• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Williamson first stage question

And think the arrangement from #177 is a winner. With 70-72 mA the zenner didn't heat more than 55C (131F). 31.5V over zener, meant 2.27W but at 55 degree is not excessive hot. I even can keep the finger on the case for seconds...The diode derate with 1.5V with summed errors 1.5x12 = 18V but copper mass have enough thermal inertia to keep it thermic steady over power transients so I think even derated voltage will be steady with musical signal so can be considered like a 31.5V diode. Of course if the screens will draw max currents for long period that will cool down the diodes and will clamp 18V down. But on the power transients I don't think will be the case. Thanks very much all of You gents for support.
Late: carefully measures should be taken when short soldering on high copper mass still.
 
45*C , 70mA , 2.15W, voltage as close to nominal 30.5V. Heatsinks 1x2 inch each.

20240626_083040.jpg
 
2W from natural convection would probably put Tj at the 200C Si limit. The case could be hot enough to burn flesh. If you were only using ONE, and it was mounted to a .062 FR4 with 4 square inches of 2 oz Cu on each end it might be good for a couple watts reliably. That would be a typical power supply application. They can also tolerate much higher non repetitive surge currents than one watters, in applications where they suppress transients.

I don’t think they had tube amp shunt regulators in mind when they made these.
For the thick pcb 3 oz - 0.0105 mm double sided not works as well as individual heatsinks. At same 2 square inch for each end (one inch per pcb side x2) I used, I ended with almost 90*C instead 50*C with 0.5 mm thick heatsinks. I.m sure it will work better with 4 square inch per side as You said but that's excessive large for my app please. All temps measured at the end of the case.
 
Last edited:
Heatsinking is trying to use low thermal resistance paths to ambient air temp. You can attempt to do that with one interposing item of metal, like post #182, which provides a low thermal resistance from the device junction to a large surface area for convection cooling into local ambient air. The inconvenient aspect of that layout is that you then have to physically mount multiple such units and electrically link them all.

In post #142 you started with all devices on a convenient single board, that could be more easily mounted and linked. However that board needed to be enhanced with added fins, and a way of lowering the thermal resistance from each device to a local fin (ie. not by a significantly long lead length). One option would be to solder a fin to each pcb section, and to simplify retrofitting without repositioning each zener, the fin could be also soldered directly to the zener lead where it exists its package. Each zener could remain as shown, or repositioned to sit on the pcb and use the pcb land as the close metal heat spreader.

A one-to-one comparison would use fins of the same surface area for both #142 and #182.

A typical rule of thumb for silicon devices is to try and keep Tj below about 100-110C, as that allows some margin against poorer ambient air flow or short duration increase in device dissipation. You can try and estimate Tj be measuring a spot temperature close to a device, but be mindful that the junction would be somewhat hotter than the measured temp, and the probing of the temp may be wicking away heat, or otherwise have inaccuracies.
 
Last edited:
  • Like
Reactions: 1 user
Thanks. Was more like that. Same area taking in consideration the both sides. But half air exposed surface.
That.s interesting- in the same way efficiency convection was loose in the case the heatsink was folded in 'U' shape. It makes the inner face of ''U' less effective, despite the fact almost 3/8 was in between. The difference was 10*C more than 'L' shape for the same heatsink.Than the spacing between fins becomes a problem. Minimal room should be preserved to keep the same heatsink efficiency for a determined air flow (natural in this case).

20240627_031213.jpg
 
Last edited:
Now, a safety measure question please.: The regulator will hang exposed out of chassis. I Intend to protect the hardware under voltage with a metalic mesh cage. That is rigid enough to prevent accidentally touch. Still just in case of shock deformation of the mesh... or some making the high voltage contact possible between heatsink under voltage and mesh cage I ask if a fused protection of circuit and a hard wiring of the metalic mesh directly to the earthing nut may be just enough please ? Thanks.
 
Last edited: