• 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.

Powerdrive

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The negative 400 volt supply is there to provide the source for the grid voltage on the output tube. I used a - 400 volt supply because it was easy to make in the 845 amplifier.

The positive 150 volt supply sources the current for the tubes grid in A2. These power supplies MUST be capable of swinging the entire range of possible grid voltages seen by the output tube (bias AND signal voltage).

The easiest way to figure these out is to look at the plate curves for the tube used. The negative voltage supply must be enough to completely cut the tube off for the plate voltage used. In my case I need -300 volts to cut off an 845 with 1200 volts on the plate. A 211 can be completely cutoff even with 1250 volts on the plate with -125 volts. I would use a -150 volt supply for headroom. If you are running a lower plate supply voltage you may get by with a less negative voltage.

The positive supply must be high enough to to drive the tube deep into saturation. +50 volts should be plenty for most tubes (except the 811A and the 833A). I used +150 volts because it is easy to make from US power (a bridge rectifier on an isolation transformer).
 
in the powerdrive schematic, the -400 supply is used to bias the 813?......

Actually the tube is an 845, I have never built an 813 amp.

The control grid of most tubes must be "biased". This means that a negative (with respect to the cathode) voltage must be applied to the grid to control the current through the tube. If the negative voltage is not there the tube will run "full throttle" and usually self destruct.

Think of this as the gas pedal in your car, except that the spring holds the pedal to the floor and you must pull it up to slow down. The negative voltage pulls it up. If the negative voltage supply is not there the 211 will have a short unhappy life (car floored all the time). It is also likely that the tube failure could blow up a lot of other parts (car crash).

If your knowledge of vacuum tube circuits limited, and this is your first amplifier, please don't start with a 211 amplifier. The voltages used in a 211 amp are instantly lethal in an accident. Amps of this nature are hard to construct, expensive, and the chances of success are slim for an inexperienced builder.
 
Hi,
This is my first post in the forum, also a newbie in electronics (both tube & SS). please kick me if I ask too many rubbish question.

I see that you RC couple the input of the source follower and DC couple the cource to the 845, wouldn't it be easier to also RC couple the source to the 845?

The reason is that I had build a 45 which is very much similar to your simple 45, B+ is ~320Vexcept that it is RC coupled and cathode LED bias and 75Hchoke plate for the 6c45p at 150Vp, -2.2V bias and zener cathode bias for the 45. I am thinking of trying out a source follower betwee the two tubes to see if it makes any difference in fidelity.

I only have IRF820 in hand. I am thinking of doing some thing like that, I will bias the Mosfet with a 5V zener from B+ (with a series 330K resistor)to the gate, then I will pop the drain of the source to the B+ and connect the source to the ground and souce connect to 45's grid via a coupling cap of 0.1uf and replace the grid resistor of 45 to 20K.

Please kick me if it's complete rubbish,

Cheers
 
This is my amp currently.
For the main 1100 V to the OPT, I use a c/tapped, mercury vapor PSU, CLC, with NOS PBC paper capacitors and a NOS PCB oil choke. Coupling caps are Russian teflons.

So, I want more power.

How much % more power will I get if I drive the 211s into A2?

I was thinking of using a 45 driver and the powerdrive circuit, or a 5881 pentode dc coupled to the grid of the 211...
 

Attachments

  • amp2.jpg
    amp2.jpg
    45.1 KB · Views: 1,023
Audio_idiot said:
Hi,
This is my first post in the forum, also a newbie in electronics (both tube & SS). please kick me if I ask too many rubbish question.

I see that you RC couple the input of the source follower and DC couple the cource to the 845, wouldn't it be easier to also RC couple the source to the 845?

Yes, it would be easier, but I really, really, really hate that. If you have a well designed amplifier, the first stage that should go into overdrive is the last stage. Now, when overdriven, the excess positive signal voltage on the grid turns on the parasitic diode formed by the cathode and control grid. This clips off the positive going peaks, and leaves a net negative DC voltage. This is how Class C RF amps derive all or most of their operating bias. This excess DC voltage on the grid moves the final(s) to a more nonlinear operating point that distorts until that excess DC leaks off the coupling capacitor. DC coupling with a follower of some sort avoids this problem. Even if hit with a fast transient, you don't hear it if it doesn't shift the operating point. You can play louder, and if your drive capability is there, much louder if you can go Class A2. In my 807 amp, I use DC coupled 6SN7s as cathode follower drivers, and that amp starts to distort when the OPTs begin to saturate. Driving the grids of the 807s positive doesn't result in a "dirty" sound.


The reason is that I had build a 45 which is very much similar to your simple 45, B+ is ~320Vexcept that it is RC coupled and cathode LED bias and 75Hchoke plate for the 6c45p at 150Vp, -2.2V bias and zener cathode bias for the 45. I am thinking of trying out a source follower betwee the two tubes to see if it makes any difference in fidelity.


It just might.


I only have IRF820 in hand. I am thinking of doing some thing like that, I will bias the Mosfet with a 5V zener from B+ (with a series 330K resistor)to the gate, then I will pop the drain of the source to the B+ and connect the source to the ground and souce connect to 45's grid via a coupling cap of 0.1uf and replace the grid resistor of 45 to 20K.


With MOSFETs, the Zener between the gate and source isn't for biasing. It's there to protect the gate from excessive voltages. You'll need an extra bias pot for setting the MOSFET's Q-Point current.
 
I am thinking of trying out a source follower betwee the two tubes to see if it makes any difference in fidelity.

It made a difference in overload recovery (now non-existent) and transients, but the 45 really doesn't respond well to A2. The tubes that I have (about 20 different old ones) saturate before the grid goes positive. I use the same PC board for 300B amps, which benefit from a little A2.

I see that you RC couple the input of the source follower and DC couple the cource to the 845, wouldn't it be easier to also RC couple the source to the 845?

Yes it might be easier, but you would lose all of the benefits that PowerDrive was trying to accomplish. RC coupling into the grid of any tube that can be overdriven is one of the major sources of distortion in a tube amplifier. This was documented by Norman Crowhurst over 50 years ago. (I can't find a link to his work)

I only have IRF820 in hand.

I tried a lot of fets. You need one with a low "transfer capacitance" the IRF820 will work, but others will work better.

So, I want more power. How much % more power will I get if I drive the 211s into A2?

I got 28 watts in A1 and 40 watts in A2 SE with 1100 volts. You will get double this power in class A P-P. If you really want power a pair of 211's in AB2 should set your speakers on fire!
 
Tubelab,

I think your website and photos are amazing!
Thank you very much for the inspiration you have given me from your web-site, and thanks for your help.

My amp just rocks so bad ***.

I think Russian teflons dont smear the sound as much as the G.E. Pyranol pio's that i have tried... Although the pio's had a very very nice organic sound to them. perfect for psu's.
 
No, I didn't think to try this. I made measurements on the completed amp early in the design phase. They are on my web page.

I "should" have next week off from work and some "extreme vacuum tube experiments" are planned. Spectral measurements may (or may not) be possible depending on the computer situation. The old computer is flakey (on a good day) and its 96KHz sound card won't play in the new whiz bang super computer (which is finally working after 2 months). I will get a new sound card as soon as I figure out which one I want. I have a USB sound module that I can use but it is only good to 20KHz.

Another problem with PowerDrive: The ROHS hysteria has forced the discontinuance of my favorite HV mosfet the 2SK2700 (it wasn't unleaded). I will be testing some replacement mosfets as soon as they all show up, and I have the measurement capabilities.
 
The IRFBC20 looks like it might be a good replacement, I seem to recall it being mentioned before.

The IRF820A looks good to me (International Rectifier version only, strangely the Fairchild spec sheet shows considerably worse Crss).

The IRF840A looks like it might perform alright, with a lower Rds(on) than the others, and apparently passible capacitance?

Perhaps the IRF5801 within it's obvious power and voltage limitations, maybe a Minidrive for triode strapped 6AQ5s or somesuch... :D

I was wondering if you ever tried replacing the grid resistor on the power tube with a CCS. If it worked, seems like you'd dramatically reduce the filtration requirements for that supply.
 
Wow, good call on the DN2540! I was going to use those cascoded for the plate CCS on the driver tube, but for some strange reason I never thought to use one as the grid driver, too. And I could use one as the CCS on the grid I mentioned above. ;) Handy device, luckily I just ordered a decent quantity of those...
 
I did try the 6Y6 as a CF. It works better than most tubes that I tried. The max plate voltage rating of 200 volts limits its use in some cases. An 845 needs over 300 volts peak to peak. I did not try the 6W6 or any tubes in parallel.

Since the PowerDrive development I have discovered that some sweep tubes make good CF's with high peak current capabilities. You also have to watch the H-K breakdown rating on any tube. Since the cathode could be swinging over 300 volts, the filament should be on its own winding referenced to the cathode.

Each design has its merits, and the PowerDrive is simple and effective. I plan further experiments, but I don't have the time in the near future.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.