isolating screws from taped heatsink for mounting TO-3 package

[Zak28]

It would be far quicker to tap holes and screw the TO-3 package but can screws be insulated from TO-3 case?

The TO-3 case will go to ~800v - the screws must make contact with heatsink threads which require screws to be isolated from case.

 

[wwfeldman]

use a plastic (nylon?) shoulder washer

examples:
https://www.fastenercomponents.com/nylon-standard-shoulder-washers-p-27.html

put shoulder washer into mounting hole, put screw into shoulder washer

 

[Zak28]

Nylon parts can be torqued to TO-3 torque specs? https://www.digikey.com/products/en...1d4&quantity=&ColumnSort=0&page=1&pageSize=25

 

[wwfeldman]

i remember watching them being used in production

the wiki article

https://en.wikipedia.org/wiki/TO-3

shows lots of plastic and fiber shoulder washers

and references this article on mounting

https://www.ti.com/lit/an/sboa020/sboa020.pdf

which has this bolded
The recommended tightening torque is 6 inch-lbs., with anacceptable range of 4 - 7 inch-lbs. ( 0.45 - 0.79 N-m ).

and this:
**broken link removed**

and this
https://www.shopaavid.com/Portals/0.../Aavid-Board-Level-Heatsinks-Catalog-2018.pdf

to answer your question:
without specific torque specification:
i expect the nylon shoulder washer to NOT handle the same torque as the metal TO3 case
i also expect that they are acceptable to do the job

 

[Zak28]

Whats the approximate thermal resistance from TO-3 case into heatsink using slightly greased mica? https://www.digikey.com/product-detail/en/56-03-8G/56-03-8G-ND/1625398/?itemSeq=293777268

It would probably be better with more torque but nylon doesn't permit it.

 

[wwfeldman]

the issue usually isn't the torque
its the contact between the heat sink versus the TO3 case

one of the references in post 4 talked about:
1) drill, don't punch the mounting holes and C E holes - punching leaves craters which reduce the contact area and the quality of the contact
2) machine the area of the heatsink to a flatness of 0.001 inch per inch - the TO3 package is o.004 inch per inch

 

[hoyt]

Why won't the metal hole in the TO-3 arc over to the hole in the heatsink (via the path through the hole in the sil-pad or mica insulator hole)? That same arc path is there even wehn the shoulder washer & screw is completely removed or inserted

 

[wwfeldman]

the dielectric strength of air is about 3 x 10^6 V/m

https://hypertextbook.com/facts/2000/AliceHong.shtml

that's 3000 V/mm
or 300 V/ 0.1 mm

the space between the screw and the TO3 is on the order of 1 mm (0.03937 inch)
at 1/64 th of an inch (0.015625 inch or 0.3969 mm) its about 1000 V
much higher than the collector voltage of most TO3 components

see page 100
https://www.shopaavid.com/Portals/0.../Aavid-Board-Level-Heatsinks-Catalog-2018.pdf
the pertinent part of shoulder wacher is about 1 mm
so there is about 1 mm between screw and metal.

 

[hoyt]

If the case is at 800V, what keeps it from arcing from there to the heatsink metal which is extremely close by?
You have a hole sandwich:
A) TO3 --with metal hole at 800V
B) Insulating pad...Very thin insulator with a hole
C) Heatsink---metal with tapped hole

There are 3 stacked holes.. The only thing separating metals A & C, is the thickness of hole B, regardless of whether a shoulder washer & screw is inserted into these stacked holes.

Is this discussed some where??

Certainly the washer keeps the screw isolated & provide "lots" of creepage. But what prevents the TO-3 from "arcing" to the heatsink? UL requires a much larger creepage distance at 800V.

 

[FvM]

All too vague. The problem can be clearly discussed in terms of clearance, creepage and dielectric breakdown voltages. Would you mind to show a drawing or photo of the respective mounting? We'll try to interpret it in IEC1010 terms.

 

[Zak28]

All too vague. The problem can be clearly discussed in terms of clearance, creepage and dielectric breakdown voltages. Would you mind to show a drawing or photo of the respective mounting? We'll try to interpret it in IEC1010 terms.

The voltage devices which heatsink will facilitate will not have sufficient voltage to break over and arc.

Is there a small inexpensive tool or method which can determine whether surface is adequately flat?

 

[wwfeldman]

@hoyt
is there an 800V rated TO3?

see this thread:
Transistor isolation question

@FvM
i have not found the information you want - at least not all in one place
every article, ap note etc on mounting transistors, not just metal TO3, but plastic -220s -243s etc
do essentially the same thing - screws, nuts, washers and shoulder washers.
they do not talk about creepage distances or arc over.
neither does Horowitz and Hill (at least, i didn't find it)

@Zak
I doubt it is small,
and i doubt it is inexpensive
especially since very few things seem to be in TO3 cases anymore