you can read what I wrote about sound (not much, being overwhelmed/saturated with overall situation in my Workshop) of Old Soul, having a look at THD Spectra
can't pinpoint on anything wrong, but sound was having some peculiar character on which I'm not exactly used......
planning to make one as Summer amp, to get used
can't pinpoint on anything wrong, but sound was having some peculiar character on which I'm not exactly used......
planning to make one as Summer amp, to get used
I was going to make my own chassis using the large heatsink I acquired but have decided to reserve it for another project I have in mind so I think I'll take the lazyman's option of buying a chassis for my THF51 follower project from Modushop. The question is which size heatsink?
I am looking at the 5U heatsink, 300mm x 40mm, rated at 0.18°C/W - I'm a simple soul so have just assumed a 36V power supply and 3A of current through the SIT:
36V x 3A = 108W
108W x 0.18 = 20°C rise in temperature of the heatsink above ambient.
Could probably get away with the 4U version of the heatsink:
108W x 0.31 = 34°C rise in temperature of the heatsink above ambient.
Does my thinking look about right?
Thanks
Your thinking is about right, but those dissipation factors are not de-rated. Anecdotally (and backed by some limited data), I'd expect about 50% of that performance on the conservative side, and maybe up to 70% on the aggressive side. I'd ballpark the 34 degree rise for the larger of the two sinks listed.
The specs might also have a time element in it. After 2 hours my F5 with two 5U sinks gets rather hot, still able to touch; it goes slowly. Luckily I use the Toshiba 2SK176 and 2SJ56 with a neg tempco.
But my experience is not science. 
Thank you guys, you're right that I didn't derate the heatsink so thanks for giving me the nudge and the estimates to work with.
So if we play it conservatively, the temperature rise of the 5U 30cm heatsink will be about 40°C above ambient so will reach perhaps 60-65°C - that's hot but not enough to burn if you touch it for a short time.
Stepping up to a 5U 40cm heatsink gives:
108W x 0.14 = 15° rise, derated by 50% = 30°C above ambient = 50-55°C, however, the 40cm heatsink is actually made up of two 20cm sections - I'm planning on using a single active device so, even with heat transfer goop between the two sections there will be uneven heat transfer across the two sections. I guess this is where using a single 2SK180 on a chunky piece of aluminium will help to spread the heat across the heatsinks.
So if we play it conservatively, the temperature rise of the 5U 30cm heatsink will be about 40°C above ambient so will reach perhaps 60-65°C - that's hot but not enough to burn if you touch it for a short time.
Stepping up to a 5U 40cm heatsink gives:
108W x 0.14 = 15° rise, derated by 50% = 30°C above ambient = 50-55°C, however, the 40cm heatsink is actually made up of two 20cm sections - I'm planning on using a single active device so, even with heat transfer goop between the two sections there will be uneven heat transfer across the two sections. I guess this is where using a single 2SK180 on a chunky piece of aluminium will help to spread the heat across the heatsinks.
I don't have my 2SK180 amps running right now as I am running my summer amplifier to limit the heat added to my room during the warm weather.
However when I had them running at around 20 C ambient, they were dissipating about 90W and the 300mm x 200mm x 40mm Modushop heatsink (0.18C/W) was somewhere in the 40s degree C. The 2SK180 is attached to an angle 2 1/2" x 2 1/2" x 3/16" x 6" long. I thought I had purchased 1/4"angle but I just measured it. So definitely go with a thick angle. The angle got quite hot, definitely much hotter than 55C, and the 2SK180 body was hotter still. So the challenge is getting the heat from the 2SK180 to the heat sink. I would recommend a large, thick, and long angle. A tee would even be better.
The 2SK180 and angle expelled heat into the interior of the chassis so ventilation in the form of vent holes in the top and bottom plates are necessary. I also installed 1 1/4" high rubber feet on the chassis to allow better air flow under and into the chassis.
However when I had them running at around 20 C ambient, they were dissipating about 90W and the 300mm x 200mm x 40mm Modushop heatsink (0.18C/W) was somewhere in the 40s degree C. The 2SK180 is attached to an angle 2 1/2" x 2 1/2" x 3/16" x 6" long. I thought I had purchased 1/4"angle but I just measured it. So definitely go with a thick angle. The angle got quite hot, definitely much hotter than 55C, and the 2SK180 body was hotter still. So the challenge is getting the heat from the 2SK180 to the heat sink. I would recommend a large, thick, and long angle. A tee would even be better.
The 2SK180 and angle expelled heat into the interior of the chassis so ventilation in the form of vent holes in the top and bottom plates are necessary. I also installed 1 1/4" high rubber feet on the chassis to allow better air flow under and into the chassis.
Foraging through my spares stash I found a couple of -12V ultra-low noise regulator modules (unused from a previous project) - is there likely to be any benefit using them in the bias supply instead of the LM337? They can only deliver 500mA instead of LM337's 1.5A but I don't see that as a factor.
In other news, my LL1694 inductors have shipped.
In other news, my LL1694 inductors have shipped.
Last edited:
I don't believe cap recommendations hold much value outside the context of a particular system. Used a 33uF Mcap in place of an electrolytic for the only reason i use the same caps in my valve amp bias, where they sound better than all other caps i tried. Whenever the follower gets a more permanent housing i'll probably revisit the parts selection again.
Even 500mA is much, much more than the current demand, which is much less than 1mA.
Another point - depending on the DCR of your load choke for the follower, the voltage needed from the bias voltage supply may be very low. The voltage drop across the load choke at the source may end up providing most of the Vgs if your SIT has low Vgs.
I recently purchased a 2SK180 to replace the one that has high gate leakage current. This 2SK180 is nearly 100 percent biased by the 1 Ohm DCR of the Hammond 193V at the source. I had to adjust the minimum voltage of bias voltage supply to near zero in order get 2.3A Iq. With this particular 2SK180 I can do without the bias voltage supply.
So beware of the SITs with very low Vgs in choke loaded follower mode. A combination of low Vgs and high choke DCR may limit Iq.
I recently purchased a 2SK180 to replace the one that has high gate leakage current. This 2SK180 is nearly 100 percent biased by the 1 Ohm DCR of the Hammond 193V at the source. I had to adjust the minimum voltage of bias voltage supply to near zero in order get 2.3A Iq. With this particular 2SK180 I can do without the bias voltage supply.
So beware of the SITs with very low Vgs in choke loaded follower mode. A combination of low Vgs and high choke DCR may limit Iq.