Question about HT fuse on PT sec center tap, JTM45 and Hiwatt DR504

[jamesmafyew]

I've been looking at the power supply schematics for the JTM45 and Hiwatt DR504, particularly at the fusing arrangement for the HT secondary.

Though the JTM45 uses a GZ34 and the Hiwatt uses diodes, they seem fairly similar to my eyes in nominal voltages, power amp configuration, biasing arrangement, and output power. In particular, both amps have the HT fuse in series with the grounded center-tap of the HT secondary winding, and both amps pull the bias voltage off of one side of the HT secondary, through diode/voltage divider/filter cap arrangement, and then off to the bias feed resistors.

I've read the Valve Wizard stuff on fusing and rectifiers, which was helpful.

So I have a couple of questions:
1. What happens a.) on the bias supply and b.) on the HT line when the CT HT fuse blows? It seems like the bias voltage would go to twice its normal value given an ideal transformer. What really happens here? What's the voltage on the HT line if the fuse blows and the HT winding is floating?

(I'm confused about whether the bias supply could become a ground reference, and whether this could endanger components in either the bias circuit or on the HT. Is there any reason that there's a series resistor BEFORE the diode in the bias supply? Would having the diode before the resistors/voltage divider make any difference?)

2. Given the similarities in the two circuits, why is the Hiwatt's HT fuse a 1A and the JTM45's only 500mA? The Hiwatt can't be pulling twice as much current for only 5 additional watts. Is it because the Hiwatt's diode rectifier doesn't limit current like the GZ34, and a higher value fuse avoids nuisance fuse-blows?

JTM45 schem:
http://www.drtube.com/schematics/marshall/jtm45tr.gif

DR504 schem:
http://www.drtube.com/schematics/hiwatt/hwpsu2.gif
http://www.drtube.com/schematics/hiwatt/hwpwr50w.gif

[Steve Conner]

I think if the fuse blows, the bias tap becomes the new center tap, and the bias voltage could theoretically go as far negative as the HT voltage was positive, destroying the bias circuit. You can argue that the HT winding might be floating, but I argue that if the fuse has just blown, it's probably because the HT is shorted to ground.

It always struck me as a really stupid place to put the fuse, and I have never put it there in any of my own circuits. It doesn't provide any protection against a shorted diode, any more than putting it in the positive rail after the diodes would.

I don't know why the fuses are different ratings, but if the designer was unfortunate enough to put it in the center tap without considering the bias circuit, I guess he is capable of choosing the wrong rating too. It's usually 500mA for a 50 watt amp and 1A for a 100 watt one, as they draw around 300 and 600mA flat out respectively.

[jamesmafyew]

Yeah that's what I thought, but when I ran some sims in LTspice, the bias never went to more than twice its normal voltage (I could be missing something major though). The thing is though that these amps don't have a separate tap for bias on the HT secondary (which I've seen in other amps and which struck me as a very bad idea), they just pull it off the AC side of the rectifier and knock it down with a voltage divider.

[Steve Conner]

Oh, OK, I missed that you were talking about the kind of bias supply that works without a bias tap. What you saw in LTSpice is correct in that case. The voltage that the bias is derived from is pretty high already, so it can never more than double.

I actually like the above circuit less than the one with a separate bias tap, which is perfectly good as long as you don't fuse the center tap. It has some embarrassing pitfalls. In the bridge rectifier version, it's easy to wire the standby switch so that the bias supply doesn't work in standby, which is great for nuisance blowing of HT fuses.

[jamesmafyew]

In the bridge rectifier version, it's easy to wire the standby switch so that the bias supply doesn't work in standby, which is great for nuisance blowing of HT fuses.

By that do you mean putting the standby switch on the AC side of the rectifier?

[Steve Conner]

It's not as simple as that, there are a few places you can put it that will inadvertently open the circuit for the bias voltage. I struggle to visualize what they are, and there are a few commercial designs that suffer from the lack of bias in standby too.

[jamesmafyew]

Cool, thanks. I did see on Valve Wizard a warning about putting the standby switch on the HT center tap, which I can see would also interrupt the bias supply in this configuration.

About the fuse values, I've compared these:

JTM45   | .5A CT                | 2x KT66 | GZ34  
JTM100  | 1A CT                 | 4x KT66 | diodes + CT
JMP50   | 2x .5A HT (1A total)  | 2x EL34 | diodes + CT
JMP100  | 2x 1A HT (2A)         | 4x EL34 | bridge rect
DR504   | 1A CT                 | 2x EL34 | diodes + CT

For the JMPs, the fuses are on the outer taps of the HT secondary winding (Valve Wizard says this is optimal). Using two fuses this way on a grounded CT gives you an effective rating of the sum of the two, correct? So 2x .5A in the JMP50 is the same as the 1x 1A on the CT in the Hiwatt?

If so, that makes me think that the higher HT fuse ratings are due to tube type, instead of rectifier type. The EL34 designs seem to have twice the fuse ratings for a given power.

JMP50 power amp (diode rect, 2x EL34)
JTM100 (diode rect, 4x KT66)
JMP100 power amp (diode bridge, 4x EL34)