Found this link while perusing the internet for 833 designs - it sounds way cool. I have no idea how old this link is, but was wondering what ever became of this idea??
This might be just the trick for driving my classic big Infinity speakers with a single-ended amp...and they said it couldn't be done (at least not without buying a WAVAC for $50K)!
Any news?
The 833A SE Amp Prototype
This might be just the trick for driving my classic big Infinity speakers with a single-ended amp...and they said it couldn't be done (at least not without buying a WAVAC for $50K)!
Any news?
The 833A SE Amp Prototype
The 833A experiments shown on my website are about 5 years old. Transcendar built the original OPT used in those experiments. It did not meet the original design requirements and both of us agreed not to proceed any further. I have an OPT and several possible power transformers, but haven't gone any further. I don't have any need for a mono 200 watt amp, and no place to put it, so no further action is planned. I may build a monster guitar amp someday but I don't need that either.
OK, thanks for the update!
In talking with Jack Elliano of Electraprint, he has said he could build an output transformer with a FR of -1db 20-30hz to 20khz if I could tell him the operating conditions I'd be running at. Right now I'm in the process of figuring that out; if you don't mind my asking, what did you find to be the optimum OP for the 833 in terms of plate voltage, current and bias?
I like the powerdrive idea - I'm a fan of using solid state where appropriate in a tube amp, and have built several preamps using SS CCS plate loads and even a 26 DHT preamp using the mu output from the plate CCS for a lower Z, with no "sand contamination" that I can hear ;-). The tube still does all the voltage amplification, so the sound should be (and is to my ear) the sound of a 26 DHT, not a SS chip.
So, to make a long story short, I'm in the planning stages of an 833 amp project now, and I'm considering a powerdrive-type setup to drive the monster tube. How does that work with you - do you sell the powerdrive boards that can be used with an 833?
In talking with Jack Elliano of Electraprint, he has said he could build an output transformer with a FR of -1db 20-30hz to 20khz if I could tell him the operating conditions I'd be running at. Right now I'm in the process of figuring that out; if you don't mind my asking, what did you find to be the optimum OP for the 833 in terms of plate voltage, current and bias?
I like the powerdrive idea - I'm a fan of using solid state where appropriate in a tube amp, and have built several preamps using SS CCS plate loads and even a 26 DHT preamp using the mu output from the plate CCS for a lower Z, with no "sand contamination" that I can hear ;-). The tube still does all the voltage amplification, so the sound should be (and is to my ear) the sound of a 26 DHT, not a SS chip.
So, to make a long story short, I'm in the planning stages of an 833 amp project now, and I'm considering a powerdrive-type setup to drive the monster tube. How does that work with you - do you sell the powerdrive boards that can be used with an 833?
My favorite OP was 1500 volts into a 5K ohm load with as much current as the tube could stand. I ran up to 275 mA which is over spec. This would work with forced air cooling. For convection cooling I would keep it around 200 mA. I got nearly 200 watts output at 275 mA
The heater draws 10 amps. I could not make it work on AC. Hum came from the speakers before I swithced ON the plate voltage.
The 845SE amp was built about 7 or 8 years ago. It is still umchanged since I built it. I don't use it much any more due to the heat that it makes.
I dont have PC boards for PowerDrive, but you could make your own or use PTP.
Was thinking along these lines and modifying for 845 filaments and upping the input voltage to 650-700, removing the 400v B+ since that will be done with it's own setup through the TSE board. It seems like a reasonably simple circuit and I like the solid state rectification in my TSE.
Sorry to the OP for the hijack.
Last edited:
I recently completed an 833A SET which I use to drive the woofers of my GR Research OB7 speakers. The design is very simple but the execution is tricky. I stacked a pair of 2C22 triodes as a choke loaded mu follower and directly drive a pair of 245s from the lower 2C22s anode which is connected directly to the grid of both the upper 2C22 and the grids of two 245s. One 245 drives the tweeter and the other 245 connects in parallel to both a 100K pot that controls the signal strength to a Hammond 5k number 126B coupling transformer that has its secondary connected between ground and the grid of the 833A, and an output transformer that drives the mid-range speaker. The impedance is not a problem: the output impedance of the lower 2C22 is much less than the two 245s it drives and the 245 that drives the mid-range and the 833A grid has low enough output impedance to drive the coupling transformer to the 833A grid and the mid-range transformer. The 833A grid has more than enough input impedance that this arrangement gives better impedance ratios than I would get if I let my mu follower pair of 2C22s drive three 245 grids.
But I suspect that most people would just use one 245 to drive the 833A and not bi-wire or, in my case, tri-wire the loudspeakers.
Now for the secret of making an 833A amplifier stage work successfully; a well filtered plate voltage is needed. So I made mine with Hammond 724 1500 Volt center tap transformer, pairs of fast reacting diodes (Mouser part number 583-5000F). These diodes are doubled up to double the current carrying capacity. So you will need two pairs for each channel. I connected these to 600 volt rated electrolytic capacitors wired three in series, two strings of 56 microfarads for a total of six per channel, a Hammond 193M ten Henry choke that has to be mounted in plastic because it can only isolate its outer case to 800 Volts, and the same capacitor arrangement as the first bank of the pi filter except I used 100 microfarads. This measures within 25 volts of 1000 Volts for the anode of the 833A. With the grid at zero potential, that is, grounded through the secondary of the coupling transformer, the 833A draws about 140 ma. So far, so simple except for the complication of having to enclose the choke in a plastic cage so that nobody can touch it and die and the necessity of stringing filter capacitors in series. The tricky part is the DC that is the only way you can heat the cathode without 120 Hz noise in the speakers. This requires enormous capacitance. I started with a Hammond 125V10 10 Volt transformer that is rated at 20 Amps and a full bridge rated at 50 Amps. A 20 Amp rated bridge gets hot enough to stink up the house so get a 50 Amp bridge. Mount it on an aluminum chassis with heat sinks and it will only get warm to the touch. Connect this to a small 82,000 microfarad 25 volt rates capacitor, then a 0.1 Ohm wire-wound resistor rated at least 25 Watts then a 20 Amp fuse, then an automotive digital 4.5 FARAD capacitor. This capacitor has to be kept partially charged because you do not want to hit it with the surge it will take if you fire up the thing with the 4.5 emulating a short circuit. So when your amplifier is turned off you maintain the charge with a 9 Volt 1 Amp AC to DC converter from Parts Express through a 100 Ohm 10 Watt resistor. You will also need a relay to allow the 833A heater to disconnect when the thing is switched off.
Finally, I use a Hammond 1642SE output transformer and it all works beautifully, with more than enough low frequency energy.
I hope this will help anyone else crazy enough to build a big one. Visitors to your home will think you have flipped your lid when they see it but the sound of this outclasses any very expensive amplifiers I have ever heard.
But I suspect that most people would just use one 245 to drive the 833A and not bi-wire or, in my case, tri-wire the loudspeakers.
Now for the secret of making an 833A amplifier stage work successfully; a well filtered plate voltage is needed. So I made mine with Hammond 724 1500 Volt center tap transformer, pairs of fast reacting diodes (Mouser part number 583-5000F). These diodes are doubled up to double the current carrying capacity. So you will need two pairs for each channel. I connected these to 600 volt rated electrolytic capacitors wired three in series, two strings of 56 microfarads for a total of six per channel, a Hammond 193M ten Henry choke that has to be mounted in plastic because it can only isolate its outer case to 800 Volts, and the same capacitor arrangement as the first bank of the pi filter except I used 100 microfarads. This measures within 25 volts of 1000 Volts for the anode of the 833A. With the grid at zero potential, that is, grounded through the secondary of the coupling transformer, the 833A draws about 140 ma. So far, so simple except for the complication of having to enclose the choke in a plastic cage so that nobody can touch it and die and the necessity of stringing filter capacitors in series. The tricky part is the DC that is the only way you can heat the cathode without 120 Hz noise in the speakers. This requires enormous capacitance. I started with a Hammond 125V10 10 Volt transformer that is rated at 20 Amps and a full bridge rated at 50 Amps. A 20 Amp rated bridge gets hot enough to stink up the house so get a 50 Amp bridge. Mount it on an aluminum chassis with heat sinks and it will only get warm to the touch. Connect this to a small 82,000 microfarad 25 volt rates capacitor, then a 0.1 Ohm wire-wound resistor rated at least 25 Watts then a 20 Amp fuse, then an automotive digital 4.5 FARAD capacitor. This capacitor has to be kept partially charged because you do not want to hit it with the surge it will take if you fire up the thing with the 4.5 emulating a short circuit. So when your amplifier is turned off you maintain the charge with a 9 Volt 1 Amp AC to DC converter from Parts Express through a 100 Ohm 10 Watt resistor. You will also need a relay to allow the 833A heater to disconnect when the thing is switched off.
Finally, I use a Hammond 1642SE output transformer and it all works beautifully, with more than enough low frequency energy.
I hope this will help anyone else crazy enough to build a big one. Visitors to your home will think you have flipped your lid when they see it but the sound of this outclasses any very expensive amplifiers I have ever heard.
Update om my 833A: Last fall I auditioned Magnepan 0.7 speakers which sounded better than any cone speakers, even speakers that cost more than $50,000. The salesman played a work by Bernstein on a proper but not all that remarkable solid state amplifier. Over a couple of months I made payments until I was ready to take them home. I connected them to my 833A amplifiers and the sound was spectacular. However some improvement was needed. I did not like the potentiometers on the transformers that drove the 833A grids so I replaced them with 40 Henry chokes, some leftover Mundorf blocking capacitors and ran parallel feed Lundhal LL1930 mu metal transformers with the primaries in series and the secondaries also in series for an 8:1 step-down from 4.5K and this made a better match, and better sound. I still could do better. My preamp is a pair of #27 Arcturus blue glass triodes which were transformer coupled by 10K 1:1 transformers for the output and I replaced these with plate loading resistors and .33 micro-Farad Mundorf silver oil capacitors and this made the music come alive. I have lost all interest in bothering to hear any other amp and speaker combination in my home after hearing it.
Because I use 5 channel SACD I needed to make a center channel amplifier so I made one with a GM70. I had originally intended to use an 845 but the vendor sent me a GM70 instead. I changed the cathode heater from 10 Volts to 19.2 Volts and got it working. It sounds almost but not quite as detailed as an 833A channel in an AB comparison with volume matched on 0.7 speakers. The center channel will use a Magnepan MMG-C which is a couple of dB more sensitive .
I am convinced the 833A is the Stradivari of all vacuum tubes. It is a perfect match for Magnepans.
Because I use 5 channel SACD I needed to make a center channel amplifier so I made one with a GM70. I had originally intended to use an 845 but the vendor sent me a GM70 instead. I changed the cathode heater from 10 Volts to 19.2 Volts and got it working. It sounds almost but not quite as detailed as an 833A channel in an AB comparison with volume matched on 0.7 speakers. The center channel will use a Magnepan MMG-C which is a couple of dB more sensitive .
I am convinced the 833A is the Stradivari of all vacuum tubes. It is a perfect match for Magnepans.
Hi Drbarney1,
You may want to check out some other threads in the tubes-valves section.
Lots of info and pictures there.
Even an 'upgrade' from the 833
TB3/1000 is alive ! ("European 833A")
Where are the 833 amps?
The Midlife Crisis - My 833C Amp Build
You may want to check out some other threads in the tubes-valves section.
Lots of info and pictures there.
Even an 'upgrade' from the 833
TB3/1000 is alive ! ("European 833A")
Where are the 833 amps?
The Midlife Crisis - My 833C Amp Build
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.