trinityamps.com

I installed a pair of KT-66 tubes in my 1x12 Triwatt combo - it worked and sounded great!

After playing for a little while (deliberately at high volume), I could tell the amp was physically hotter than usual (e.g. top of the cabinet was warmer, I could even feel it a bit on the knobs). This makes some sense of course, the KT66s are bigger, more powerful tubes, and presumably put off more heat.

But one of the big chassis-mounted filter caps is really close to the power tubes. Furthermore, in a combo, it seems the natural ventilation is probably worse than in a head unit.

Is this a potential problem for the life of the big filter caps? Or am I worrying about a non-problem?

Thanks!

I built my Triwatt into a combo and it has the KT66's in it. Yes it does get hot.
I've used it in peak summer at various shows, indoor and outdoor with zero issues. I was thinking of possibly hooking up a small computer type cooling fan like my Mesa MK3 has but I've had it two years, constant use and zero issues so whether it's more trouble than it's worth I can't say at this point.

@intothefloyd - thank you for that voice of reassurance. To be clear, I don't expect any kind of failures any time soon, but I have a tendency to overthink things, along with a neurotic desire for overbuilt stuff.

@coco - I don't think the big 50-50uf/500v filter caps have any markings on them. Do you happen to have the details on them? In particular, temperature rating and ESR?

FWIW, I posed this question over at The Amp Garage. The gist of the responses is, yes, in general, this is a potential concern. But can be mitigated a number of ways, such as component layout, tube orientation, capacitor quality/temp rating, and airflow.

I'm thinking about drilling a 120mm hole in the rear cabinet panel (the one that covers the tubes), and mounting a 12v computer fan to it, powered by the low-voltage PT secondaries (i.e. what the tube filaments use). I think that would be a fairly simple modification, and should substantially reduce temps in the tube area. I don't see how it can hurt?

Looking at the Triwatt schematic, the C20 is the big 50-50uF/500v filter cap closest to one of the power tubes. Based on the schematic, couldn't this be replaced with a single 100uF capacitor? My understanding is that you often want multiple smaller caps (versus fewer big caps) to lower ESR (equivalent series resistance). If that's the reason for using two caps here, we need to see the specs of that 50-50uF cap to see if there is a single cap that replace it. For example, Rubycon 500LXW100MEFR18X45: that's a 100uF/500v cap, 12k hours life rating, 105C temp rating - should be fairly robust. Or, CDE 381LX121M600A032: 120uF/600v, 3k hours, 105C rated. That Rubycon line is available in smaller capacitance values too, so for e.g. C21 or to keep using 2x caps for C20, it could probably be done with a little mounting creativity.

It's unlikely I'll actually change those caps, it's more just for discussion. I think the simpler fan mod will go a long way to improving component life. But I feel that a big part of the Hiwatt idiom is their overbuilt nature.

These are JJ TC 529 50+50

The specs are here:
https://www.jj-electronic.com/images/st ... EN_web.pdf

If you are concerned, wrap a little reflective aluminum tape around the side facing the tubes.

I have yet not heard of a heat related issue
in combos (or heads for that matter)

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

Sounds like heat isn't an issue, but adding reflective tape, as Stephen said, wouldn't hurt.
Hiwatts had an aluminum heatshield between the output tubes and the filter capacitors. Like on this DR504 clone with EL34s that I built several years ago.

I like the isea of the heat shield for the bigger tubes.

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

Regarding the heat shield: after the amp has been running for some time, won't the shield itself ultimately be just as hot as the tubes? And in turn, won't the shield radiate the same heat to the caps?

With the Triwatt specifically, at least in stock form, there really isn't room for a heat shield. Here is a bird's eye view:



(This picture was taken after I replaced the stock JJ 50+50 caps with pairs of Nichicon UCY2H560MHD.)

With the stock caps, I could barely get my pinky in between the bigger KT66 tubes and the capacitor. And you can see the mounting for the tube clamps, as well as the screw for the capacitor clamp, prohibit any kind of shield from being mounted. Of course the cap clamp could be rotated 180 degrees, and if I was willing to drill the chassis, I could also rotate the tube clamp 180 degrees. But I really don't want to drill the chassis!

The JJs are rated for 1k hours at 70C (158F). The Nichicons are rated for 12k hours at 105C (221F). With KT66 tubes, the surface temp of the PSU cap does exceed 70C, as measured by an IR thermometer. I wrapped the stock cap in aluminum tape, and took measurements - that didn't seem to help, and in fact caused wide variations in my temp measurements. I think this is a limitation of using an IR thermometer.

At any rate, the stock JJ caps are rated for 1k hours at 70C (158F), whereas the Nichicons are rated for 12k hours at 105C (221F). The mounting probably isn't optimal, but hopefully good enough.

You can also see in the picture, the length of red and black twisted wire coming out from the chassis, next to the power transformer. This is wired to the PT 6v3 secondary (i.e. for the tube filaments). This will be used to power a fan. I've been trying to think of a good way to mount the fan. What I'm thinking now is the following: four total lengths of bailing wire, one for each of the fan's four mounting holes. Two of those wires will attach to the output transformer, and the other two will attach to the power transformer. The result is that the fan will be "floating" just above the PSU caps, blowing towards the power tubes. I'm also planning to replace the back piece of cabinet with a grill; the fan plus grill should significantly improve ventilation and exhaust some heat. I feel this is good not only for the caps, but the power tubes as well (all components should benefit from lower temps).

I was initially turned off by the idea of a fan... But, I've been building my own computers for decades, so I have a drawer full of 12v fans. You can buy ultra-high quality 12v fans for about $20, with fluid-dynamic bearings with a crazy MTBF of like 100k hours. The one I plan to use is a 120mm fan, 1200 RPM at 12v (audible but very quiet), and completely silent at 9v or lower.