10A fuse in the filament wiring? Geloso 1020

[modman]

Many of the later 60s Geloso PA heads (in particular the G1/10xx series) suddenly have an 10A fuse in the filament wiring. Like in the 1020:

Screen Shot 2019-09-15 at 21.30.59.png

If you look at earlier version of this amp (it's a 15W pushpull EL84 amp) like the Geloso 3215 or 215AN, this fuse is absent.

Valvewizard writes however:

Heater Fusing
An argument can be made for not fusing the heater winding, provided it supplies nothing but AC heaters and lamps. You know, simple stuff. There is very little to go wrong with a simple AC heater supply, and the likelihood of a fault creating a permanent short is slim. Also, the inrush to heaters at switch-on is so heavy that it may be difficult to find a fuse that will withstand it. I've never seen a 20mm fuse rated at more than 10A. Beyond this you might consider automotive fuses. Traditional amp designs don't use heater fuses, and I have never heard of it being a problem.
However, if the heater supply has a rectifier for DC heaters or ancillary circuits then fusing is suddenly and strongly recommended; rectifier diodes nearly always fail short. If this means adding inrush-limiting to achieve a soft start and stop the fuse from blowing every time you switch on, so be it.

Merlin is the number one authority on tube amps, but his prose is somewhat convoluted .
Actually the first line is saying: if the AC line only feeds filaments and lamps, it is best not to put a fuse.

And I believe him. I had two amps with this configuration, on both the fuse tested good but failed in how way to conduct. Still I cannot believe Geloso, after building this amp for many years without the fuse, suddenly added it for no reason. Geloso enthousiasts I contacted told me: if they put it there, it should be there, but still they failed to explain why.

I removed them in my amps... but maybe somebody knows why they were there...
thanks
J

[pdf64]

It’s probably good practice to fuse all PT windings (other perhaps than a dedicated bias winding/tap).
Guitar amp PT heater windings tend to be a somewhat smaller fraction of the total PT load than the HT, so a fault current load on the heater may not draw sufficient current to blow a primary fuse. Hence a fault mode n the heater circuit of an unfused heater winding may result in the winding overheating and being damaged without the primary fuse blowing.
I think that RG Keen has long suggested the use of automotive blade type fuses for the heater winding.

[roberto]

I would not fuse the bias winding, because it could be worst than without fuse.

[modman]

Guitar amp PT heater windings tend to be a somewhat smaller fraction of the total PT load than the HT, so a fault current load on the heater may not draw sufficient current to blow a primary fuse.

.... and this amp the AC heater wiring only serves the EL84s and the pilot light, the 12AX7 filaments are run on negative DC from a separate transformer tap. This must be the reason. Thanks so much!

Better put the fuses back, if I find new ones 10A...

[R.G.]

Merlin certainly puts a lot of stuff on tube amps on the net, all right.

One important thing to think about with fuses is what you're trying to protect. The AC line primary fuse is not there to protect the transformer or the amp. It's there to prevent a failure from burning down the building. Any protection of the amp or the parts in it is completely accidental. Primary fuses or some other forms of current limiting are effectively required by safety standards.

To protect your power transformer - which is, after all, the single most expensive part in your amp, excepting maybe the output transformer, you might consider a fuse per independent winding section. The current rating of each winding section is perhaps unique, and a fuse may be a help in protecting that winding section from burning out and losing your transformer.

The fuse ratings will probably be different per winding. The obvious illustration is the high voltage B+ winding and the filaments. Filaments need amps, B+ may need a partial amp. The same fuse value can't protect each one. I ran into a self-proclaimed tube amp expert who simply refused to believe that a filament fault would not always blow the primary fuse. He argued for days, carefully avoiding the math that showed he was wrong, and finally simply vanished off the net rather than believe the math.

I suspect the manufacturer simply decided that protecting their power transformers from filament shorts was worth the cost of a fuse holder and wiring.

[tubeswell]

At startup, when the filaments are cold, they will draw more than nominal operating current. So using a 10A fuse on the heater winding is not so daft.

[trobbins]

The turn-on current of the heaters may be a bit different depending on the valves used. I recently tested some 5U4 and they had a 5x cold surge level. The hassle with fusing the heater winding is that the cold surge level takes some time to drop - typically longer than a power supply capacitor charging - so even if a slow blow fuse variety is available, it's current rating still has to be a few times the hot/idle current rating in order to survive.

The logical link with the winding powering the 12AX7's may be that they started using a different transformer, and determined that the chance of accidental connection of one side of the heater to chassis would damage the new transformer - perhaps also related to the incandescent bulb holder or the wiring.