Digi-key sells Berquist Silpads to fit a TO-247 case. These should also fit a TO-3P case. It would help to know exactly which cse you really mean by a TO-3P, whether it is the old TO-3P case with exposed metal tab like the TI TIP34 series or something else. In any case, the TO-247 insulator should work just fine.
These are On-Semi MJL3281's (TO-264 case) for a Class A amp.
I wanted mica because I really expect to be experimenting a bit, and a sil-pad would have to be cleaned off and replaced every time I remount the device. Mica, on the other hand...
wrenchone...I honestly did not see the TO-247 Berquist pads at Digikey. They appear that they will work, but I'd still rather have mica.
I wanted mica because I really expect to be experimenting a bit, and a sil-pad would have to be cleaned off and replaced every time I remount the device. Mica, on the other hand...
wrenchone...I honestly did not see the TO-247 Berquist pads at Digikey. They appear that they will work, but I'd still rather have mica.
Sil pads
RS Components have these
2015-122 TO-3P 283-3672 £1.58 £1.34
http://rswww.com/
At £15 pounds for 10 they seem expensive
technical specification
Thickness: 0.38 mm
Thermal resistance: 0.20°C/W (for TO-3 case)
Electrical insulation: 4 kV
Thermal conductivity: 3.5 W/m-k
This looks like they conduct heat well.
These are a lot cheaper.
3223 07 FR-122 TO-3P 283-3638 £0.28 £0.24
At 28p they are a lot cheaper but you have to buy 50 so you still have to spend £14 pounds.
technical specification
Thickness: 0.18 mm
Thermal resistance: 0.40°C/W (for TO-3 case)
Thermal conductivity: 0.9 W/m-k
Electrical insulation: 4 kV
I presume that there not much point in having a massive expensive heat sink if the heat from the transistor is no passed to it???
Looking a some of the posts I may has well just stick some insulating tape to the bottom of the transistor and bolt it up.
I'm confused
RS Components have these
2015-122 TO-3P 283-3672 £1.58 £1.34
http://rswww.com/
At £15 pounds for 10 they seem expensive
technical specification
Thickness: 0.38 mm
Thermal resistance: 0.20°C/W (for TO-3 case)
Electrical insulation: 4 kV
Thermal conductivity: 3.5 W/m-k
This looks like they conduct heat well.
These are a lot cheaper.
3223 07 FR-122 TO-3P 283-3638 £0.28 £0.24
At 28p they are a lot cheaper but you have to buy 50 so you still have to spend £14 pounds.
technical specification
Thickness: 0.18 mm
Thermal resistance: 0.40°C/W (for TO-3 case)
Thermal conductivity: 0.9 W/m-k
Electrical insulation: 4 kV
I presume that there not much point in having a massive expensive heat sink if the heat from the transistor is no passed to it???
Looking a some of the posts I may has well just stick some insulating tape to the bottom of the transistor and bolt it up.
I'm confused
What if the Heatsinks are hard coated (hard anodized?) and the transistor just mounted directly to the sink? In another thread, there was confidence that the electrical insulation and the mechanical resistance of the coating is good enough to ensure safe operation.
And I guess, the thermal resistance of the hard coating isn't higher than that of a insulating pad. But I don't have any numbers.
Ciao, Tino
And I guess, the thermal resistance of the hard coating isn't higher than that of a insulating pad. But I don't have any numbers.
Ciao, Tino
zinsula said:
when you drill through you have to make sure that you use an insulated shoulder washer -- and you still need the thermal paste -- the paste is thermally conductive for any voids or imperfections either on the heatsink or on the semiconductor.
if I had to guess I would think that not using an insulating pad would raise a potential hazard such that Underwriter's Lab (in the U.S.) would not certify the product as "safe".
I am unaware of any standard thermal paste that is electrically conductive at the kinds of voltages we see in audio gear.
If any of the the usual products were conductive, sloppy me would have left a trail of non-op amps behind me, and it didn't happen! ;-)
neckro of the year?
when doing metal to metal thermal contact, AND you need an insulator, relying on a paste for electrical isolation is foolhardy.
do you want paste on thermal pads? the word "pads" implies insulator with thermal coatings.
use plain old white thermal grease and a proper insulator , IMO some of the pastes tend to dry out sooner than others.
if you explain in more detail what your application is "the questions" will become clearer.
when doing metal to metal thermal contact, AND you need an insulator, relying on a paste for electrical isolation is foolhardy.
do you want paste on thermal pads? the word "pads" implies insulator with thermal coatings.
use plain old white thermal grease and a proper insulator , IMO some of the pastes tend to dry out sooner than others.
if you explain in more detail what your application is "the questions" will become clearer.
Last edited:
redjr,
I think that aluminum thermal compound is electrically conductive. Good for CPUs but I wouldn't recommend it for amps.
The type that is normally used is silicon thermal paste. It is non-conductive. Simply spread it evenly on both sides of the mica insulators and bolt the transistors to the heatsink.
If you don't want the mess, go for something like silicon pads. No thermal paste needed.
I think that aluminum thermal compound is electrically conductive. Good for CPUs but I wouldn't recommend it for amps.
The type that is normally used is silicon thermal paste. It is non-conductive. Simply spread it evenly on both sides of the mica insulators and bolt the transistors to the heatsink.
If you don't want the mess, go for something like silicon pads. No thermal paste needed.
redjr, they make commercial silicone non-conductive heat sink grease, and it is readily available.
If you are not using a pad or mica wafer, then the conductivity is not an issue, because as infinia said, expecting paste of any sort to act as insulator is foolish.
Examples:
8616-3ML MG Chemicals | Mouser
10-8108: GC ELECTRONICS: Test, Tools & Supplies
If you are not using a pad or mica wafer, then the conductivity is not an issue, because as infinia said, expecting paste of any sort to act as insulator is foolish.
Examples:
8616-3ML MG Chemicals | Mouser
10-8108: GC ELECTRONICS: Test, Tools & Supplies
I didn't think I could make myself any clearer, but apparently I do. I don't need any paste simply to use as an insulator, nor do I expect any to do so. You can use paste and an insulator together though. Maybe I don't need to, but I like the idea of paste filling in the imperfections of the metal before the thermal pad.
For pieces of aluminum angle or channel, since they will be used as an extension for the heatsink, I simply want to use a good thermal paste between them for better heat transfer - not any electrical conductivity.
Rick
How well do the alumina (oxide) insulators work? Does anyone have data or a reference?
It would seem that the much increased thickness would work against overall thermal conductivity, even with alumina being a better thermal conductor than mica in general. I looked, but couldn't find data comparing mica insulators vs. alumina insulators. The alumina ones look more 'serious' though!
It would seem that the much increased thickness would work against overall thermal conductivity, even with alumina being a better thermal conductor than mica in general. I looked, but couldn't find data comparing mica insulators vs. alumina insulators. The alumina ones look more 'serious' though!
why even mention non-conducting pastes if it's a don't care?I didn't think I could make myself any clearer, but apparently I do. I don't need any paste simply to use as an insulator, nor do I expect any to do so. You can use paste and an insulator together though. Maybe I don't need to, but I like the idea of paste filling in the imperfections of the metal before the thermal pad.
For pieces of aluminum angle or channel, since they will be used as an extension for the heatsink, I simply want to use a good thermal paste between them for better heat transfer - not any electrical conductivity.
Rick
thermal pads don't need anything in fact by you adding stuff spoils the performance, read the data sheets. Do the thermal analysis! (it's as easy as ohms law) if one thermal interface is real problem, then a 'magic grease' isn't the right solution. IMO full stop. If one thinks the commercial stuff isn't good enough... looking all the CPU exotic pastes really aren't any better, 'Tom's Hardware' has done several articles you can look up. Spend the money on more surface area and getting the machining / mechanics right. Angle stock thickness , flatness, proper hardware torqueing sequence and all that. Any improvements to ventilation / and air flow is HUGE in the area of heat management.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.