Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Transistor isolation question

Status
Not open for further replies.

hoyt

Newbie level 5
Joined
May 22, 2019
Messages
10
Helped
2
Reputation
4
Reaction score
2
Trophy points
3
Activity points
91
Hello:

I have a unit with a very thick aluminum box (earth grounded), which is used also as the heatsink. The transistor is a standard TO-220, running at line voltage (115Vac). The all-plastic style TO-220 was tried, but simply gets too hot (much worse thermal parameters than standard to-220). The TO-220 uses a large sil pad & shoulder washer & gets very warm/hot, but not "burning up".
So the unit is operating very fine.

However, even with a large area sil pad & big shoulder washer isn't something missing for creepage/clearance?

It seems to me that at the hole edge, the metal tab of the transistor is only a hair away from the metal of the housing (separated by the hair-thin thickness of the sil pad). Even with the shoulder washer, how does this provide enough creepage/clearance? So how do shoulder washers offer any help when this path still exists? The path from the tab to panel still exists on the outside of the smaller portion of the shoulder washer. I haven't seen this mentioned anywhere in terms of UL testing or anyone trying to resolve it...maybe not an issue?
 

I've used standard nylon shoulder washers on TO-220 devices at over 1KV between tab and heatsink with no problem. The manufacturers specify breakdown voltage of mounting washers and heat sink pads but it is normally far higher than line voltage.

Brian.
 

I've used standard nylon shoulder washers on TO-220 devices at over 1KV between tab and heatsink with no problem. The manufacturers specify breakdown voltage of mounting washers and heat sink pads but it is normally far higher than line voltage.

Brian.

The washer and silpad have "nothing" to do with it. The transistor tab hole (inside bottom perimeter) & the edge of the threaded hole are only kept from touching by the thickness of the silpad (through the hole in the silpad. The shoulder washer does nothing to alleviate this. Is the a common problem for UL approval?
 

I think you are confused about what the washer does. It has a wide 'top' and a narrower center tube that goes right through the transistor tab and into the heatsink hole. The mounting screw head sits on the 'top' (usually with a washer to spread the load) and the thread passes inside the washer tube to a nut or threaded section on the heat sink. If you look at the shoulder washer side on, it looks like a letter 'T' with it's center line running through the transistor and into the heatsink.

The mounting screw and tab cannot touch each other and there is no open gap between the tab and heatsink.

Brian.
 

I am NOT talking about the screw at all. Put the screw & washer in the trashcan a moment. I am talking about the tab hole to heatsink thread THROUGH the silpad hole...there is only a thin air gap between between these metals, of the silpad thickness--correct? This does not change when the shoulder washer is inserted. TO220.PNG
 

correct the silpad thickness is the isolation thickness - this is why you need to know the limitations of the exact sil-pad you are using - 3500Vac for one min? The "top hat" insulating bushing is also very important - the thicker the bushing and longer the more creepage you have ... I'm not sure any TO-220 bushing system gives proper mains isolation. To earth may be OK as it will just blow the fuse for a short - but i/p to o/p -- a no go ...

- - - Updated - - -

it is usual to countersink the Al h-sink underneath the TO-220 to give even more creepage - the countersink bevel must be longer than 1.0 mm to count in the creepage path ( standards )
 

it is usual to countersink the Al h-sink underneath the TO-220 to give even more creepage - the countersink bevel must be longer than 1.0 mm to count in the creepage path ( standards )

Do you have any examples of this (links, app notes, etc)? I have never seen that described or mentioned anywhere
 

I have several thousand examples out in the field - IEC60950 gives ample explanation of creepage - in the world of power electronics ( as other endeavours ) one is limited only by one's imagination - usually ...

- - - Updated - - -

countersinking will still not provide the i/p to o/p separation required by 60950 for example - nor the 2.5kV AC input to earth, for TO-220.
 
  • Like
Reactions: FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
Thanks for the app note...I have seen it a while back & it does not mention countersinking the hole under the tab (but perhaps a viable idea)...In fact the diagram you posted shows exactly the issue I was worried about (high voltage jump along green line) & no countersinking. I've seen lots of articles & catalogs about shoulder washers, but never see this mentioned.

Creep1.PNG
 

what's o/p i/p?

Is my concern a real issue? I don't want to "invent a problem"...if it is a problem, why can't I seem to find any mention of it anywhere? Is no one else having this issue? All I find is "use a shoulder washer" type suggestions, but the shoulder washer isn't pertinent. How can a hole through widely-used sil-pads, or mica provide enough creepage for everyone?
 

Hi,

o/p = ouptut
i/p = input

I´m with you. I see the creepage as a safety problem.

Currently I have no idea how to mount the TO220 with a screw through it´s heat tab.

Thus my workaround is to not use the hole at all.
Use a solid (without hole) sil pad with the size to meed safty creepage specifications.
Then use an isolating part (glass filled plastics? 3mm FR4? Check mechanical stability!) that presses the tab with the use of (2?) screws to the aluminum case.

Klaus
 
Another alternative is to use "steel clips" to fix the power device onto heat-sink.High voltage devices are generally fixed with this technique.

Since this is mounted inside & to the grounded case, clip legs sticking out might get knocked loose or cut someone.
Our line voltage transistor is connected by wires, so it is only held in place by the screw/clip/whatever is being used.

I thought of using one of those "thermal sleeves" (looks like heat shrink tubing)---which gives perfect isolation. However I can't find a screw in clip/spring that would securely hold the TO-220 in place. Need a screw down springy finger that "wraps around" the transistor sides/thermal sleeve to prevent motion. Having a line voltage transistor flopping around is bad news!

Again, it seems strange that until I thought of this issue, I've never seen it brought up or mentioned at all these years (& I've seen lots of articles about heatsinks and shoulder washers).
 

Most companies design their own wrap around spring steel clips for TO-220 in a "tube" of insulation ...
 

Most companies design their own wrap around spring steel clips for TO-220 in a "tube" of insulation ...

Do you have some Examples or links I can take a look at? If this is a common issue, it seems the part would be readily available.
 
clips are very application dependent - hence the enormity of differing ones in different products - we get to see a fair few at our work - once again the designer is limited only by his / her imagination and knowledge
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top