Heatsink
Good question about heatsink, maybe I should just go and buy a silicon pad..............and throw some marmite in for good measure😀
Good question about heatsink, maybe I should just go and buy a silicon pad..............and throw some marmite in for good measure😀
In the case of separate power supplies... YES. Combining amps on a common heatsink would create problems.
😉
GEEZE... I hate being quoted when I msispell something.
dfdye... do the math... the the thermal resistance is not pretty. Easily 10-20 C from chip to sink with insulators in place.
😉
GEEZE... I hate being quoted when I msispell something.
dfdye... do the math... the the thermal resistance is not pretty. Easily 10-20 C from chip to sink with insulators in place.
Endzone, those heat sinks look SWEET!!!! I have no idea how they will work but they will look great doing it.
Taj, Yes, the sweat will indeed be conductive. Electrolytes and all. . . 😀 (I just love answering silly questions like that! 🙂 )
Poobah, I must admit I haven't done the math, so I'll defer to your numbers, but that seems a little high. What insulators are you assuming? Either way, I think you are right that for a chip this small relative to its required heat dissipation make skipping the insulator a good idea if you plan to drive the chip hard.
David
Taj, Yes, the sweat will indeed be conductive. Electrolytes and all. . . 😀 (I just love answering silly questions like that! 🙂 )
Poobah, I must admit I haven't done the math, so I'll defer to your numbers, but that seems a little high. What insulators are you assuming? Either way, I think you are right that for a chip this small relative to its required heat dissipation make skipping the insulator a good idea if you plan to drive the chip hard.
David
dfdye,
You will find that mica insulators start out better than silicone and then gradually degrade. Silicone starts out a near second but will improve.
In both cases, benefits are derived by spring based mounting.
I think if you look at the datasheet for an LM3875 there are some typical values for thermal resistance through mica, grease, etc...
What I would like to see is for some to reflow solder the chip onto a copper heat sink... that would ROCK!
You will find that mica insulators start out better than silicone and then gradually degrade. Silicone starts out a near second but will improve.
In both cases, benefits are derived by spring based mounting.
I think if you look at the datasheet for an LM3875 there are some typical values for thermal resistance through mica, grease, etc...
What I would like to see is for some to reflow solder the chip onto a copper heat sink... that would ROCK!
I'm half way there. I have a home made copper heat spreader/riser block on my LM4780's since I made the pcb's a shade too deep on the back side. Unfortunately, no reflow solder for me! 🙁poobah said:What I would like to see is for some to reflow solder the chip onto a copper heat sink... that would ROCK!
(And yes, I know the additional surface transition will drop the sinking efficiency, but sometimes you gots to do what you gots to do. . . . )
Poobah
Cheers for that. So..............uninsulated means seperate heatsinks and a floating heatsink from main case. If i insulate the two chips it should be o.k to keep them on the same heatsink, although i still run the risk of the pads breaking down at some stage and destroying the chips.
dfdye: yes they are quite cool, they came with the case, it was an older home theatre unit which i stripped. Managed to get a tranny out of it as well. This might come in handy with the LM4702 board, secondaries are 30-0-30, maybe a little low?
Brett
Cheers for that. So..............uninsulated means seperate heatsinks and a floating heatsink from main case. If i insulate the two chips it should be o.k to keep them on the same heatsink, although i still run the risk of the pads breaking down at some stage and destroying the chips.
dfdye: yes they are quite cool, they came with the case, it was an older home theatre unit which i stripped. Managed to get a tranny out of it as well. This might come in handy with the LM4702 board, secondaries are 30-0-30, maybe a little low?
Brett
dfdye said:
4780 > paste > spreader > silpad > heatsink could be better than 4780 > silpad > heatsink depending on the particulars.
If I have to buy an oversize sheet of silpad, I might go that route.
I'm going with no insulation and floating the sinks. I figure it's just as easy to insulate the sink as the chip. Besides, Poobah did the math, and who am I to argue! 😀jimbo1968 said:
Actaully df,
I didn't do the math for this one... but I have done too many times before.
Semiconductor life is halved for every 11 C rise in operating temp... so cooler is better.
A fair rule is that if you can't hold your finger on a semi for at least 5 seconds... you should expect trouble.
😉
I didn't do the math for this one... but I have done too many times before.
Semiconductor life is halved for every 11 C rise in operating temp... so cooler is better.
A fair rule is that if you can't hold your finger on a semi for at least 5 seconds... you should expect trouble.
😉
live heatsinks
Because of the predrilled and prethreaded holes on the heatsink tunnel I used for my Krell clone (its pictured on the link in my sig line) I implemented "live" heatsinks with just some thermal paste (guess it doesn't matter to use conductive paste on this but at the time I had the normal white nonconductive stuff) between the TO-3 package and heatsink. Works well but-
a live heatsink is a big antenna, it will pick up noise (from xfo, from other sources) no big deal if its isolated electrically but if its in the circuit you have to be careful to shield it from other stuff. Mine is in a separate compartment shielded from the XFO's by a steel plate which works pretty good. To shield from the xformer, use a metal with iron in it (tin and steel) not aluminum.
It is dangerous and you really have to be careful, I know my sinks are live but still I've caused them to arc one time. Had the DMM probes in the current measuring positions and probed the sinks to see what voltage they were at...not pretty! I still have to think carefully when I go in there to measure stuff. You will want bleeder resistors to get the sinks down to zero volts after about 30 seconds so its safe to work in there.
Because of the predrilled and prethreaded holes on the heatsink tunnel I used for my Krell clone (its pictured on the link in my sig line) I implemented "live" heatsinks with just some thermal paste (guess it doesn't matter to use conductive paste on this but at the time I had the normal white nonconductive stuff) between the TO-3 package and heatsink. Works well but-
a live heatsink is a big antenna, it will pick up noise (from xfo, from other sources) no big deal if its isolated electrically but if its in the circuit you have to be careful to shield it from other stuff. Mine is in a separate compartment shielded from the XFO's by a steel plate which works pretty good. To shield from the xformer, use a metal with iron in it (tin and steel) not aluminum.
It is dangerous and you really have to be careful, I know my sinks are live but still I've caused them to arc one time. Had the DMM probes in the current measuring positions and probed the sinks to see what voltage they were at...not pretty! I still have to think carefully when I go in there to measure stuff. You will want bleeder resistors to get the sinks down to zero volts after about 30 seconds so its safe to work in there.
Re: live heatsinks
You can throw a bag over them while you fiddle, like in the car repair business. (an old Wall Mart bag, or your mom)
Live heatsinks shouldn't be a problem in a steel chassis.
lgreen said:
You can throw a bag over them while you fiddle, like in the car repair business. (an old Wall Mart bag, or your mom)
Live heatsinks shouldn't be a problem in a steel chassis.
Live sinks are indeed an antenna. But, this is a power rail... it is dirty in the first place. And, local decoupling closer to the chip should prevent any problems.
🙂
🙂
Netasinks
David,
Have you got a picture of your chips mounted onto your heatsinks.
It would nice to someone elses work, to compare
Brett.
David,
Have you got a picture of your chips mounted onto your heatsinks.
It would nice to someone elses work, to compare
Brett.
When listening at high (Rock Concert) levels, I have no problem with placing a fan on my heat sink.
When my buddies want to hear "Green Grass and High Tides" as if they were in the front row, I defy ANYONE to tell me the 80mm fan is producing noise.
And the heat sinks stay cooler than ambeint.
When my buddies want to hear "Green Grass and High Tides" as if they were in the front row, I defy ANYONE to tell me the 80mm fan is producing noise.
And the heat sinks stay cooler than ambeint.
davidlzimmer said:
My latest chipamp, "sparky" is a Sun SPARC workstation that has a fan included, so i'm going to use it. I'm sure its old and noisy but i've put in a speed control to slow it down if I need to. Assuming I dont blow up all my chips.
There are times when I want a zero ambient noise level. The sound of the central air makes more noise than the fan. Once EVERTHING has been silenced, I turn off the fan.
Silence, modulated by wonderful music.
I think way too much has been made about fan noise. Turn the thing off if you can hear it! How bout a PC based system? Put a fan on it and forget about it.
Silence, modulated by wonderful music.
I think way too much has been made about fan noise. Turn the thing off if you can hear it! How bout a PC based system? Put a fan on it and forget about it.
Silicon Pads
My silicon sheet turned up today....
It's 300x300mm, and with me guessing a LM4780 pad would be roughly 30x20mm; that works out to be 150 silicon pads at$0.33 ea. Not that I'm ever going to use 150 of the little buggers.😉
So it looks as though I wont be worring about floating the heatsink.
Brett.
My silicon sheet turned up today....
It's 300x300mm, and with me guessing a LM4780 pad would be roughly 30x20mm; that works out to be 150 silicon pads at$0.33 ea. Not that I'm ever going to use 150 of the little buggers.😉
So it looks as though I wont be worring about floating the heatsink.
Brett.
- Status
- Not open for further replies.