Hiwatt DR103 Copy Bias Circuit and HT Fuse Questions

[ThreeChordWonder]

So the schematic below is the bias circuit for the Mojotone DR103 copy.

As you can see, the bias is set by physically swapping out resistors, 4.7k to 6.3K, fed off the 4007 diode, to achieve the plate current. This is how the original Hiwatts were wired, I believe.

I've modified mine by fitting a 5k pot, wired as a variable resistor not a potential divider in series with a 1k resistor (which will probably get swapped for a slightly higher one later to give the sort of sweep I want).

The question is, should I instead leave, say a 5.6k fixed resistor where the circuit shows the 4.7k to 6.3k, and use, say, a 500k or 1meg pot in place of the 470k to ground?

What, if any, would be the advantage?

The other change I've made is to move the HT fuse from the negative side of the bridge rectifier to the positive. A fused ground potentially leaves the entire DC circuit hot, and a pudgy finger in the wrong place could create a new circuit to ground. Fusing the positive means a blown fuse kills the DC positive, which I think is far safer.

The second (set of) questions are (a) am I right? (b) do I still need an HT fuse in the negative / ground line? There are no center taps on this PT except on the 6VAC tube heater windings.

Thanks.

 

[Pete Farrington]

I've modified mine by fitting a 5k pot, wired as a variable resistor not a potential divider in series with a 1k resistor (which will probably get swapped for a slightly higher one later to give the sort of sweep I want).

The question is, should I instead leave, say a 5.6k fixed resistor where the circuit shows the 4.7k to 6.3k, and use, say, a 500k or 1meg pot in place of the 470k to ground?

What, if any, would be the advantage?

Your mod puts the charging current for the 100uF reservoir cap through the 5k bias pot. Consider that at power up, the uncharged cap initially presents a dead short to the pot. The wiper contact of which won't be intended for that kinda service. If the pot fails, resistance creeps up, the amp will run hotter and hotter.
It's much better engineering to make the 47k load resistance variable; it's not havjng to pass the cap charging current so under less stress, and if the resistance creeps up or opens somehow (most likely failure modes), the bias supply still works, output valves just run a bit colder.
ie a fail safe arrangement, good engineering
So try replacing the 47k with eg a 22k pot wired as variable resistor, in series with 33k.
Tweak values as required

The other change I've made is to move the HT fuse from the negative side of the bridge rectifier to the positive

The Mojo design is just bad engineering; fuses, switches etc are generally designed to break AC rather than DC circuits. So wherever feasible, move the HT fuse to the feed from the HT winding to one of the AC inputs to the rectifier.

 

[ThreeChordWonder]

Your mod puts the charging current for the 100uF reservoir cap through the 5k bias pot. Consider that at power up, the uncharged cap initially presents a dead short to the pot. The wiper contact of which won't be intended for that kinda service. If the pot fails, resistance creeps up, the amp will run hotter and hotter.
It's much better engineering to make the 47k load resistance variable; it's not havjng to pass the cap charging current so under less stress, and if the resistance creeps up or opens somehow (most likely failure modes), the bias supply still works, output valves just run a bit colder.
ie a fail safe arrangement, good engineering
So try replacing the 47k with eg a 22k pot wired as variable resistor, in series with 33k.
Tweak values as required

The Mojo design is just bad engineering; fuses, switches etc are generally designed to break AC rather than DC circuits. So wherever feasible, move the HT fuse to the feed from the HT winding to one of the AC inputs to the rectifier.

Good advice on both points, thank you.