Logically, I believe my adapters must have better thermal performance than simply bolting a TO-3P straight to a heatsink drilled for TO-3. There is simply more copper heat-spreader, and more surface area in contact with the heatsink. And the double-thick copper sandwich, formed by the TO-3P heat-spreader stacked on top of the adapter, has a lot more copper mass than even a regular TO-3.
But if course I need to SCIENCE this hypothesis to be sure! I don’t sell untested assertions.
The other question—and this is difficult to prove—Is the Adapter + TO-3P as good as, or better than TO-3? That’s a hard question to answer, due to the variety in construction of TO-3 packages. Some are all steel, some have a copper heat-spreader inside, and some have a copper heat-spreader that goes all the way through the case and is exposed to the heatsink. I make no claims that the adapter is always superior to TO-3. Depending on their construction, there may exist TO-3 devices with better thermal performance than these adapters. You must make that determination for yourself, based on the evidence presented. I believe that most cases, the adapter will be superior to an actual TO-3, due to its high copper mass, but since there is so much variability in the TO-3 package, and the technology of the semiconductor within, I make no guarantees. Some TO-3 packages have more or less copper inside. Their cases are made from different types of steel. Some semiconductor dies are rated for 150C, but some are 200C, and that too affects a device’s thermal considerations.
These are variables I cannot control. So, pack your own parachute!
AFAIK, all TO-3 cases are made of steel. Steel doesn’t conduct heat super well, but some alloys are better than others, so I’d guess the steel they use for transistor cases is at least 50W/mK. (Copper is 400W/mK)
High-power TO-3 transistors usually have a copper heat-spreader inside, like this ON Semi MJ21193.
And this Toshiba 2SB554
Some TO-3’s have a copper heat-spreader that goes all the way through the case, as on this Toshiba 2SB554. The rest is steel.
Due to it’s higher copper mass, the Toshiba 2SB554 should have slightly better thermal performance than the On Semi. This isn’t the whole story though; The MJ21193 is actually rated for more power than 2SB554; 250W @200C junction temperature, and the 22B554 is rated for 150W @150C. They both have the same size die and the Toshiba probably has a lower thermal resistance. So it would seem that the extra 100W rating of the MJ21193 comes from allowing the junction temperature to go 50C higher. Is the On Semi die actually able to reliably withstand higher heat than the Toshiba? Maybe it is. Technology moves on. Maybe Toshiba and On Semi have different ideas of safe power levels. In any case, both of these transistors are famously robust devices.
The experiment:
Very simple: We’re going to mount the transistors to a large heatsink, and plot the ∆ temperature rise with a constant 50W dissipated.
We’re comparing:
- TO-3P NJW1302 mounted straight to the heatsink.
- TO-3P NJW1302 on Hoppe’s Brain Adapter/Heat Spreader.
- A TO-3 Toshiba 2SB554 (Kind of apples-to-oranges, but interesting.)
Test setup:
A large heatsink is drilled and tapped for TO-3 mounting. A hole is drilled in the center for the thermocouple. Thermal pads for both devices are Parker CHO-THERM T441, 8 mil thick.
A 5K resistor is tied from V+ to base. Power supply is set to deliver constant power at 50W. This develops approximately 13V from C-E, at 3.8A.
Comparing Apples to Apples:
I’m taking the transistors temperature at a point directly below the die.
In order to take the transistor’s temperature from the exact same spot when the adapter is installed, a hole for the thermocouple is also drilled in the adapter. The thermocouple is touching the back of the transistor. This way, the measurement is taken at the same point, with 1.5mm of copper between the die and the thermocouple.
Comparing Apples to Pears:
The situation with the TO-3 device isn’t so simple. The TO-3 case is physically different from the TO-3P, making a direct comparison impossible. This Toshiba 2SB554 has a steel case, with a 2.8mm thick copper heat-spreader, that goes all the way through the case.
The TO-3P device has a 1.5mm thick copper heat-spreader, and the Toshiba’s heat-spreader is 2.8mm thick, so I drilled down 1.3mm so that the thermocouple would sit approximately 1.5mm beneath the die, same as the TO-3P. This is not a perfect comparison.
Results:
TO-3P NJW1302 mounted directly to heatsink – Start 23C, End 59.3C = ∆ 36.3C
TO-3P NJW1302 mounted to Hoppe’s Brain adapter – Start 23C, End 57.8C = ∆ 34.8C
TO-3 Toshiba 2SB554 – Start 22.9C, End 62.4C = ∆ 39.5C
Plots:
TO-3P NJW1302 mounted directly to heatsink
TO-3P NJW1302 mounted to Hoppe’s Brain adapter
TO-3 Toshiba 2SB554
The TO-3P levels out at 59.3C, and the adapter levels out 1.5C cooler, at 57.8C. That isn’t much, but the big difference is in the steepness of the temperature rise. Without the adapter, the temperature gets to 45C in about 6 seconds, and with the adapter, 8 seconds. 23C to 30C takes twice as long!
Audio has lots of peak power dissipation events, and the heat-spreader helps move the heat away from the die faster. Copper conducts heat almost twice as quickly as aluminum, so the more that heat can be spread out before hitting the aluminum, the better.
The Toshiba’s results are interesting, but again, taken with a big grain of salt because the physical construction is different. What I think is most interesting, is that the steepness of the temperature rise is about the same as the TO-3 adapter.
Conclusion:
Hoppe’s Brain TO-3 to TO-3P Adapters/Heat-spreaders perform better than mounting a TO-3P straight to a TO-3 heat sink. They are probably equal to, or better than, an actual TO-3 device.