12V is first boosted to as much as 400V or so, then used to operate the ignition transformer as a forward converter. That apparently lets them get higher peak power and allows them to use leaner mixtures for better fuel economy. Hyundai was the first one to use such a design in a mainstream car. I'd imagine it's much more common now that automakers are being pressured to improve the MPG on their conventional cars in order to make them more competitive against the hybrids and EVs.

[Circlotron]

What I want to find out by opening the thing is NOT how to drive the coil; that's easy. The electronics has two earth, a power earth and a trigger signal earth. The power earth goes to engine metal and the signal earth goes to the computer along with the usual trigger wire. This allows the two earths to be at slightly different voltages and yet not cause triggering problems because of noise because the trigger is referenced to the signal ground, not the power ground. Obviously they have done it in a very simple manner and I want to use their method as a starting point.

We are actually going to market a high performance replacement version of the OEM coil. The company I am doing it for has already had prototypes of the coil section made with a core double the normal cross section and had the plastic case mouldings done too, so it is well along. Energy output is up to 190mJ if you shove it hard enough. That's more than double the OEM version.

[Richard Crowley]

So what is the DC resistance you measured between the two grounds?
Unless it uses an opto-isolator for the trigger input (which seems unlikely) chances are the two grounds simply connect together at some appropriate point in the circuit.

[Circlotron]

So what is the DC resistance you measured between the two grounds?

For the LS1 type coil there was 36.9K both meter lead polarities.
LS2 coil = infinite ohms and < 100pF capacitance.
Yukon truck coil (exposed heatsink type) = infinite ohms both polarities and 100nF in one meter lead polarity only. Reverse polarity < 100pF capacitance.

See pics for coil types and connections.

Edit -> Yukon coil 100nF in one direction, not 100pF in both directions.

[Circlotron]

Unless it uses an opto-isolator for the trigger input (which seems unlikely)

I also think unlikely. From external measurements the engine computer signal output appears to have a 2k7 pullup resistor to +5v. Not really enough to drive an opto properly, particularly a schmitt triger type.

[max_torque]

Does one actually need more spark energy that the std coil can supply?

[Circlotron]

^^ Under ideal conditions probably not. With a race car engine though you may have poor fuel atomisation because of low inlet temperatures caused by inter coolers, nitrous oxide injection, and evaporating alcohol. In particular, blown alcohol engines can be difficult to start when cold. You need all the spark you can get. Even if your 1000hp engine only picks up 5hp, if you are one foot ahead of the other guy at the far end of the quarter mile then that's all that matters.

[mzzj]

There's a trend towards using a two stage topology where the 12V is first boosted to as much as 400V or so, then used to operate the ignition transformer as a forward converter. That apparently lets them get higher peak power and allows them to use leaner mixtures for better fuel economy. Hyundai was the first one to use such a design in a mainstream car. I'd imagine it's much more common now that automakers are being pressured to improve the MPG on their conventional cars in order to make them more competitive against the hybrids and EVs.

Trend? I think capacitor discharge ignition was more popular on european cars in 80's and 90's

[NiHaoMike]

It's not CDI, as the intermediate rail does not discharge. More like a high voltage switching power supply that is powered by another high voltage switching power supply.

[mzzj]

It's not CDI, as the intermediate rail does not discharge. More like a high voltage switching power supply that is powered by another high voltage switching power supply.

ah, makes sense.

[Circlotron]

What model Hyundai did that system appear on? I want to look it up.