Thermal transfer and shimming adhesive

For the three and more legged things
Post Reply
User avatar
Dr Bunsen Honeydew
Old Hand
Posts: 1358
Joined: Mon Sep 21, 2009 10:50 am
Location: Muppet Labs

Thermal transfer and shimming adhesive

Post by Dr Bunsen Honeydew » Fri Jan 12, 2018 11:13 am

Posted at HFS but inf might help DIY solid state builders here.

Forum lie - I use ordinary glue to fix output transistors and they fail because of it. In the past seen at both PFM and at AoS and used to try and show how badly NVA amps are made. Latest put on a 2nd hand A40 listing by Emporium Hi-fi at ebay which was libel that I have got him to remove.

I have avoided publishing this before as why should I help people design and build amps, it is far easier to use than the tradition method if more expensive. For me the primary motive is I need to use it to maintain ClassII double insulated appliance regs.

http://www.seifert-gmbh.de/Loctite_315.pdf

User avatar
ed
retired
Posts: 3656
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

Re: Thermal transfer and shimming adhesive

Post by ed » Fri Jan 12, 2018 12:44 pm

Interesting angle Richard. I’ve used the loctite stuff on TI class D chips but never dared to use it on very hot fets. I sleep easier with nuts and bolts and Sil-pad 400 or mica and goop, although I accept that the loctite is a proven solution.
I have avoided publishing this before as why should I help people design and build amps,
Whats interesting is the perceived need to share this with the diy community. Surely all you need to do to combat the dissenters is deny that you use normal glue and divulge the name of the product you do use. You don’t have to donate to the diy community at all in this instance.


In what way would the use of nuts and bolts defeat your class ii builds?
There's nowhere you can be that isn't where you're meant to be

User avatar
Dr Bunsen Honeydew
Old Hand
Posts: 1358
Joined: Mon Sep 21, 2009 10:50 am
Location: Muppet Labs

Re: Thermal transfer and shimming adhesive

Post by Dr Bunsen Honeydew » Fri Jan 12, 2018 6:24 pm

It takes away one layer of insulation. The TO-247 cased output transistors are glued to black anodised aluminium heatsinks. The anodising is one layer of electrical insulation, by tapping and screwing into it you lose this as the screw holding the transistors is now part of it and not separately insulated from it. So the rail voltage to the transistor which goes onto the case of it only has one layer of insulation, so defeats the double insulated regs.

Regarding giving the inf, I am just trying to be helpful member of the forum believe or not.

Once again if you wish to understand then pay for and download the regulatory documents. Google is your friend.

User avatar
Nick
Site Admin
Posts: 11409
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Thermal transfer and shimming adhesive

Post by Nick » Fri Jan 12, 2018 6:43 pm

But do you not get the extra insulation back by isolating the screw from the transistor with a plastic spacer?

Or is it because the anodizing and adhesive are two layers of insulation, the screw and washer mean its just the one?

Would the same not be achieved by using a nylon screw to hold the transistor down on the insulating pad?

Or would a solid insulating pad and holding the transistor in place with a clip give the same two layers of insulation?
Resistance isn't futile it's V / I.

User avatar
Dr Bunsen Honeydew
Old Hand
Posts: 1358
Joined: Mon Sep 21, 2009 10:50 am
Location: Muppet Labs

Re: Thermal transfer and shimming adhesive

Post by Dr Bunsen Honeydew » Fri Jan 12, 2018 6:52 pm

Nick I am certain there are other ways to do it to solve the problem. It is even possible to produce Class II product with a metal case but just more faff. I love simple solutions and this is a simple solution.

Same as with Class I the main thing is coping with a single failure, mainly a wire with potentially dangerous voltage breaking off and touching some thing. The traditional output stage fixing method would mean the head of the screw would only have one layer of insulation to the outside, so a failure of the anodising such as chip or deep scratch would allow a shock.

User avatar
ed
retired
Posts: 3656
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

Re: Thermal transfer and shimming adhesive

Post by ed » Fri Jan 12, 2018 7:37 pm

I think Nick has illustrated exactly what I was thinking when I asked the question.
Re: the attached photo I thought the plastic shield back and front would prevent the screw ever touching any part of the transistor. I may be missing the point entirely but I see your benefit as not having to tap the sink.

as you say, it is possible.
mp1620 mn2488.jpg
mp1620 mn2488.jpg (4.77 KiB) Viewed 1233 times
There's nowhere you can be that isn't where you're meant to be

User avatar
Dr Bunsen Honeydew
Old Hand
Posts: 1358
Joined: Mon Sep 21, 2009 10:50 am
Location: Muppet Labs

Re: Thermal transfer and shimming adhesive

Post by Dr Bunsen Honeydew » Fri Jan 12, 2018 8:01 pm

Sorry but you didn't understand, you need to read my last post again. It nothing to do with touching the transistor it is the exposed head of the screw and losing internal insulation from a short circuit fault due to loose cable, which is what the regs are about. They presume there can be a single fault so there should be double protection as the chance of two associated faults at once is next to impossible, OR as in Class I the fault condition is taken to earth.

User avatar
ed
retired
Posts: 3656
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

Re: Thermal transfer and shimming adhesive

Post by ed » Fri Jan 12, 2018 8:19 pm

got it, thanks.....bluddy finniky these regs innit.
There's nowhere you can be that isn't where you're meant to be

Post Reply