How do you tie the LM3886 to the heatsink when thermal pad is used. Can 6-32 Nylon or plastic screw really do a solid job ? Or you still need metal screw with some kind of insulation.?
A nylon screw would probably hold well enough, but the best thing would be some sort of clamp that goes across the front of the case and has two screws into the heat sink. A metal bar of small piece of metal plate would work good.
Peace,
Dave
a "nylon" screw is unlikely to give adequate clamping force when cold.
I have no idea whether it can maintain any significant clamping force at elevated temperatures.
The clamp bar using a pair of screws placed closely to either side of the chipamp does a better job of thermally clamping to the heatsink. The bar must be stiff enough to reduce bending when it is clamped up. The bar and bolts will be at the same voltage as the heatsink.
I have no idea whether it can maintain any significant clamping force at elevated temperatures.
The clamp bar using a pair of screws placed closely to either side of the chipamp does a better job of thermally clamping to the heatsink. The bar must be stiff enough to reduce bending when it is clamped up. The bar and bolts will be at the same voltage as the heatsink.
The strength of any bar as a clamping device is going to be directly related to it's thickness, and how close together the screws are. If you use a thick enough piece of aluminum, and keep the retaining screws close to the case of the device (without touching it) then you will get maximum strength. Of course in doing this you will then need to be careful not to over torque the screws and break the case of the device being held.
I agree with AndrewT's statement about the bar being at the same voltage potential as the heatsink, this is why it is important to use a proper isolation pad between the heatsink and the device. Of course we all knew that
Peace,
Dave
P.S. you might use the nylon screw as an initial hold down, to make it a little easier to keep the device (LM38xx) in place while you install the clamp bar.
I agree with AndrewT's statement about the bar being at the same voltage potential as the heatsink, this is why it is important to use a proper isolation pad between the heatsink and the device. Of course we all knew that
Peace,
Dave
P.S. you might use the nylon screw as an initial hold down, to make it a little easier to keep the device (LM38xx) in place while you install the clamp bar.
Member
Joined 2009
Paid Member
a "nylon" screw is unlikely to give adequate clamping force when cold.
I have no idea whether it can maintain any significant clamping force at elevated temperatures.
The problem is that overtime the nylon stretches and the clamping force can not be relied on.
Why don't you use the insulated LM3886TF? This package will work for most people.
My TFs have been abused for about four years without any thermal issues or failures. The only mounting I use is a 3mm Allen head screw. You might be over-thinking this.
Abs
The nylon just like any and every bolt stretches as soon as the nut applies tension to the fixing.The problem is that overtime the nylon stretches and the clamping force can not be relied on.
The difference in E (Young's Modulus) value from steel to nylon will be many decades of stiffness.
i.e. the cold nylon will stretch, a lot. The cold nylon will creep. The cold nylon will not provide adequate clamping force, cf. a steel bolt/screw of similar diameter.
As I said 4 years ago, I have no idea what happens when it gets hot.
DON'T use nylon clamping bolts.
Nylon bolts to locate PCBs will be OK for most lightweight builds.
Why don't you use the insulated LM3886TF? This package will work for most people.
Yeah, and use some of that silver conductive grease between the chip and the heatsink.
My TFs have been abused for about four years without any thermal issues or failures. The only mounting I use is a 3mm Allen head screw. You might be over-thinking this.
The thermal resistance of the TF package is about 2 K/W. That of the T package is 1 K/W. Both are junction-to-case numbers. Even when you account for the thermal washer, it means that if you're pushing significant power through the LM3886, you really need the non-isolated T package.
If you're using an external heat sink, keeping its temperature to safe levels (<60 ºC), you can generally get away with the TF package if using +/-28 V rails and 8 Ω loads...
Some of the TO-247 sized pads are actually large enough for the LM3886. The Bergquist SP900S-0.009-00-114 for example (Digikey P/N: BER180-ND). The TO-220 type shoulder washers should work with the LM3886 as well. That way you can use a metal machine screw. Using nylon screws is a recipe for disaster. You'll never get enough clamping pressure and the thermal cycling will weaken the screw.
~Tom
Last edited:
Best is to punch it in a male/female die and that´s what factories do; at home you can drill a clean hole through with a sharp new drill and going slow,so you let drill edge actually *cut* kapton, which, after all, IS "plastic".
You´ll end with a round hole but a slightly raised burr around, which you remove by twisting a larger diameter drill (8-10mm) by hand.
Do it both sides.
You´ll end with a round hole but a slightly raised burr around, which you remove by twisting a larger diameter drill (8-10mm) by hand.
Do it both sides.
Laser cutter will do a fairly good job, though you have to clean the edge:
Processing Kapton(R) Using Laser Technology
Processing Kapton(R) Using Laser Technology
Attachments
Why not just use a TO-247 or TO-3P sized SilPad (or Keratherm Red)?
As Mark points out above, the purpose of the various pads and gap fillers is to fill the air gaps. The pad material is not super thermally conductive. Neither is thermal grease. But they're much more conductive than air.
Many of the thermal pads are electrical insulators. "Many". I.e., not all. It's worth double-checking the dielectric strength of the pads before you buy many of them.
If using Kapton pads, I would put thermal grease on both sides. Just as I would use thermal grease on both sides of a mica washer. I'm not sure about the adhesive pads, though.
If you're using silicone pads, there's no need for thermal grease. That's the point of the SilPads.
Tom
As Mark points out above, the purpose of the various pads and gap fillers is to fill the air gaps. The pad material is not super thermally conductive. Neither is thermal grease. But they're much more conductive than air.
Many of the thermal pads are electrical insulators. "Many". I.e., not all. It's worth double-checking the dielectric strength of the pads before you buy many of them.
If using Kapton pads, I would put thermal grease on both sides. Just as I would use thermal grease on both sides of a mica washer. I'm not sure about the adhesive pads, though.
If you're using silicone pads, there's no need for thermal grease. That's the point of the SilPads.
Tom
Does anyone have a model number of a shoulder washer usable with the LM3886T (metal tab) to be used with a 4-40 screw ?
There are so many of these things, it is mind boggling. I dont want to make a mistake so someone who may already have gone thru this experience will know much better than me.
Thanks
There are so many of these things, it is mind boggling. I dont want to make a mistake so someone who may already have gone thru this experience will know much better than me.
Thanks
- Home
- Amplifiers
- Chip Amps
- LM3886 thermal pads/insulators