Univox U235PB B+ Secondary Fusing

[astronomicum]

I am adding conditioning, safety, and protection mods on a Univox U235PB including:

•   Plug with ground
•   Y1 rated cap on the PT input- neutral to ground (line noise filtering)
•   2A 220 ohm Current Limiter on the PT input (to limit inrush current) (added benefit of dropping a few volts bringing the input closer to the 115V spec)
•   220K across the Standby (to partially charge the B+ Caps)
•   220K across the Caps (to discharge the B+ Caps after power down)
•   270 ohm 5W across the OT output (replacing the original 20 ohm 5W Resistor on the Speaker Out shunt to ground)

I would also like to fuse the B+ secondaries. My question is, based on the uncommon nature of the U235PB voltage doubler (I call it a Voltage Adder), do I need to fuse the 0V tap in addition to the 250V and 90V taps that have the rectifiers (see attached schematic). Thanks.

[PRR]

> uncommon nature of the U235PB voltage doubler

I don't get it. At all.

[astronomicum]

I have not had a great deal of experience and maybe "uncommon" is not be the best choice of words, but I thought it is more common to see a voltage doubler using two taps and the result being a doubling of the voltage. This one uses three and it does not actually double the voltage. My question then is if I want to fuse the secondaries, do I need to fuse all three taps?

[pdf64]

The scenario of the anode supply fusing (perhaps spuriously) whilst the screen grid supply continues should be avoided.

[PRR]

I don't think it is any sort of voltage doubler. My best interpretation is two half-wave rectifiers. Which on an amp this size, is commercially stupid.

If this is truly an isolated 3-terminal winding, simple network theory says only n-1 fuses are needed.

But listen to pdf64. Breaking plate power and leaving screens hot is liable to complete melt-down.

On one hand, the thing is clearly still alive after 50 years so I do not see any reason to rush into fuses.

The other hand is to admit this is a 21st Century, break out a +520V DC through a full wave bridge rectifier, then MOSFET-drop a screen rail.

[pdf64]

I don't think it is any sort of voltage doubler. My best interpretation is two half-wave rectifiers...

Yes, my thinking is that the 250V winding gets one half wave rectified. That charges the cap up to 350VDC unloaded. Then the other half wave gets tapped off at 90VAC and rectified to 127VDC across its reservoir cap. That 127V is referenced to the 350V cap, adding to a total of 477V.
So, the same method of operation as a half wave voltage doubler.

Which would be fine for a lightly loaded supply (eg bias) but will probably be rather saggy and ripply with an AB amp, especially at high power outputs. 

[astronomicum]

U235PB voltage doubler (I call it a Voltage Adder)

That 127V is referenced to the 350V cap, adding to a total of 477V.
So, the same method of operation as a half wave voltage doubler.

Unless there is another name for this configuration, maybe "Half Wave Voltage Adder" would be appropriate for future reference.

But listen to pdf64. Breaking plate power and leaving screens hot is liable to complete melt-down.

Understood

break out a +520V DC through a full wave bridge rectifier, then MOSFET-drop a screen rail.

I like the idea of breaking out +520V DC and would like to learn how to do it with this PT. Can you schetch it for me?

[pdf64]

Relying on the switching contacts of a jack for ensuring that a safety resistor for the OT secondary kicks in seems like the wrong way to go. Just providing a reasonable degree of resistance to make sure that secondary current can always flow, come what may,  is all that’s necessary.
eg 220ohm hardwired across the winding, no switches, impedance selector etc.
Unless you’re intending on using the safety resistor as a dummy load, eg for silent amp overdrive?

[astronomicum]

I see the advantage in the case of a break or a blown speaker while the jack is still plugged in. I have some 270R 5W resistors. I will use that as you suggest. Thanks.

[PRR]

... as a half wave voltage doubler....

This is a linguistic, not technical, quibble.

I understand "doubler" as the same source, twice. "2" is baked into it.

Here we have two different taps and will get arbitrary ratios. "Adder" covers that but also a lot of other things.

I've never seen this plan. _I_ don't think it needs a name.

It has run 50years this way. I would not mess with it. Fuse the primary and play it with respect.

[astronomicum]

I understand. Thanks all for the feedback.

[trobbins]

As indicated, you shouldn't fuse the screen supply unless that also stops the plate supply. 

You could fuse the 90V diode feed, as that only supplies the extra plate supply.  The difficulty would be to identify the max continuous diode rms current, as the fuse is only of practical benefit if it just exceeds that (if using an IEC type fuse, not a UL type).  You could use PSUD2 with an estimate of the max likely plate supply current (eg. 120mA for 6L6GC PP), which would indicate a diode current of 250-300mArms, so you could aim for a 315mA T IEC type fuse (ie. 5x20mm, 250VAC).

Given the age of the amp, it could be worthwhile confirming that the diode reverse leakage currents are ok, and the main e-cap leakage currents and capacitances are ok - however that requires specialist testing for leakage currents, so some would rightly argue that replacing them could have a protection benefit