Denso EFI potting=ack!

[Kalle Pihlajasaari]

Hi Chris,

Long booring post that is off topic (almost) about getting to look inside
potted ECUs.  I once was asked to 'expose' the insides of a Outboard
motor ECU and they had done a good job of making them irrepairable.

>   The Suzuki TL1000S arrived on Friday.  It's fitted with a Denso ECU.
> There's no EPROM slot, and the production version of the box is 
> completely potted with the most satanic stuff I've ever seen.  It's 
> like RTV with lots of ground glass mixed in... Is this an industry 
> standard potting material?
> 
> It's hell to chip off and the crunching noises sound exactly like the
> cruching a DIP would give off while being destroyed. (!)
> 
> The catch is, we have to be very careful, because the bike is a brand new 
> model and there are no parts in the pipeline yet.  (and it has it's first
> race in 4 weeks.) 
> 
> Has anyone encountered this horrible stuff and if so, is there any safe 
> solvent for it that won't harm the board?

Solvent, yeah.  Safe, well now theres the rub :-)

The fillers are put in for a few reasons, sometimes overlapping and
there are a few fillers that will perform many/most of the desired
functions.  As mentioned so far Glass for mechanical property matching
with ICs and glass diodes.  Whatever for reducing cost of potting compound.

Then there are a few more reasons.  Composite mechanical strength, one
can use anything that is larger than a powder.  Optical properties,
almost anything will obscure stuff even if you started with a clear
epoxy.  Thermal properties, many materials will improve on the potting
compounds heat transfer capacity.  Tamper resistance, this is something
I have come across (and once made use of) and here you just want to add
something pretty hard, the stuff I saw made use of carborundum and when
diging into the mix your knife blade would become an ex-blade very quick.

Now a cheap material to add is washed river sand available from petshops
and it will work pretty well with all of the above needs.

RTV Rubber type compounds are pretty inert but are softer so the knife 
and as sugested Hot air attack are perhaps a better idea, I have 
not seen any RTVs mixed with fillers to date.

Now for getting the stuff out, I make use of this real toxic stuff that
is also a 'suspect carcinogen' called methelene chloride or di-methyl
chloride depending on where one gets it.  I have heard that it is the
active ingridient in a commercial unpotting compound so it might be what
the 'Uresolve' as mentioned contains.  They might add some other modifiers 
to attack some plastics faster and reduce the evaporation but that is my 
own speculation.  

Use this stuff outdoors and avoid breathing the vapours, gloves are
a plus but not sure what type to use (your hands get real funny if
you get some of the dirty stuff on as the solvent evaporates in a moment
and you are left with epoxy stains on your fingers that resist
all friendly solvents).  The action is quite slow and you leave the
whole thing immersed.  Expect to work through about 1 mm after 20-30
minutes of soaking, heating helps but the vapour pressure goes up
fast and the stuff boils below 37degC so one just has to keep it at about
22degC as it likes to get cold from the evaporative cooling.  I have not
done this in a while but on the last thing I unpotted I came to the
conclusion that one needs to agitate the container and continuous brushing
to remove the surface as it softens.   It does not disolve, just 80% of
the plastic bonds become weak and then it you take it out for more than
a minute the stuff evaporates ant you are back to almost what you had
before.  If you let it soak it swells a bit as it softens and so helps
with getting around the filler material.

Be warned that it will also remove the colors from resistors and numbers
from components as well as possibly destroying small coil formers and
similar thermo plastic materials.  Usefull for opening something
to fix or investigate, risky if you hope to keep it working afterwards.

> I used to work for a defense contractor.  Everything was conformal coated
> or potted.  We had a refrigerated freon tank to dissolve the coatings.
> Hang it in the vapor for a while, and it would drop off.  Drop it in the
> liquid for a few seconds, and it's gone.  I don't recall which refrigerant
> was used, but if you can get your hands on some, try spraying the board.
> I think some "cool-it" type sprays are still available with freon.
> If it's an epoxy type material, all you can do is chip away at it.

Epoxies are much harder than solvent coatings but possible with
the required work.

> There is no "solvent" for disolving the potting however there is a chemical
> that will remove it. It is called Uresolve. I am at home right so can give
> no other data.  I will look at the stuff tomorrow for the mfg address.  If
> you send me an Email I will reply from the computer I will be at tomorrow
> night.  The glass is probally microspheres.  They are added to lower the
> density of the epoxy and to make it more gentle on the components.  Contrary
> to many peoples expectations potting in epoxy can be extremely stressful to
> electronic components.  When an assembly is depotted with Uresolve it is
> hard to tell that it was ever potted.  As a side note there are vendors
> around that will do the job for you.

Interesting to find out if the active ingredient is listed.
 
> The same stuff was used in some Ford ignition modules. I think the reason
> they add the silicone to the potting is that sand is cheaper then RTV,
> 
> I found that a hot air pencil worked to soften the RTV and then it just 
> crumbled away. It was still a lot of work.

RTV can often be worked off with blades and it does not stick as well
if it is of the silicon variety, probably more resistant to 'solvent'
type of attack though.

> Most pottings are epoxies, I imagine this is the same.  Although it is a 
> thermoset polymer, enough solvent can do the trick.  Acetone, or even 
> worse, Methyl Ethyl Ketone (MEK) will soften  the epoxy, if as rag wetted 
> with it is placed on top. MEK is particularly nasty, I would wear gloves 
> and an organic vapor mask if I was using the stuff for a long time.  
> Boiling nitric acid will digest epoxy, but that's a bit extreme.  The MEK 
> or acttone will help weaken the potting, but i wouldn't dip it, the 
> actone and MEK dissolve many thermoplastics like ABS.

MEK Sounds like something as agressive as the methelene chloride I mentioned.
Acetone is almost uselessly slow unless the epoxy has not quiet set, then it
works a treat.

> Good luck, and wear gloves!

Yeah.  What type of glove is good for MEK ?

Cheers
-- 
Kalle Pihlajasaari   kalle at ip.co.za   http://www.ip.co.za/ip
Interface Products   P O Box 15775, DOORNFONTEIN, 2028, South Africa
+ 27 (11) 402-7750   Fax: 402-7751    http://www.ip.co.za/people/kalle

DonTronics, Silicon Studio and Wirz Electronics uP Product Dealer