A pair of 2 foot tall, 175 pound 8974/X-2159
will make 2.015 MegaWatts as an AB1 audio amplifier. 22,500V 125A in the plates, 16V 600A in the filament (directly-heated tetrode). Mostly cooled by water but they want forced air in a few places the water won't reach.
"It was also noted how much more quietly water-cooled transmitters operate than new ones with air blowers......"
> with zero (!!!!!) input power
Same again. However if you look at the input capacitance and grid-swing, it needs a couple hundred VA of unreal power to hit the top of the audio band. In an untuned circuit, this amounts to hundreds of watts of potential real power: forget 12AU7 or even 6550, think a pair of 833.
> WLW
More entertaining than accurate. The 8-foot horn on the mike dates from the era before tubes. (Water-cooled carbon-mikes in series with a spark-gap or RF-alternator.) Film-sound was using automatic limiters long before WLW was pushing a half-megawatt, and it would be insane to expose all that gear to un-limited audio. (Crosley was nuts, but not flippin' insane.)
> "Crosley noted, bitterly, that 50 kW wasn't that many more horsepower than a couple of speeding Buicks."
He was wrong. 50,000 Watts is 66 horsepower. A 1938 Buick would indeed go pretty fast (55-60MPH was fast for most roads of the day) with 30-35 horsepower, but the straight-six engine was as big as some KIAs and quite advanced (forced-oiling, valve-in-head): it would burst to 80-100HP (~70KW).
When I was a boy, I lived in sight of a 2-Buick (50KW) AM tower. In those days, these rigs were always tube and almost always plate-modulated, which means there was also a 25+KW Audio Amplifier in the shack. This little 50KW rig stunted my electronic studies: I could pick up the station with any darn bit of circuitry, without trying, whether I wanted to or not; and no simple "tuner" would bring in any other station. Even on a 5-tank superhet, other stations always had some of that station behind them.
In the town down the road, some total idiot was building megawatt low-RF buzz-boxes to melt metal. I've been inside the place: aside from the fright that goes with ton-pots of liquid steel, there is the thought that they were melted without any fire or arc, just RF. Inductotherm became a prime source of ultra-pure alloys.
An externally hosted image should be here but it was not working when we last tested it.
will make 2.015 MegaWatts as an AB1 audio amplifier. 22,500V 125A in the plates, 16V 600A in the filament (directly-heated tetrode). Mostly cooled by water but they want forced air in a few places the water won't reach.
"It was also noted how much more quietly water-cooled transmitters operate than new ones with air blowers......"
> with zero (!!!!!) input power
Same again. However if you look at the input capacitance and grid-swing, it needs a couple hundred VA of unreal power to hit the top of the audio band. In an untuned circuit, this amounts to hundreds of watts of potential real power: forget 12AU7 or even 6550, think a pair of 833.
> WLW
More entertaining than accurate. The 8-foot horn on the mike dates from the era before tubes. (Water-cooled carbon-mikes in series with a spark-gap or RF-alternator.) Film-sound was using automatic limiters long before WLW was pushing a half-megawatt, and it would be insane to expose all that gear to un-limited audio. (Crosley was nuts, but not flippin' insane.)
> "Crosley noted, bitterly, that 50 kW wasn't that many more horsepower than a couple of speeding Buicks."
He was wrong. 50,000 Watts is 66 horsepower. A 1938 Buick would indeed go pretty fast (55-60MPH was fast for most roads of the day) with 30-35 horsepower, but the straight-six engine was as big as some KIAs and quite advanced (forced-oiling, valve-in-head): it would burst to 80-100HP (~70KW).
When I was a boy, I lived in sight of a 2-Buick (50KW) AM tower. In those days, these rigs were always tube and almost always plate-modulated, which means there was also a 25+KW Audio Amplifier in the shack. This little 50KW rig stunted my electronic studies: I could pick up the station with any darn bit of circuitry, without trying, whether I wanted to or not; and no simple "tuner" would bring in any other station. Even on a 5-tank superhet, other stations always had some of that station behind them.
In the town down the road, some total idiot was building megawatt low-RF buzz-boxes to melt metal. I've been inside the place: aside from the fright that goes with ton-pots of liquid steel, there is the thought that they were melted without any fire or arc, just RF. Inductotherm became a prime source of ultra-pure alloys.
PRR, that sounds like a nice match for LS3/5As. Great links.
Had the same childhood experience, WBAL's transmitter a hundred meters from our back yard. That RFI challenge will toughen up a youngster.
Had the same childhood experience, WBAL's transmitter a hundred meters from our back yard. That RFI challenge will toughen up a youngster.
AM radio transmitter
I also grew up a block from a 10KW AM radio station. WVCG Miami. Is this a coincidence? Or did the large RF field cause a genetic mutation that resulted in a fascination with big vacuum tube amplifiers?
I remember picking this station up with a crystal diode connected directly across a small speaker, using the chain link fence for an antenna. After lightning zapped the phone lines, it took 2 years for the phone guys to finally get the radio station out of the phones.
Same deal as far as trying to pick up any other station. Tuned circuits, hah. Solid state, yeah right, intermod city.
I also grew up a block from a 10KW AM radio station. WVCG Miami. Is this a coincidence? Or did the large RF field cause a genetic mutation that resulted in a fascination with big vacuum tube amplifiers?
I remember picking this station up with a crystal diode connected directly across a small speaker, using the chain link fence for an antenna. After lightning zapped the phone lines, it took 2 years for the phone guys to finally get the radio station out of the phones.
Same deal as far as trying to pick up any other station. Tuned circuits, hah. Solid state, yeah right, intermod city.
> did the large RF field cause a genetic mutation that resulted in a fascination with big vacuum tube amplifiers?
Yes.
> I remember picking this station up with a crystal diode connected directly across a small speaker, using the chain link fence for an antenna.
Fence Antenna? A few inches of wire was plenty. A diode was not always needed if there were an imperfect contact anywhere.
> took 2 years for the phone guys to finally get the radio station out of the phones.
We always had radio in the phone. Everybody in the area complained, the telco came around several times and put RF suppressors on the line and phone-base and in the handset. It would be less for a while and then get worse again.
Are we drifting away from the original chip-driven killer-amp question?
Yes.
> I remember picking this station up with a crystal diode connected directly across a small speaker, using the chain link fence for an antenna.
Fence Antenna? A few inches of wire was plenty. A diode was not always needed if there were an imperfect contact anywhere.
> took 2 years for the phone guys to finally get the radio station out of the phones.
We always had radio in the phone. Everybody in the area complained, the telco came around several times and put RF suppressors on the line and phone-base and in the handset. It would be less for a while and then get worse again.
Are we drifting away from the original chip-driven killer-amp question?
Yes, we have. It seems that my comment relating to "the biggest tube amplifier on the planet" provoked a "my story is bigger than yours" response. I am fully aware of how plate AM modulators work. I am a transmitter design engineer (albeit, on a much smaller scale).
My comments were intended to hopefully convince the originator to build a more realistic amplifier for his first attempt. There is no quicker way to kill a new hobby than an ultra expensive failure (or worse) on the first attempt.
My comments were intended to hopefully convince the originator to build a more realistic amplifier for his first attempt. There is no quicker way to kill a new hobby than an ultra expensive failure (or worse) on the first attempt.
> hopefully convince the originator to build a more realistic amplifier for his first attempt
Yeah, I am still not keen on the idea. I've been bit by 600V, no fun. I have been bit by a low-power 30,000V supply, and don't know why I am alive. Besides which, it seems like an awful lot of work and expense and special-parts for a power level that will rarely be needed (though SY seems to have the load for kilo-amps). And at "serious" power levels, the tubes are not the biggest expense: iron and capacitors get insanely priced, especially for custom winding. Considering the killer sound you can get from a couple 2A3 or one 300B, or a quartet of 6550, I would go in those directions before investing in the outer limits of hollow-state audio.
But I am curious in theory, did give one retail source, and here's another.
GreenStone claims to be a major supplier and rebuilder of big tubes, and the link lists many big-tank tubes. Sadly they don't list prices, you have to call/email for a quote. However that page does link to short-form data for many big tubes, including all four variants of 3CX3000:
For all:
Filament: Thoriated Tungsten mesh
Voltage 7.5 Volts
Current 51.5 Amps
3CX3000A1/8238 - Amplification factor 5
3CX3000A7 - Amplification factor 160
3CX3000F1/8239 - Amplification factor 5
3CX3000F7 (wrongly linked to 3CX3000A7) - Amplification factor 160
OK: "A" suffix means Coaxial socket, 9 inches tall; "F" means Flexible leads, 18 inches tall. "1" suffix means Mu=5, good for SET A1 audio; "7" means Mu=160, good for some RF amps and for zero-bias Class B2 audio amplifiers.
All are air-cooled. Sucking thousands of watts of heat through a 4-inch space suggests HUGE air-flow, such that you would not want this in the same wing of your house as you listen in. I suppose ~800 Watts of available power means you can tolerate some loss in long lines. Water-cooled might be quieter, but water-cooling kiloVolt plates has its own difficulties, dangers, and cost. (Might be a good use for the radiator from that Buick.)
Total filament+plate power may be in excess of what a common US branch power circuit can deliver.
Yeah, I am still not keen on the idea. I've been bit by 600V, no fun. I have been bit by a low-power 30,000V supply, and don't know why I am alive. Besides which, it seems like an awful lot of work and expense and special-parts for a power level that will rarely be needed (though SY seems to have the load for kilo-amps). And at "serious" power levels, the tubes are not the biggest expense: iron and capacitors get insanely priced, especially for custom winding. Considering the killer sound you can get from a couple 2A3 or one 300B, or a quartet of 6550, I would go in those directions before investing in the outer limits of hollow-state audio.
But I am curious in theory, did give one retail source, and here's another.
GreenStone claims to be a major supplier and rebuilder of big tubes, and the link lists many big-tank tubes. Sadly they don't list prices, you have to call/email for a quote. However that page does link to short-form data for many big tubes, including all four variants of 3CX3000:
For all:
Filament: Thoriated Tungsten mesh
Voltage 7.5 Volts
Current 51.5 Amps
3CX3000A1/8238 - Amplification factor 5
3CX3000A7 - Amplification factor 160
3CX3000F1/8239 - Amplification factor 5
3CX3000F7 (wrongly linked to 3CX3000A7) - Amplification factor 160
OK: "A" suffix means Coaxial socket, 9 inches tall; "F" means Flexible leads, 18 inches tall. "1" suffix means Mu=5, good for SET A1 audio; "7" means Mu=160, good for some RF amps and for zero-bias Class B2 audio amplifiers.
All are air-cooled. Sucking thousands of watts of heat through a 4-inch space suggests HUGE air-flow, such that you would not want this in the same wing of your house as you listen in. I suppose ~800 Watts of available power means you can tolerate some loss in long lines. Water-cooled might be quieter, but water-cooling kiloVolt plates has its own difficulties, dangers, and cost. (Might be a good use for the radiator from that Buick.)
Total filament+plate power may be in excess of what a common US branch power circuit can deliver.
PRR said:
My first job was at a radio station that still had an RCA BTA-50F1 as backup, installed in 1947 long before remote controls were available (or legal.) On-site technicians watched meters and turned dials all day. The room had to be quiet.
The supply fans for the F1 stood four around feet tall and the noise issue was solved by placing them in the basement and ducting the supply through the floor. The exhaust was simple convection out the ceiling. Air entered the cabinet via the white ceramic base shown in the top photo. (The site has had a custom hand jitney for replacing the 100 pound plus modulator tubes.)
http://hawkins.pair.com/wknr.html#More about the 5671 tube
No stereo 3cx3000 amp is going to be portable anyway. If you get one of those past your sigificant other blasting holes in the floor should be a cakewalk.
(BTW, thx PRR for the model info. Interesting stuff.)
Less-short form data for 3CX3000F1
Maximum Ratings
Plate Voltage (Volts) 6,000
Plate current (Amps) 2.5
Typical Operation
AB1 Grid Driven AF Amplifier or Modulator {clearly push-pull}
Plate Voltage (Volts) 5,500
Plate Current (Amps) 2.2
Drive Power (Watts)
Output Power (kiloWatts) 8.25
"Characteristics and Operating values are based on performance tests. Figures may change without notice as the result of additional data or product refinement. CPI Eimac Division should be consulted before using this information for final equipment design."
This page seems to list the -A7 as a catalog item, "Preço (R$) 11.660,71". I'm not finding surplus sources of this tube; it may be a closed ecosystem where most used-up tubes get rebuilt and put back in service.
-----
Is it at all possible a key stuck, and the plan is 3CX300?
Typical Operation,
Class A, Audio Amplifier (single tube)
Plate voltage 500 V
Grid voltage -55 V
Peak grid drive 120 VP-P
Plate current, no signal 80 mA
Plate current, max signal 95 mA
Effective load resistance 1600 ohms
Distortion at 1 watt into 8 ohms 0.67 %
Power output at 5% distortion 15 W
Typical Operation,
Class AB1, Audio Power Amplifier, Push-Pull
Plate voltage 500 V
Grid voltage -45 V
Peak grid drive 100 VP-P
Plate current, no signal (both tubes) 300 mA
Load resistance, plate-to-plate 2000 ohms
Power output 40 W
Hmmm, performs like a 6550 except ceramic-metal, no warm orange glow, yet much more heater power. These ratings may be inside the 30W limit without a blower; just 12CFM and 1" water will let you run 300 watt dissipation. 1.5KV and a blower into two tubes will get you around 500 watts. The ~300Vpp grid drive is not out of sight for a power-amp chip driving a 6KCT:8 transformer backwards. The load resistance will be a couple of Kohms, not an impossible impedance for hi-fi even with the excess insulation needed at these voltages. There may be over 100W SET if you blow hard enough (hisssss).
----
> fans for the F1 stood four around feet tall
Drifting again.... a dedicated school Music Building was built about 1928, with climate control (a Carrier water-system "air conditioner"; freon coolers were not common and ammonia far too dangerous). The main fan stands 8 foot tall and spins a couple turns per second, VERY quietly, furthur quieted by a muffler that IS a room. They are replacing it with banks of 1-foot screamers into 1'x1.5'x6' lined ducts. I'm going to hate it. I'm going to beg for a bypass for recording sessions.
> 5671 ...a huge number of them built for induction-heating service during WWII
Crankshaft hardening.
> thx PRR for the model info. Interesting stuff
??? I may be confused with someone else. Thank them, whoever they are.
Maximum Ratings
Plate Voltage (Volts) 6,000
Plate current (Amps) 2.5
Typical Operation
AB1 Grid Driven AF Amplifier or Modulator {clearly push-pull}
Plate Voltage (Volts) 5,500
Plate Current (Amps) 2.2
Drive Power (Watts)
Output Power (kiloWatts) 8.25
"Characteristics and Operating values are based on performance tests. Figures may change without notice as the result of additional data or product refinement. CPI Eimac Division should be consulted before using this information for final equipment design."
This page seems to list the -A7 as a catalog item, "Preço (R$) 11.660,71". I'm not finding surplus sources of this tube; it may be a closed ecosystem where most used-up tubes get rebuilt and put back in service.
-----
Is it at all possible a key stuck, and the plan is 3CX300?
Typical Operation,
Class A, Audio Amplifier (single tube)
Plate voltage 500 V
Grid voltage -55 V
Peak grid drive 120 VP-P
Plate current, no signal 80 mA
Plate current, max signal 95 mA
Effective load resistance 1600 ohms
Distortion at 1 watt into 8 ohms 0.67 %
Power output at 5% distortion 15 W
Typical Operation,
Class AB1, Audio Power Amplifier, Push-Pull
Plate voltage 500 V
Grid voltage -45 V
Peak grid drive 100 VP-P
Plate current, no signal (both tubes) 300 mA
Load resistance, plate-to-plate 2000 ohms
Power output 40 W
Hmmm, performs like a 6550 except ceramic-metal, no warm orange glow, yet much more heater power. These ratings may be inside the 30W limit without a blower; just 12CFM and 1" water will let you run 300 watt dissipation. 1.5KV and a blower into two tubes will get you around 500 watts. The ~300Vpp grid drive is not out of sight for a power-amp chip driving a 6KCT:8 transformer backwards. The load resistance will be a couple of Kohms, not an impossible impedance for hi-fi even with the excess insulation needed at these voltages. There may be over 100W SET if you blow hard enough (hisssss).
----
> fans for the F1 stood four around feet tall
Drifting again.... a dedicated school Music Building was built about 1928, with climate control (a Carrier water-system "air conditioner"; freon coolers were not common and ammonia far too dangerous). The main fan stands 8 foot tall and spins a couple turns per second, VERY quietly, furthur quieted by a muffler that IS a room. They are replacing it with banks of 1-foot screamers into 1'x1.5'x6' lined ducts. I'm going to hate it. I'm going to beg for a bypass for recording sessions.
> 5671 ...a huge number of them built for induction-heating service during WWII
Crankshaft hardening.
> thx PRR for the model info. Interesting stuff
??? I may be confused with someone else. Thank them, whoever they are.
"Total filament+plate power may be in excess of what a common US branch power circuit can deliver."
During my brief insanity with the prototype 833A guitar amplifier, I blew the breaker twice. This was with a 200 watt SE amplifier, being abused by my guitar playing. That session was the source of my avatar. I came to the conclusion that when I build the stereo version of this amp, I will use a fan and a close dryer vent to port the hot air out of the house.
During my brief insanity with the prototype 833A guitar amplifier, I blew the breaker twice. This was with a 200 watt SE amplifier, being abused by my guitar playing. That session was the source of my avatar. I came to the conclusion that when I build the stereo version of this amp, I will use a fan and a close dryer vent to port the hot air out of the house.
SY> issue is the required turns ratio for a suitable OPT. The tube wants that ratio very high, but that makes producing a decent transformer a very big challenge.
It isn't the ratio. It is the maximum impedance. A 0.1Ω:1K (1:100) is easy. A 100K:100K (1:1) with wide-band response is hard.
rdf> 1953 Eimac catalogue has these operating recommendations ... effective load plate-to-plate of 1170 ohms
1K2 is not even a high impedance. Especially for a fat core, which will give more inductance per turn than a small core. It is true that it will be a physically large winding, implying high capacitance. High compared to our little 2-inch core-tongue windings, at least. I don't think the AC problem is a big deal. After all, the better AM transmitters used such bottles and iron and got far past 10KC, even though they had to be over-wound and gapped to handle the large DC flowing to the final RF amp (as this one needs to be gapped to handle SET DC current). The 10KV insulation requirement is tougher: up to 300V, a good coat of paint is enough, and 1KV just needs careful layering and some paper, but 10KV needs serious attention to breakdown which may reduce coupling (increase leakage inductance). Double-bobbin is traditionally bad for wide-band iron, but good insulation, and I have seen multi-bobbin transformers do much better than I would have expected. If the dude worked for Triad as a transformer engineer, I'd accept his claims with only token token checking. After all, back in the Glory Days of AM tube-transmitter sales, Triad would have been invited to bid the iron for every new design.
It isn't an impossible job. It is just far outside the scale of things that we little people are used to.
In fact, you would almost get audio if you stole the pole-transformer from out at the street (someone else's street; you need yours). A 2880:240VAC transformer isn't a Standard (13KV used to be the standard street voltage), but can be ordered. 240V into 8 ohms is just 30 Amps, we have plenty of Amps so we have low resistance losses. Working around 60V of audio in a 240V-rated power winding, we are far from saturation, distortion is low. DC current is a problem, but a pole transformer is 50,000VA and we are putting maybe 2,000VA to 4,000VA across it, it won't saturate bad. Losses are low to reduce utility company overhead, which means ample inductance at 60Hz and probably high impedance to 20Hz, also low leakage reactance and capacitance to reduce imaginary power (which is real power lost in the feeder), so bandwidth will be far-far above 60Hz (rather lame small power iron does 5KHz).
And if you must fret about the ratio, 1K2:8 is only 12:1 turns ratio, much less than the popular and manageable 8K:8 (32:1 turns ratio).
It isn't the ratio. It is the maximum impedance. A 0.1Ω:1K (1:100) is easy. A 100K:100K (1:1) with wide-band response is hard.
rdf> 1953 Eimac catalogue has these operating recommendations ... effective load plate-to-plate of 1170 ohms
1K2 is not even a high impedance. Especially for a fat core, which will give more inductance per turn than a small core. It is true that it will be a physically large winding, implying high capacitance. High compared to our little 2-inch core-tongue windings, at least. I don't think the AC problem is a big deal. After all, the better AM transmitters used such bottles and iron and got far past 10KC, even though they had to be over-wound and gapped to handle the large DC flowing to the final RF amp (as this one needs to be gapped to handle SET DC current). The 10KV insulation requirement is tougher: up to 300V, a good coat of paint is enough, and 1KV just needs careful layering and some paper, but 10KV needs serious attention to breakdown which may reduce coupling (increase leakage inductance). Double-bobbin is traditionally bad for wide-band iron, but good insulation, and I have seen multi-bobbin transformers do much better than I would have expected. If the dude worked for Triad as a transformer engineer, I'd accept his claims with only token token checking. After all, back in the Glory Days of AM tube-transmitter sales, Triad would have been invited to bid the iron for every new design.
It isn't an impossible job. It is just far outside the scale of things that we little people are used to.
In fact, you would almost get audio if you stole the pole-transformer from out at the street (someone else's street; you need yours). A 2880:240VAC transformer isn't a Standard (13KV used to be the standard street voltage), but can be ordered. 240V into 8 ohms is just 30 Amps, we have plenty of Amps so we have low resistance losses. Working around 60V of audio in a 240V-rated power winding, we are far from saturation, distortion is low. DC current is a problem, but a pole transformer is 50,000VA and we are putting maybe 2,000VA to 4,000VA across it, it won't saturate bad. Losses are low to reduce utility company overhead, which means ample inductance at 60Hz and probably high impedance to 20Hz, also low leakage reactance and capacitance to reduce imaginary power (which is real power lost in the feeder), so bandwidth will be far-far above 60Hz (rather lame small power iron does 5KHz).
And if you must fret about the ratio, 1K2:8 is only 12:1 turns ratio, much less than the popular and manageable 8K:8 (32:1 turns ratio).
> I will use a fan and a clothes dryer vent to port the hot air out of the house.
That's interesting. The tube is 4 inches, a dryer vent is 4 inches; the tube is 2,000-4,000 Watts and that is what a dryer puts out when the clothes get dry and it does not stop. So we have a low-low-cost prototype system that will move the required heat without burning the house down. In fact if the dryer bucket didn't squeak, if buttons and pennies didn't clank, the fan-noise is not SO loud that it could not just be in the next room with the door sealed. At worst, you want to seal the inlet and muffle it or run it outside, and double-wall the ducts.
And we also have a killer combination product. Pull air over the tube, then through the dryer bucket, and out of the house. Play your music AND dry your clothes at the same time on the same electric bill. You will have the driest clothes in town. If you can't justify this at your house, find a bar-band playing next to a laundromat and sell the scheme to both places: free dryer-heat in the laundromat and the laundry owners can slip next door for a drink while the clothes dry.
That's interesting. The tube is 4 inches, a dryer vent is 4 inches; the tube is 2,000-4,000 Watts and that is what a dryer puts out when the clothes get dry and it does not stop. So we have a low-low-cost prototype system that will move the required heat without burning the house down. In fact if the dryer bucket didn't squeak, if buttons and pennies didn't clank, the fan-noise is not SO loud that it could not just be in the next room with the door sealed. At worst, you want to seal the inlet and muffle it or run it outside, and double-wall the ducts.
And we also have a killer combination product. Pull air over the tube, then through the dryer bucket, and out of the house. Play your music AND dry your clothes at the same time on the same electric bill. You will have the driest clothes in town. If you can't justify this at your house, find a bar-band playing next to a laundromat and sell the scheme to both places: free dryer-heat in the laundromat and the laundry owners can slip next door for a drink while the clothes dry.
You might have missed my point.
I live in South Florida. The current outside temp is 94 degrees F. This is typical from June untill September, regardless of what they tell the tourists on TV. The house has central air. During the day the best it can do is keep the house at 80 F. My lab/ listening room is small. Ten by eleven feet. Turn on the computer and a small single ended amp and the temperature rises quickly. To remedy this I have added a small window air conditioner in this room. This helps and thermal equilibrium as maintained when I have the computer and the 845SE amp running. That amp dumps about 400 watts of heat into the room. So now I pay for the electricity to make the heat (and some sound) and I pay again for the electricity to remove the heat. I figure that an 833SE amp will produce about 1000 watts of heat. Thermal equilibrium is now lost, the room will heat up.
To solve this problem I have made a hood that goes above the amp (similar to a range hood) out of thin plywood. This is connected to a plastic clothes dryer vent from Home Depot. I added a 4 inch muffin fan to the dryer vent to move the hot air. The dryer vent has little plastic flappers on it to keep the critters out when it is off.
When I build the 833A amp I will design the housing in such a way that I can connect the vent hose directly to the amps cooling system.
Years ago I built an amplifier that used a large power transformer as an output transformer. The transformer was from a Globe King transmitter. I used the center tapped HV winding for the primary, and the 110 volt winding for the speaker. Tubes were 4CX250 in push pull. The thing put out lots of power, but had no high frequency response. I had no measurement equipment at the time (at least 25 years ago), so I have no hard numbers. It was used as a guitar amplifier. Well one day the guitar player had it turned up all the way, doing his best Jimi impersonation, when the speakers blew. The speakers opened up, giving the amp an open circuit. I wasn't there but I heard about the "wicked fireball" for years. Most of the amp, including the output and power transformers were toast!
My current guitar amps have MOV's across the output. Pick a voltage that will never conduct with a normal load. If the load goes open, the MOV will conduct, and usually fail to a short, saving the transformers. These things can be made to create all sorts of fuzz tone effects if connected in the right places.
I live in South Florida. The current outside temp is 94 degrees F. This is typical from June untill September, regardless of what they tell the tourists on TV. The house has central air. During the day the best it can do is keep the house at 80 F. My lab/ listening room is small. Ten by eleven feet. Turn on the computer and a small single ended amp and the temperature rises quickly. To remedy this I have added a small window air conditioner in this room. This helps and thermal equilibrium as maintained when I have the computer and the 845SE amp running. That amp dumps about 400 watts of heat into the room. So now I pay for the electricity to make the heat (and some sound) and I pay again for the electricity to remove the heat. I figure that an 833SE amp will produce about 1000 watts of heat. Thermal equilibrium is now lost, the room will heat up.
To solve this problem I have made a hood that goes above the amp (similar to a range hood) out of thin plywood. This is connected to a plastic clothes dryer vent from Home Depot. I added a 4 inch muffin fan to the dryer vent to move the hot air. The dryer vent has little plastic flappers on it to keep the critters out when it is off.
When I build the 833A amp I will design the housing in such a way that I can connect the vent hose directly to the amps cooling system.
Years ago I built an amplifier that used a large power transformer as an output transformer. The transformer was from a Globe King transmitter. I used the center tapped HV winding for the primary, and the 110 volt winding for the speaker. Tubes were 4CX250 in push pull. The thing put out lots of power, but had no high frequency response. I had no measurement equipment at the time (at least 25 years ago), so I have no hard numbers. It was used as a guitar amplifier. Well one day the guitar player had it turned up all the way, doing his best Jimi impersonation, when the speakers blew. The speakers opened up, giving the amp an open circuit. I wasn't there but I heard about the "wicked fireball" for years. Most of the amp, including the output and power transformers were toast!
My current guitar amps have MOV's across the output. Pick a voltage that will never conduct with a normal load. If the load goes open, the MOV will conduct, and usually fail to a short, saving the transformers. These things can be made to create all sorts of fuzz tone effects if connected in the right places.
If you are referring to the incident many years ago where the amp self destructed, no the band was practicing at night in a warehouse with no one around.
Even though I live in a residential neighborhood, I generally don't get complaints from my neighbors. They have become used to loud sounds coming from my house. When my daughter lived at home she played the drums and piano daily.
I did not get any "complaints" but the 833A guitar amp experiment, particularly the feedback got a few "What the #%$@ was that?" comments from as far as a block away.
http://www.tubelab.com/833SE.htm scroll down to the bottom.
I love Jimi, and the 845SE is the only SE amp that I have heard do Jimi right.
I actually saw Jimi perform live at the Orange Bowl (an American football stadium) in 1967 or 1968, although it is hard to remember back that far. He did light his guitar on fire, while it was laying on the stage screaming through multiple Echoplex units. As the flames died down he picked it up and continued playing.
Even though I live in a residential neighborhood, I generally don't get complaints from my neighbors. They have become used to loud sounds coming from my house. When my daughter lived at home she played the drums and piano daily.
I did not get any "complaints" but the 833A guitar amp experiment, particularly the feedback got a few "What the #%$@ was that?" comments from as far as a block away.
http://www.tubelab.com/833SE.htm scroll down to the bottom.
I love Jimi, and the 845SE is the only SE amp that I have heard do Jimi right.
I actually saw Jimi perform live at the Orange Bowl (an American football stadium) in 1967 or 1968, although it is hard to remember back that far. He did light his guitar on fire, while it was laying on the stage screaming through multiple Echoplex units. As the flames died down he picked it up and continued playing.
While searching for a 3CX3000 datasheet, this thread showed up.
I'm getting a pair of RF amplifiers tomorrow with 3CX3000's in them. But I read this old thread, very interesting and amusing.
Whew! Ok. it's all been said.. it's a zero bias triode. At 3KV on the anode, it looks like it will draw about 100mA. -25V will cut it off. +25V to the grid you get about 600mA. Not linear at all for class A unless you keep some positive DC on the grid and raise the plate current into a more linear region. Grid will dissipate up to 225 watts so that shouldn't be a problem.
How to use it for a single ended class A audio amp?
DC conditions:
anode voltage 3KV
grid voltage +23V
anode current 500mA
anode dissipation 1500W (don't wear it out!)
grid current 80mA (approx.)
grid dissipation 2 watts
load resistance 3000 ohms ?
output 375 watts
anode voltage 3KV
grid voltage +40V
anode current 1A
anode dissipation 3000W (wear it out!)
grid current 200mA (approx.)
grid dissipation 8 watts
load resistance 1500 ohms ?
output 750 watts
Either case, the driver would have to supply a positive DC voltage and regulate it according to the input waveform, regardless of how the grid current varies perhaps from 0 to 0.5 amp. It would be easy for a loaded pass transistor setup. Need some protection in case of a flashover so the drive circuit does not get blown to ashes. Or use a bank of 6L6's or 6550's as the drive source, sure theyd explode under the same conditions but at least they fit in sockets and can easily be replaced. Besides, it's a tube amp, something has to glow.
But where to get the OPT, and get it from someone who can wind it for the high voltage (12KV test insulation) and deal with the core magnetization issues? Sell your wife's new mercedes and call Peter Dahl. He'll probably throw in the obligatory ball-type spark gap for the primary protection for free on such an order.
Also will be required a 7.5VAC @ 51.5A center tapped transformer for the filament. A socket, unless you have the flying lead (F type). It needs minimum airflow of 67CFM@1.2" H2O. Along with this, airflow switch interlock to the heater voltage, and a timer for heater voltage to plate voltage. Also a circuit to monitor for clipping, underload, overload, so the OPT does not get spiked. I leave the HV supply to the imagination. I prefer choke input filters and mercury vapor rectifiers, perhaps 673's for this job. I think you could get a monoblock into a 7FT rack if crafty with tools.
Sorry, I just had to comment on this project.
I'm getting a pair of RF amplifiers tomorrow with 3CX3000's in them. But I read this old thread, very interesting and amusing.
Whew! Ok. it's all been said.. it's a zero bias triode. At 3KV on the anode, it looks like it will draw about 100mA. -25V will cut it off. +25V to the grid you get about 600mA. Not linear at all for class A unless you keep some positive DC on the grid and raise the plate current into a more linear region. Grid will dissipate up to 225 watts so that shouldn't be a problem.
How to use it for a single ended class A audio amp?
DC conditions:
anode voltage 3KV
grid voltage +23V
anode current 500mA
anode dissipation 1500W (don't wear it out!)
grid current 80mA (approx.)
grid dissipation 2 watts
load resistance 3000 ohms ?
output 375 watts
anode voltage 3KV
grid voltage +40V
anode current 1A
anode dissipation 3000W (wear it out!)
grid current 200mA (approx.)
grid dissipation 8 watts
load resistance 1500 ohms ?
output 750 watts
Either case, the driver would have to supply a positive DC voltage and regulate it according to the input waveform, regardless of how the grid current varies perhaps from 0 to 0.5 amp. It would be easy for a loaded pass transistor setup. Need some protection in case of a flashover so the drive circuit does not get blown to ashes. Or use a bank of 6L6's or 6550's as the drive source, sure theyd explode under the same conditions but at least they fit in sockets and can easily be replaced. Besides, it's a tube amp, something has to glow.
But where to get the OPT, and get it from someone who can wind it for the high voltage (12KV test insulation) and deal with the core magnetization issues? Sell your wife's new mercedes and call Peter Dahl. He'll probably throw in the obligatory ball-type spark gap for the primary protection for free on such an order.
Also will be required a 7.5VAC @ 51.5A center tapped transformer for the filament. A socket, unless you have the flying lead (F type). It needs minimum airflow of 67CFM@1.2" H2O. Along with this, airflow switch interlock to the heater voltage, and a timer for heater voltage to plate voltage. Also a circuit to monitor for clipping, underload, overload, so the OPT does not get spiked. I leave the HV supply to the imagination. I prefer choke input filters and mercury vapor rectifiers, perhaps 673's for this job. I think you could get a monoblock into a 7FT rack if crafty with tools.
Sorry, I just had to comment on this project.
The easiest way to get audio out of the thing would be to use it making AM modulated RF for a plasma speaker. Heinously inefficient, but quite HiFi.
Im afraid Peter Dahl retired and closed up shop a little while back.
Im afraid Peter Dahl retired and closed up shop a little while back.
Ya unfortunately Peter got cancer. But he sold his business but the new transformers will be made by MagCap Engineering in Canton Massachusetts.
RCA BTA-10K
About ten years ago I hauled two complete RCA BTA-10K AM (10 kWatt) broadcast xmitters home. One I stripped for parts at the transmitter site, the other came home before being parted out. I still have the unmolested RF final cabinet (of four in a complete system) and one of the matching cabinets with nothing inside. This is looking for a new owner. It used three of the RCA air cooled triodes rated at 3.5 kW pD each. I forget the number. They were replaced by a drop-in 5 kW pD EEV tube somewhere along the way. I have the EEV tubes tucked away and one dud original RCA which is in my tube collection display. I also have the heated oven xtal master oscillators from four transmitters and two frequency assignments. These use a single 807 tube. I could go on the air with just one of these alone! 😉
I made a LF RF power oscillator with three of these 5 kW tubes in parallel. Dropped a long secondary into the huge tank coil and made a nice Vac Tube Tesla Coil operating at 100 KHz. What a power hungry fat hissing plasma! That would have been a nice plasma woofer if plate modulated!
These circa 1954 transmitters were a real good source of high tube power components and quality ceramic tube sockets, filament xfmers, a lesser plate xfmer and choke for the driver stuff and chokes, both power and modulation.
A complete 3-cabinet 10 kW Collins broadcast transmitter recently taken out of a radio station storage room in Ontario went on ebay for about $690.00 less than a month ago (way deal!). It used a pair of 3CX2500 (modulators) and three 3CX3000 tubes in the RF final which were included. I was tempted as it was less than an hour's drive away but I don't have the room, and I don't play with this dangerous voltage and power any more because I decided I like being here.
About ten years ago I hauled two complete RCA BTA-10K AM (10 kWatt) broadcast xmitters home. One I stripped for parts at the transmitter site, the other came home before being parted out. I still have the unmolested RF final cabinet (of four in a complete system) and one of the matching cabinets with nothing inside. This is looking for a new owner. It used three of the RCA air cooled triodes rated at 3.5 kW pD each. I forget the number. They were replaced by a drop-in 5 kW pD EEV tube somewhere along the way. I have the EEV tubes tucked away and one dud original RCA which is in my tube collection display. I also have the heated oven xtal master oscillators from four transmitters and two frequency assignments. These use a single 807 tube. I could go on the air with just one of these alone! 😉
I made a LF RF power oscillator with three of these 5 kW tubes in parallel. Dropped a long secondary into the huge tank coil and made a nice Vac Tube Tesla Coil operating at 100 KHz. What a power hungry fat hissing plasma! That would have been a nice plasma woofer if plate modulated!
These circa 1954 transmitters were a real good source of high tube power components and quality ceramic tube sockets, filament xfmers, a lesser plate xfmer and choke for the driver stuff and chokes, both power and modulation.
A complete 3-cabinet 10 kW Collins broadcast transmitter recently taken out of a radio station storage room in Ontario went on ebay for about $690.00 less than a month ago (way deal!). It used a pair of 3CX2500 (modulators) and three 3CX3000 tubes in the RF final which were included. I was tempted as it was less than an hour's drive away but I don't have the room, and I don't play with this dangerous voltage and power any more because I decided I like being here.
- Status
- Not open for further replies.