Full wave bridge for bias supply- does this look right?

[DIY_Guy]

Thread title says it all.... I'm sure the values will need to be tweaked once i get it running to get the voltage range correct. Just need to know if I'm on the right track with this....

[tubeswell]

Thread title says it all.... I'm sure the values will need to be tweaked once i get it running to get the voltage range correct. Just need to know if I'm on the right track with this....

Hi there DIY_Guy

That sure uses lots of caps, diodes and pots. Is there any particular reason you don't just want to do a more conventional and economical HW bias supply (i.e. with one diode, and a couple of resistors and a filter cap)?

(Edit - Oops I just noticed you have 4 bias supplies, sorry. All your bias pots are in parallel - effectively reducing the available bias range quite a bit, and when you tweak one pot, it will affect all the other values, defeating the purpose of having 4 separate pots - will it not? The decoupling caps won't actually change this, unless you alter it to have corresponding parallel top resistors for each pot divider/decoupling cap combination, with each 'top' each resistor taken from the 1st (reservoir) 47uF filter cap)

[Vortexion]

As tubeswell has pointed out, the single diode half-wave approach is usually quite sufficient, because in a grid bias circuit there's so little current being drawn that the amount of ripple generated is very small and easly smoothed. However, AFAIK there's no harm in using a bridge rec if that's what you want to do.

I'd also suggest that having individual filter caps for each bias adjuster is overkill. It'll be fine with just a single 47uF after the 15k resistor, then run all your adjustment preset pots after that.

Basically, you're over-filtering like crazy!

[DIY_Guy]

Hmmm.... I'm a bit confused about it not working properly, since I've seen this same setup before in production amps(minus the bridge rectifier). I think the Randall RG series uses it(only 2 pots of course) and it was recommend in a few places in TUT2....

I guess my brain is seeing it the wrong way. If all the pots are tapping off of that same resistor (R1), shouldn't they all see the same voltage?

Oh and the reason I'm using a full wave bridge is because I just wanted to be different. I've got tons of parts laying around, and I'm all about overkill when it comes to things like this. Won't it give me a bit more hum reduction?

[Vortexion]

Oh and the reason I'm using a full wave bridge is because I just wanted to be different. I've got tons of parts laying around, and I'm all about overkill when it comes to things like this. Won't it give me a bit more hum reduction?

As I said, there's no harm in using a bridge rec - I've done the very same myself - but the low current draw of the bias circuit means it's not actually necessary. Another thing to keep in mind is that any ripple that does occur on the bias supply will appear pretty much equally, in the same phase, on both sides of the push-pull output stage, so most of it will be cancelled out anyway.

[black_labb]

i would go without the caps on the wiper of each pot. these caps will mean that your wipers will take a while to reach the desired voltage and cause lag in the adjustment while the caps charge up. this could be annoying when adjusting bias. could also cause some interaction with the other bias controls when the caps charge up/discharge when adjusting the bias.
i would replace those with a cap on the downstream side of R1 (along with the one on the upstream) to make a PI filter and leave it at that.

keep in mind that the 4 50k pots are in parallel causing a resistance of 12.5k, so not much adjustment (may be plenty). in my bass amp build i have a similar circuit but use 3 250k pots for the equivalent of 83k resistance.

does anyone know any ways to modify this circuit to keep the

[DIY_Guy]

keep in mind that the 4 50k pots are in parallel causing a resistance of 12.5k, so not much adjustment (may be plenty). in my bass amp build i have a similar circuit but use 3 250k pots for the equivalent of 83k resistance.

Ah I see. Can't I raise and lower the 33k resistor to get the range better?

[defaced]

Here's a quad bias supply I'm working on; it does not include the winding, rectifier, or first filter cap because they are mounted off this circuit board. I use it with a full wave bridge for no particular reason. The first pot allows adjustment of the four supplies up and down, while each supply is indepenant. I got my prototype working again, and this does work well. Values are only suggestions and the basic circuit configuration was pulled from a thread where Paul P and Enzo went back and fourth about a dual bias supply.

http://music-electronics-forum.com/a...-quad-bias.png

http://music-electronics-forum.com/t5415/

[Vortexion]

Can't I raise and lower the 33k resistor to get the range better?

That's not a great way to alter the range of adjustment. In order to make the 12.5k of pot track resistance constitute a larger proportion of the voltage divider, you'd have to reduce the value of R2 significantly. This would cause more current to be drawn from the bias winding - not an especially desirable outcome. The better way to achieve it would be to use larger value pots.

However, I must say that defaced's quad bias supply is the best solution of all. It achieves everything you're aiming for, plus it has the advantage that should any pot wiper lose contact with its track, the bias on that valve will default to the maximum negative voltage, avoiding any risk of lost bias and redplating.

[DIY_Guy]

Thanks guys. Once again I learn some things from this forum that otherwise would of had me chasing the issue on a breadboard for hours.

I love this forum.

[defaced]

However, I must say that defaced's quad bias supply is the best solution of all.

Aw, shucks. You guys are too nice. Glad I could help.

[tubeswell]

And the key difference between Mike's scheme and your original one DIY Guy, is those resistors (R2, R4, R8 and R11 in Mike's scheme) that I was talking about earlier. Otherwise your pots and decoupling caps are all just in parallel.

[Steve Conner]

I installed almost this very same circuit shown in the first post, in my Crown SXA restoration. The only difference was that I didn't use an individual cap on each wiper, rather a single one after the 15k resistor, as suggested by Vortexion. I did this because I didn't think it needed four.

I used a bridge rectifier because: The original SXA schematic showed one, it produces 120Hz ripple instead of 60, thus twice as easy to filter, and at the end of the day, diodes aren't that expensive.

The four pots don't interact, because the tube grids (at least in theory) don't draw any DC current. In practice they do interact, because they set the plate current draw of their respective tubes, which affects the B+ voltage of the whole amp, which impacts back on the plate current of all the other tubes. So you have to go through and reset them all two or three times.

A tip I got from one of my old Mech Eng lecturers: When you're working with several adjustments that interact, only turn them half as far as you think they need. This greatly increases the odds of "converging" on the right "solution".

[tubeswell]

Ah yes I see now that the wipers are not tied to one side of the pot, which stops the pots appearing in parallel. But this is riskier if the wiper contacts fail.

[mooreamps]

I used a bridge rectifier because: The original SXA schematic showed one, it produces 120Hz ripple instead of 60, thus twice as easy to filter, and at the end of the day, diodes aren't that expensive.

You may be good at quoting electrical theory, but in this application is not a factor. The current load on a grid bias circuit is so low, filtering is not an issue. Not that I am any real big fan Ken Fisher, but go look at a trainwreck schematic and see what I mean.