Heater supply - real compared to artificial centre tap

[pdf64]

I used to think that an artificial CT (2 x 100 ohm resistors to give the heater circuit an balanced ground reference) was preferable to a real centre tap on the heater winding of the power transformer. Reason being that if there was a fault in a (power) tube, such that the plate or screen elements somehow contacted the heater filament, then the current from the B+ circuit would blow the 100 ohm resistors as it made it's way to ground.
That would save the B+ winding of the power transformer from driving into a short for too long (and maybe also the primary winding of the output transformer).
But is it really a good thing?
My thinking now is, that will then pull the heater circuit up to the B+ voltage level (because it's lost it's ground reference), which could well damage the heater/cathode insulation of every tube in the amp.
By that rationale, a real heater CT will protect the tubes from this hazard, and so is preferable than an artificial one. Because a real CT will cope with the B+ fault current (thereby keeping the heater circuit tied with a low impedance path to ground) and the B+ or line fuse will eventually blow.
Is my rationale good?
Has anyone worked on an amp with the 100 ohm resistors burnt out, and co-incidentally all the tubes in there also had a hum problem or something?
Apologies if this is general knowledge that I've just discovered.

[R.G.]

Rather than dinking with trying to make the incoming primary fuse blow, it makes a lot more sense to put a fuse in each **secondary** winding to protect that winding.

Primary fuses are not there to protect the power transformer. They may do that coincidentally in some cases, but they're actually there to prevent a failure in the amp, including inside the power transformer, from starting a fire.

Fuse the secondaries, don't sweat whether the center tap is artificial or not.

[mikepukmel]

Rather than dinking with trying to make the incoming primary fuse blow, it makes a lot more sense to put a fuse in each **secondary** winding to protect that winding.

Primary fuses are not there to protect the power transformer. They may do that coincidentally in some cases, but they're actually there to prevent a failure in the amp, including inside the power transformer, from starting a fire.

Fuse the secondaries, don't sweat whether the center tap is artificial or not.

This is a really old post and maybe no one is watching any longer. In case someone is watching: how would I go about calculating what size fuses to put on each sec PT winding? I made some really rough guesses based on old tube data sheets, but those are based on the designs at the time, and also, design limits.

[cjenrick]

copy somebody else. what tubes are you using? Marshall uses B+ fuses in a lot of their new models, you could check out an amp with the same type of tubes,

[pdf64]

Yes, if you can find something similar that's got fusing then use that as a starting point, trial and error after that.
For the HT, if FWB the current on the ac side of the rectifier will be 1.414 x the dc side, whereas with bi-phase full wave, ac current = dc current.
The nature of the loads means that T type fuses are generally best suited.
Don't fuse a bias winding or the heater CT, as applicable.
I guess I didn't word my query well, as RG's response seems at something of a tangent to the issue I was aiming at. In another thread Mick Bailey confirmed that having the heater circuit pulled up to HT voltage could be damaging to the tubes.
Don't worry that info sources may be old, as tubes are an obsolete technology.

[cjenrick]

i see a lot of 500 ma fuses in the B+ lines for four EL34, so figure maybe 250 ma in each line for a pair of tubes,

[R.G.]

First you have to know what DC current goes out of the first filter cap when the amp is at full-max-pedal-to-the-metal-hyperdrive - or at least whatever the highest level/power output that you'll ever want to run it at. The only accurate way to figure that out is to run it at hyperdrive and measure. You can get a decent estimate by computation if you have to, but clamping a true-RMS current clamp on the two sides of the transformer leads would tell you a more real-world value.

Once you have a good estimate of the current in each outside transformer lead (that is, not the CT), fuse each one at about 150% of the reading with a time-delay or slow-blow fuse. Use two fuses, one on each lead for a typical full wave center tapped power secondary.

You could also use hunt-and-peck. Put in the fuse holders, put in fuses that are probably too small, then run the amp up to max and see if they pop. If they do, use the next fuse size bigger. Eventually, you'll get to where the fuses don't pop on power up surges and hyperdrive. Then quit increasing fuse ratings.

[trobbins]

PSUD2 (soon to be PSUD3) allows fuse design to be a bit more rigorous nowadays.

http://dalmura.com.au/projects/Valve%20amp%20fusing.pdf

[nickb]

Yes, if you can find something similar that's got fusing then use that as a starting point, trial and error after that.
For the HT, if FWB the current on the ac side of the rectifier will be 1.414 x the dc side, whereas with bi-phase full wave, ac current = dc current.

I don't think that is correct for a typical capacitor filter type of power supply. The rms current in the transformer will be about 2 x the DC output for a full wave bridge rectifier and 1.4 for bi-phase, other factors being the equivalent. The actual figures depend on the capacitance and the total series resistance. I mention this as the thread has crept from AC heaters to DC HT supplies- quite a leap.