What I'm wondering is whether there would be any downsides to having a transformer wound with slightly larger wire. Maybe Audio Note would do it, or maybe I could send the core to Heyboer and have them wind a custom part for me. Seems to me a power transformer is a pretty straightforward wind, and there wouldn't be any special considerations.

If you were told that it's relatively uncommon for this particular transformer to fail, the slightly lower copper losses of larger secondary wire most likely won't save you enough heat gain to make any difference. That is, if it would even fit in the existing winding window.

The trafo probably has something like M19 laminations, which has significant core loss and runs hot, even if moderately loaded. It might be best to send it to Heyboer, and have them build you a similar trafo with M6 lams to ensure long-term reliability.

Thermionic

Sorry for taking so long to respond! I've been so busy lately I haven't spent very much time online.

All other things remaining constant, just going from M19 to M6 will make for a significantly cooler running trafo than using the next gauge larger wire on the high voltage secondary. In many cases, you simply can't fit the correct number of turns of larger diameter wire into the window anyway.

Transformer builders have detailed charts that tell them how much core they need to carry a given volt-amp rating, as well as charts that tell them which size wire to use for the application. Many builders use M19 when possible to save the customer some money. In some cases, however, it may be borderline iffy as to whether or not M19 will work in a certain application, and there's only one way to find out for sure. If it gets too hot when bench-tested under full load, they pull it apart and restack it with M6. Problem solved, provided everything else was designed correctly.

Your Kit 2 has quite a high secondary voltage and draws a good bit of current as well, plus a significant power tube filament current draw. And, it has two separate filament windings. All that adds up to a WHOLE lotta wire, and I'd guess from looking at the size of the power iron that it's probably stuffed to the gills already. Actually, it may already have M6 lams, in which case you're just kinda stuck. You of course can't have a trafo built on a thicker stack, or go up to the next lamination size, because it wouldn't fit your existing mounting holes.

IMHO, your best bet would be to get a Hammond 166N6 filament transformer (under $20) to run your power tube heaters from. The 166N6 is fairly compact and wouldn't eat up a lot of real estate inside the chassis, and would take a big load off your power transformer. A very good example is a set of paralleled SV83 monoblocks I just finished, using the same power transformer I use on some of my EL84 integrated amplifiers. The high voltage secondary and rectifier filament windings see precisely the same current draw in either amplifier. The only difference was that in the monoblocks, I used a regulated DC filament supply operated from a separate transformer, because they were to be used on 104dB efficient speakers and had to be silent at all costs.

The transformer runs fairly hot in the EL84 integrated amplifier, where it also powers the filaments for two EL84s and a 6922. You can put your hand on it and keep it there, but it's quite uncomfortable to do so for more than a few seconds. In the monoblocks, the same power trafo runs nowhere even near hot. It barely even gets warm at all!

This demonstrates the core loss/heat gain thing very clearly, as the DCR of a typical 6.3V @ 3A or 4A secondary might only be a couple tenths of an ohm. In your Kit 2, the I²R heat gain would only be 3.2² x .2 = 2.05 watts of heat. Only 2 watts of DCR heat production is saved by moving the 6550 heaters to another trafo, but a lot of core loss heat production would be saved.

Thermionic

I've noticed a similar phenomenon in loudspeakers when used at very high power levels. A significant amount of heat is generated within the core from magnetic loss, adding to the heat transferred from the coil by radiation. The biggest thing we all watch out for is voice coil heat, and the most popular cooling method is convection using vents. But we often overlook center pole heating caused by radiation and magnetic loss, which sometimes makes the core reach temperatures in excess of 200°F.

Inside a speaker run at high power levels, the core temperature gets hot enough to cook with. It's surrounded by the magnet, forming a sort of thermos bottle that is very effective at holding the heat in. Over time, it bakes the voice coil adhesive and eventually causes it to break down, the coil unwinds and the speaker fails. This is the most common cause of speaker failure, far more more common than voice coil fusing.

This is exactly the same thing you're saying happens inside transformers. Makes sense.

I've been concerned about this amp though, because it regularly blows fuses. Not every day, but maybe once a month. Not enough to say there's a problem for sure, but after a couple times, it's worth looking into. I had time this week to open it up and check everything out, and sure enough, I found a problem that may be the cause of the blown fuses. In fact, in hindsight, I'll bet it's what caused the transformer to fail in the first place. Time will tell.

By the way, I wrote Brian Cherry about this when it happened. He sold me the PS transformer replacement, shipped it quickly, and has provided excellent support. When the fuse blew after that the first time, he told me to check the fuse size, and gave some thoughts about primary and secondary windings, stuff like that. I mean, what can you really do to troubleshoot an amplifier problem over the phone? You really have to get out the scope or meter and check stuff out, without that, it's all just guesswork. But Brian was as helpful as you can be in a situation like this, more willing to help than I think most companies are. Great guy.

So anyway, last weekend I opened the amp and checked some voltages. Found a problem right away - the left channel output tube was biased harder into conduction than the right side, about twice as high. So I retired the amp for a few days until I had time to look into it further. Last night, I had time to fix the amp. A coupling capacitor had become weak and was passing DC, so I swapped it (and all the other coupling caps) with new ones. Now, the bias levels are all correct, so I'll bet the fuse stops blowing every month.

• Paper-in-oil capacitors

• Fused cathode mod for Audio Note Kit 2