(SOLVED) Power supply help, for "preamp only" 3-little tubes

[Geezer]

(See reply #15 below for solution)

I am building the DuMars OD amp (full schematic of my preamp only version here>> http://www.el34world.com/Forum/index.php?topic=14201.msg134491#msg134491 )

What is the best way to adapt the power supply to running only 3x tubes? (2x 12A_7-types & a 5879 pentode). I'll only be using 2x nodes, "D" & "E" as shown on the drawing.

I have attached a drawing of what I might try....is this the best way?

Should I put a bid sand resistor to ground to simulate the load of the PI & the power tubes?

Open to all suggestions.....

TIA, G

[Tiny_Daddy]

I would keep those 20uF filter sections to make sure the ripple is at an acceptable level for the preamp. If there's too much voltage just make one of the resistors larger.

[Geezer]

So just build the original PS & use only the nodes I need, adjusting the dropping resistors as I normally would to adjust the voltage(?)

I have the caps & space to do that.....just want to be sure of your suggestion.....

G

[HotBluePlates]

I would keep those 20uF filter sections to make sure the ripple is at an acceptable level for the preamp. If there's too much voltage just make one of the resistors larger.

I'd do the same. The response of the original preamp, and the low level of ripple, is likely due to the total power supply. So keeping the extra filter caps, even though you won't be using the extra nodes, is probably a wise choice.

[Geezer]

Thanks guys, that makes sense to me, so I'll stay with all of the original PS components.

I've drawn up the layout with the complete PS & will post it over on the original DuMars thread.

G

[LooseChange]

Will you be building this from scratch and providing a "new" PT?
I build my preamps with appropriate PT's (Less voltage than a full amp) and three node power supply filtering (20uf caps). Seems to work the best with no hum.

[sluckey]

Just another option... Parallel nodes worked well in my 'Warbler'. Very easy to set the voltage for each node.
http://home.comcast.net/~seluckey/amps/warbler/warbler_rev1.pdf

[Geezer]

Will you be building this from scratch and providing a "new" PT?
I build my preamps with appropriate PT's (Less voltage than a full amp) and three node power supply filtering (20uf caps). Seems to work the best with no hum.

I'm building it into a gutted Valve Jr head cab & chassis & will be using the existing Valve Junior PT, which is 260-0

[jjasilli]

It appears that you are building a stand-alone preamp to run off a 260VAC PT.  After rectification 260VAC X 1.414 = 368VDC B+, which will not be drawn down by the tiny current demands of the preamp tubes.  Another option is a voltage regulator which will both drop B+ and squelch AC ripple in the power supply.  The latter might be useful for the hi gain circuit.  Just another option.

(I thought I posted this earlier,  but it seems to have vanished.)

EDIT - I did post this in the Dmars thread

[Geezer]

JJ, do you have a drawing or schematic of the voltage regulator circuit you are suggesting?

Geezer

[LooseChange]

What's wrong with 368vdc on the preamp? Its probably optimal on an OD circuit.

[Geezer]

What's wrong with 368vdc on the preamp? Its probably optimal on an OD circuit.

http://www.el34world.com/Forum/index.php?topic=14201.msg134580#msg134580

I'm trying to replicate the voltages on nodes D and E of Tubenit's D'mars amp, found in that thread.

G

[jjasilli]

JJ, do you have a drawing or schematic of the voltage regulator circuit you are suggesting? Geezer

Probably the simplest thing would be the good 'ol VVR.  This might actually be a pretty cool tone shaping device as high voltage might also be desirable as LC says. 

But if you want to go crazy checkout the attachment:  OC3 tubes require no power supply; and substitute SS rectification for the tube rectifier; maybe add another OC3 to make up the difference. 

[PRR]

> big sand resistor ...simulate the ...power tubes?

That's like building a 160MPH sports-car, with no throttle, then dragging the parking-brake to hold 25MPH around the neighborhood. You can do it that way; there may be a better way.

You do need more than 1 or 2 R-C filter stages to feed sensitive preamps.

"Adjust this R?" is a good plan. You can rough-compute the change on sight. We see that the 250r must be passing 72mA, and the C-D R (which feeds the part you'll keep) must be passing 7mA. So 250 times 72/7 is a bit over 2.5K.

Similar scribbles suggest values for the other stages.

This ~~10X more resistance is 10X more filtering per R-C stage. You could use smaller caps, but in 450V caps there's a sweet-spot around 32-40uFd. You can leave-out one R-C stage and merge a couple resistors.

After B you are almost down to just the part you will keep. There's apparently an omitted 1mA load on C but "correcting" the 3.3K for that gives 3.77K which is just too odd.

40uFd pre-A
"4.2K" (*)
20uFd B
3.3K
20uFd C
2.2K
16uFd D
2.2K
10uFd E

(*)This is your fudge-point. I assumed pre-A would be 354V but with 9/10th load on the power transformer it may be closer to 370V. To get 370V-324V= 46V drop at 7mA you may need 6.57K here. The power dissipation could be 46V*0.007A= 0.322Watts. Get some extra 2.2K and 3.3K several-Watt resistors, tack-solder series some up to 8K or 10K, smoke-test. Use clip-leads (power off!) to short one out until you get near 324V at B, then re-build nice and pretty.

[DummyLoad]

redesign for the task.

attached should work.

[Geezer]

UPDATE: I went simple "old school" & found a solution that gave perfect results.

1) Went with an EZ80/6V4 recto, ran as a "hybrid" bridge (w/2x SS "rear" diodes to form the CT, since my PT has a 260-0 HT secondary)

2) Changed the 1st 2x string resistors to 4.7k (between nodes A & B) and 2.2k (between nodes B & C)

This yielded a perfect 280v & 274 volts on nodes D & E

Thanks for all the help & suggestions! Now to get the pots & relays wired up .....

G