Insulating hardware, esp.washers

I've got an electrically and thermally hot TO-220, that needs mechanically securing to a grounded heatsink.

Anodizing isolates the TO-220 from the heatsink, but what do you guys use to isolate the mounting hardware to prevent it from shorting the device to the sink? Something that'll maintain the mechanical mating force, despite long-term operation at >80c? .----. |////| |////| ~ ~ TO-220 ~ ~ split | |////| washer V |////| | .-. |////| nut V | | |////| / .-. | | |////| .---./ _ | | | | |////| |___| bolt / || | | | / _ '-^^^^^^^^^^^^^^^^^^^^^^^^^./ _| | .-vvvvvvvvvvvvvvvvvvvvvvvvv' \_|| | |___| | | | | |////| | | '-' / | |////| '---' / | |////| / | |////|

Reply to
dagmargoodboat
Loading thread data ...

McMaster Carr sells ceramic shoulder washers for a nickel.

formatting link

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Of course none of them is small enough to fit inside a TO220 mounting hole, but maybe a nylon sleeve and either one of those ones or an insulating washer to take the compression:

formatting link

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Nylon and FR4 will both cold flow and loosen up the hardware.

A bigger package is better, TO-247 or some such. TO247s have the screw insulator built-in.

formatting link

All that needs is a screw and a lockwasher or Belleville to clamp it to an anodized sink. The bigger footprint is a huge thermal advantage.

For un-repairable stuff, skip the hardware and use epoxy or krazy glue.

There are isolated-silicon TO247 types, which would save anodizing the heat sink.

Or clamp them down:

formatting link

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators
Reply to
John Larkin

I don't trust the anodizing over time and thermal cycling. I make servo amps with either 4 or 6 transistors on a heat sink. I use the best Bergquist elastomer insulating pads and use a nylon step washer to insulate the screw. If the transistors are running hot enough to melt a nylon washer, they are running too hot for longevity. But, my rule of thumb is that the transistors should never exceed 50 C for best longevity.

If you put the transistor directly on the anodized heat sink, the thermal conductivity is awful without a heat transfer material such as the zinc oxide loaded silicone grease, which is kind of messy to deal with.

Jon

Reply to
Jon Elson

On Sat, 21 Dec 2013 10:34:12 -0800 (PST), snipped-for-privacy@yahoo.com Gave us:

No... HARD Anodizing creates an electrically insulative layer. Used to be three levels in the mil spec days.

Ordinary chassis type anodizing, I would not trust as an electrical insulator when fitting coined edged heat sink tabs onto it.

Also, if you DO 'hard anodize', and do so after the holes are punched, they are also insulated.

I would use a compression cup type fastener retention method, and a thread locking compound. You could sleeve the screw in the hole region if you are worried about poor assembler aptitude for correctly attaching a device with the fastener floating in the hole, instead of abrading against its walls. A larger, flat washer to spread force on the bottom, and *maybe* a thin teflon strip on that side of the sink in the hole locations. NO lock washer on a surface you do NOT want to make electrical contact with.

So, the underside flat washer sits on top of the teflon film to keep a badly chosen flipped washer placement to not abrade the surface. No need to worry about a properly flipped flat washer (reference to 'coined' edges). The cup type compression washer on the device side does not particularly need to be insulated from the tab at this point as you have insulated the bolt with a sleeve or surface treatment already.

While only sacrificing as little as possible of the thermal efficiency of the sinking assembly.

Yes... you definitely want hard anodizing.

Bad device choice with all the modern fasteners in the channel.

IF you were actually WANTING electrical contact, you would choose an internal or external star type washer. They also retain fixturing with less set torque applied a bit better than a single split notch would.

There are a lot of better polymers in use these days than the old, crusty stuff, your old, crusty ass remembers. :-)

This old, crusty ass just happens to still be working with discreet elements like this.

Even do a lot of conduction cooled stuff on sealed, fully canned 'card assemblies'

I have gotten some of the media after it expires, and it makes nice layers of "thermal putty" to pass heat from a hard drive case to an enclosure, for instance.

Reply to
DecadentLinuxUserNumeroUno

Nylon 6/6 GF30 (30% glass-fiber reinforced) is really durable.

formatting link

PPS is a good/better alternative, this is more available:

formatting link

Reply to
bloggs.fredbloggs.fred

I like Mica with a full body clamp or the package sandwiched between heat sink. Jamie

Reply to
Maynard A. Philbrook Jr.

Use mica insulators and a bar clamp. You can also get spring clamps from mouser.

CHeers

Reply to
Martin Riddle

What we used in automotive electrical connection applications (such as mounting alternator regulators) were Belleville (aka Belville) washers. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Yep, constant spring pressure is the best. As long as the spring can keep it's K.

Cheers

Reply to
Martin Riddle

As long as they don't get too hot... or they get over-torqued... as I did to them on several occasions :-( (But the boss has to do something to keep the technicians amused ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

We do that, 1 mil hard anodize with grease. Theta is a fraction of what you can get with Bergquist pads.

Don't believe the Bergquist specs, unless you have a full-time hydraulic ram to keep the contact pressure up. Derate 2:1 for real-life use.

The great thing about silicone grease is that it compresses down below 100 microinches with moderate pressure. The pads will always be ballpark 10 mils thick, and that's terrible for heat transfer.

Unfilled grease is less messy than the white stuff and almost as good.

The TO220 package is terrible thermally and mechanically. Given that transistors are cheap and heatsinks are expensive, it usually pays to use a bigger package. Or several transistors instead of one.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators
Reply to
John Larkin

Which they usually don't!

Jamie

Reply to
Maynard A. Philbrook Jr.

Beryllium Copper contacts have good spring tension. Run high current thru them and they loose their K. In my Tahoe the dash lights used to blink when it was hot, it turned out to be a bulkhead connector contact, the main B+ for the instrument cluster. Re bent the one terminal an all was good. There is some thing about those auto designers that they never seem to get electrical stuff right.

Cheers

Reply to
Martin Riddle

Thanks Phil. Those are nice to know about. Here I'm chicken ceramic will crack.

The link redirects to a listing of parts. I didn't see any for a nickel--if this link works, 5 pcs are ~$13.

formatting link

Cheers,

James Arthur

Reply to
dagmargoodboat

This unit was designed to run cool, loaded, as operated by a sophisticated user (it drives part of a process experiment). But, the user is proving electronically inexperienced.

Normally it switches 50A (screamingly audible--hearing protection required) and dissipates a few watts. Unloaded--i.e., under fault conditions--the FETs dissipate the entire input power. So instead of 3W conduction loss I'll get 150W of avalanche, and probably all in one device. That's the rub.

I included thermal shutdown, but I didn't expect to actually need it--the design wasn't meant for this kind of abuse.

That looks like the way to go. Easy, and all sorts of advantages...

Cheers,

James Arthur

Reply to
dagmargoodboat

Sounds good. A colleague told me woodworking outlets sell 24x24 inch mica panels, cheap. They use it for decoration.

Cheers,

James Arthur

Reply to
dagmargoodboat

I've seen that, but forgot. Sounds like a good combo.

Thanks,

James

Reply to
dagmargoodboat

Yes, hard anodized.

Teflon squeezes out in slow-mo. I like PTFE, but not compressed. Thanks for the thoughts.

Cheers,

James Arthur

Reply to
dagmargoodboat

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.