• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Bogen RP40A transformers....what to build with them?

A bias supply draws very little current. So you need a large resistance to drop the appropriate volts. In this case, however, there's 14mA being drawn by the 1/2 6DJ8, so the resistance values will be much lower. You need to use Ohm's Law to figure what the voltages will be.

In other words, you can't just drop one circuit in place of another, especially if the two circuits are entirely different.

I would use a single stage CRC to make the main negative supply, then an additional RC for each 1/2 6DJ8. Why? Because you want to filter that supply well enough that you know it won't inject any power supply hum into the 6CG7.

IMPORTANT STUFF FOLLOWS:

I need to put this warning out there -- Trial and error with these (relatively) high voltages can be dangerous. Or at the very least, smelly and expensive. This is a public forum, so I just need to state that this stuff really is dangerous. Please be careful. If (God forbid) something bad was to happen, I would not want to have had a part in causing that. I also don't want to be the cause of damaged parts, burnt up resistors, capacitor electrolyte all over the walls, etc. If you have any doubts about all this, please stop now.

Please understand that this is all for fun, but due to the dangers involved, I have to state that I cannot be held liable for any damages resulting from these experiments, etc. There, I said it. (I had to, sorry)

--
 
Last edited:
:)

I've built several dozen amps at about double these voltages. Just never decided to try something different. i.e. not off the net.

Thanks for the warning. My amps always have drain resistors on the PS rails or B+.

I just copied the design on my schematic. What I meant by saying that I scope all my PS is that I sim them in PSDII and build them. Then I scope them under load to check for noise issues and proper voltages and adjust accordingly.

Blair
 
OK,
Another thought is that if dropping a ton of voltage is going to get ugly, why don't I just use the bias tap and get my -80 or 90v and use that?

Because in order to use a tube as a CCS, that tube has to work with a fairly high voltage. That's the way tubes are, right? -90V would only be enough for the 6DJ8 itself. It also needs to work into a load (usually a resistor), which will drop volts too. So we need something like -200V right there.

--
 
I think the first step is to get good at applying Ohm's Law to finding voltages in circuits.

I = V/R

Or you can invert it and it becomes I x R = V
Or V/I = R

So looking at the two 6DJ8 CCS combined...

Ia = 28mA

We have a raw supply of something like -300V.

We want a raw supply of -260V, and we're happy to drop that -40V across resistors in decoupling (filter) networks to get our supply free of hum. So we need to drop -40V with 28mA.

V/I = R

40V/.028A = 1428.57 ohms

What if we used the standard 1500 ohm value?

I x R = V

.028A x 1500 = 42V

Close enough? That depends. But that's how it works.

--
 
Hi Ron,

Sorry, I did not mean to imply that I was planning on using the values there. I just poster the layout for review. Here is one with simulated values from PSDII. I suppose I may consider dropping that first 500 ohm resistor a little to keep the 6CG7 voltage up higher.

I got some Clarostat locking pots for the AC balance, but they are a bit large, so I may skip that addition. I like the concept though of being able to adjust the AC symmetry.

Oh, can the LEDs be the miniature variety? I have several of them already.

Thanks!
0bea3af8.jpg
 
Last edited:
That LED will be passing 14mA, so you just want to make sure your mini-LED is capable of passing that kind of current. If not, a simple resistor can be used to set the bias (cathode bias). I figure a 150R resistor would be very close. If you feel safer that way, you can replace the red LED with a 150R 1/2W resistor.

Some questions...

Are you planning to run both channels from a single plate supply? If so, you might want to re-think that for the first stage 6DJ8's. A common-cathode amplifier is going to need as stiff and low impedance a power supply as possible. Unlike the LTP, it doesn't have all that much PSRR (power supply rejection ratio), especially with the smallish plate resistor values forced on us by the low plate voltage. It's easy enough to make two separate power supply rails for the two 6DJ8 first stages.

Why the super-short time constants for the power supply decoupling? 500 ohms and 22uF is not a whole lot of filtering. I think you will need more like 150uF to 220uF in those places, if you stick with 500 ohm resistors. You want the time constant to be long enough to filter to a -3dB point at 1Hz or thereabouts. This may not be super-critical for the LTP (it rejects power supply noise very well all on its own), but the first stage voltage amplifiers will need as clean a power supply rail as possible.

What did PSUDII predict the ripple would be for those stages? Did you put the current sinks in to show it what the loads will be?

Since you increased the negative voltage of the LTP tail supply to -244V, the bias on the 6DJ8 may need to be changed a little. You see, that 6DJ8 needs to deliver 14mA, no less, no more, because it sets the bias on the 6CG7 LTP. If you bias that hotter, then the 4V at the 6CG7 could become 3.5V or lower, and now we may not get enough swing to the 6L6 plates. It's all interdependent, so you can't mess with it too much. Not a big deal -- just remember that you may need to adjust the cathode resistor in the 6DJ8 CCS (which is now a red LED to give a 1.8V reference there) if you increase the voltage of the tail supply. You may need to adjust the 6DJ8 cathode resistor that so that it draws near enough to 14mA, and not (say) 17mA, which would be too much current in this case.

The bias on the first stage 6DJ8 is not nearly so critical, as it is RC coupled to the next stage, so its parameters are independent of the LTP bias. You merely want to bias that stage to a place where it performs well with its limited B+.
--
 
Last edited:
Hi Ron,

Yes, I added the current sinks at each stage. At the end, where the 6DJ8 plate supply is, the ripple is very low. I can add a little capacitance if necessary. I was planning to have each 6DJ8 on its own plate leg for better isolation.

I think I would rather use the LED. I can get those from radio shack, so no big deal.

Other than a few adjustment possibly, it sounds like it is more less go time. I can take measurements after I get it fired up.

What was the negative supply voltage you selected the values for on the CCS? I can make it that.

Thanks!

Blair
 
Just noticed the Schade feedback. Let's hold off on that. I don't know if that will work with a medium-mu driver tube like a 6CG7. My understanding (from reading the Baby Huey thread) is that you want a tube with a high mu driving the plate feedback to the output stage, and you want the output stage to have high transconductance (gm). Unfortunately, 6L6 is not a high gm output tube. For that you'd look to EL84 or EL34.

We can play with the Schade feedback later.
--
 
What was the negative supply voltage you selected the values for on the CCS? I can make it that.

-235V

It would also be nice to put an LED in the cathode of the 6DJ8 input tube (voltage amp). That should let the 6DJ8 operate with lower distortion (as in fixed bias instead of cathode bias). But of course, if the amp winds up being too sensitive in your system, you can swap that cathode LED to an unbypassed cathode resistor and reduce the gain of that first stage. How much gain you wind up needing there depends on how much NFB you decide you want.

--
 
Last edited:
Cool. I guess you know the drill then. Start up the amp on the variac set at 0V, raising the voltage s-l-o-w-l-y, make sure no smoke or sparks fly. If anything goes wrong, switch off power asap.

If nothing goes wrong, when you're at full 120VAC, check your DC voltages in the amp to make sure all is as expected. Right?
--

Wow, you can access a distortion analyzer? I'm envious
--
 
Can you tell me what currents you calculated you'd get for the 6L6 plates and screen grids?

The voltages on the schematic don't jibe with the voltages I calculated.

I figure about 60mA plate current and about 5mA screen current per 6L6. Is that what you used?
--
 
I would worry about the negative supply for the current source. At power up there will be no voltage there. Current can't flow through the return diodes from ground to the transformer windings as they are in opposition to the negative supply diodes.

Once everything starts conducting, there will be a current path through the current source, through the LTP and back, however I doubt the minus supply will ever function properly.