MT-200 Thermal properties?

[dshortt9]

Using mica and grease what temperature rise can be expected? I would think somewhere around 0.5 to 1 degree C per watt besides the 0.65 junction to case rise but would like to know what you have experienced so I can decide how many pairs of 2sa1216 2sc2922 to use for a Krell Clone at 60V rails and 3.5A of idle current. I am using a huge 12" x 20" x 3" fin heatsink per channel.

[MikeB]

Did you say 60v Rails wih 3.5A idle ?
Are you aware that you are talking about 840watts idle dissipation ? Not only that this amp will keep your room very warm, the costs for electricity will be horrible...

About the case to sink, the TO247 from IRF is rated 0.24°C/W, i guess the MT200 is better. You should be on the save side using 1°C/W total per device.
Okay, 420 watts to transfer for one channel, having the heatsink at 60°, keeping the junction at maybe 100° you need a total transfer of 0.095, that's ~10devices/5pairs at least. (per channel) That's only for idle dissipation... Are your speakers 4 or 8ohms ? You also need to do SOA calculations.

Mike

[dshortt9]

Well, if i use it around 3 hours a day that is about 90 hours a month times .840 kw = 75kwh @ 8.5 cents per kwh = $6.37 a month extra. Not too bad. If I left them on 24 hours a day it would be too much. $51 bucks. This is Wisconsin so the heat will be welcome 9 months out of the year.

[jacco vermeulen]

somewhere around 0.5 to 1 degree C per watt[/B]

Try again, nothing beats square inches.

That's the beauty of these MT200s, 70% more case to heatsink cooling surface than TO247s. Making them the best BJT devices for class A use, not counting TO3s.
It depends on the quality and thickness of the mica, the MT200 mica's i got are crystal clear and 0.0025" thick.

I used ceramic aluminium oxyde insulators on the KSA, two TO247 AlOx insulators from Avid Thermalloy fit under an MT200 if one of them is shortened. (shortening ceramic insulators is a PITA)
I have measured the transfer resistance of that combination to see how many watts i could get out of one device if i used those insulators.
2 shortened AlOx insulators under an MT200 case leaves enough room to glue a temperature reference transistor in the middle.
A piece of teflon sleeve and a heatsink with one big hole for the TO92 to pass and 2 smaller holes for the mounting bolts was enough to measure the temperature difference between sink and the MT200 case metal surface.
50 watts dissipation and 0.1 C accurate measuring translates to determining the Rcs with better than 0.01 C/W accuracy.
As thermal resistance has a reverse linear relationship with square area theoretically i've extrapolated the number i got to the full size MT200 cooling surface.
As Mike indicated, you should expect somewhere below 0.20 C/W for a MT200 mica.

[AndrewT]

Hi,
I tend towards Jacco's estimate. Definitely better than 0.5Cdeg/W
Various sources suggest To3 at 0.33 to 0.5Cdeg/W and I would expect MT200 to be better since the junction and it's mounting are in the middle whereas the To3 has the junction offset to miss the holes for emitter and base.

Is there any possibility of direct mounting the case (collectors) to an isolated sink? Even better (<0.1Cdeg/W) with just thermal compound.

[MikeB]

Okay, with a 0.2C°/W transfer and 0.36C°/W junction to case, you end up with a 0.56°C/W total. In this case you could get away with 6 devices / 3 pairs.

Mike

[jacco vermeulen]

If you compare the metal heatsink touching area of a TO3 and an MT200 device you'll find that they differ very little.

The MT200 case is little prone to warping, the regular TO3 case is.
(ref : Timoshenko, Mechanics of Materials)
The Hitachi TO3 Mosfets were pretty thick plated, measured at 0.35 C/W R case-to-sink mica insulated.
But that number only counts for the smoothest heatsink area, good quality thermal goop, and no air bubbles !

[MikeB]

"measured at 0.35 C/W Rcs."

Case to sink or total ? (total = junction to sink)

Mike