I'm in the process if acquiring a pair of 1.5K 200W outputs. They have 40% taps.
Clearly PPP arrangements is the way to go, but what tubes would be optimal here?
6 X KT88
8 X EL34
8 X 6L6GC
Other options? Sweeps, etc.?
I have a few options for PS voltages.
A 1.2KV CT transformer that can be used with choke input for 520-560V at 500mA. Obviously KT88 or 6550 variety for UL here.
A pair of transformers that will probably do 3A or so at around 330-360V. Thinking 6L6 or EL34 for this.
Suggestions?
Thanks!
Blair
Clearly PPP arrangements is the way to go, but what tubes would be optimal here?
6 X KT88
8 X EL34
8 X 6L6GC
Other options? Sweeps, etc.?
I have a few options for PS voltages.
A 1.2KV CT transformer that can be used with choke input for 520-560V at 500mA. Obviously KT88 or 6550 variety for UL here.
A pair of transformers that will probably do 3A or so at around 330-360V. Thinking 6L6 or EL34 for this.
Suggestions?
Thanks!
Blair
4 x JJ EL509, PL519 or 6P45S gives >200 W in pentode connection.
This reguires 120...150 V regulated screen voltage and 450...500 V at anode.
I just few weeks ago built this circuit and tested it also with EL509 with 4k to 8 ohms OPT and 120 V at screen, not UL. I got 100 W.
Just douple the output tubes and reduce screen series resistors to some 47...100 ohms.
That's all.
This reguires 120...150 V regulated screen voltage and 450...500 V at anode.
I just few weeks ago built this circuit and tested it also with EL509 with 4k to 8 ohms OPT and 120 V at screen, not UL. I got 100 W.
Just douple the output tubes and reduce screen series resistors to some 47...100 ohms.
That's all.
An externally hosted image should be here but it was not working when we last tested it.
With two EL509/519 a 200W amp can't survive very long,Pa max=35W !
Input power 300W with classB, 2/3th (200W) to the output lieves 100W (50W pro tube) to be dissipated.
It could be possible with a pair of YL1350 (sold by Ask Jan First).The Pa max= 75W.
Only small problem,heater on 12,6V.
Mona
Input power 300W with classB, 2/3th (200W) to the output lieves 100W (50W pro tube) to be dissipated.
It could be possible with a pair of YL1350 (sold by Ask Jan First).The Pa max= 75W.
Only small problem,heater on 12,6V.
Mona
Attachments
Last edited:
Right,
I'm not bent on the 200W output as much as 150-175. They are Sowter, so they are rated at full power at full bandwidth, but avoiding saturation is a good thing.
The 1.5K A-A is why I'm leaning towards an 8 pack of EL34s or 6L6 tubes. I cannot find the UL data for 6L6 though. I was thinking at around 340-360V in UL, if each pair is good for 30W, then 8 should get me 120W. The GC variety even more.
The EL34 is attractive because its easier to drive and 8 should give me 150W or better.
Regarding the oddball YL tube, I have two filament windings, so a 12V tube is no big deal.
Thanks!
I'm not bent on the 200W output as much as 150-175. They are Sowter, so they are rated at full power at full bandwidth, but avoiding saturation is a good thing.
The 1.5K A-A is why I'm leaning towards an 8 pack of EL34s or 6L6 tubes. I cannot find the UL data for 6L6 though. I was thinking at around 340-360V in UL, if each pair is good for 30W, then 8 should get me 120W. The GC variety even more.
The EL34 is attractive because its easier to drive and 8 should give me 150W or better.
Regarding the oddball YL tube, I have two filament windings, so a 12V tube is no big deal.
Thanks!
Hi,
Yes, that's where I got the data, but they are 20 years old and Brian said he did not remember or have documentation of the tubes intended.
They have a static DCR of 11.5 ohms across both primaries, so I would think they can handle a good amount of current making the possibilities a bit mor open.
Yes, that's where I got the data, but they are 20 years old and Brian said he did not remember or have documentation of the tubes intended.
They have a static DCR of 11.5 ohms across both primaries, so I would think they can handle a good amount of current making the possibilities a bit mor open.
I would go with 8 6L6/807 types. They enjoy AB2 and they are cheap enough that you can buy a fair few and match them nicely. Check out Chris and Georges design in the 6l6 AB2 thread:
http://www.diyaudio.com/forums/tubes-valves/133034-6l6gc-ab2-amp.html
I built a couple of versions of this and squeezed 100W out of four 807 types in triode. That was tube burning though, I reckon 160W from eight at around 450V would be fine.
Cheers Matt.
http://www.diyaudio.com/forums/tubes-valves/133034-6l6gc-ab2-amp.html
I built a couple of versions of this and squeezed 100W out of four 807 types in triode. That was tube burning though, I reckon 160W from eight at around 450V would be fine.
Cheers Matt.
There have been a few valid choices presented, so I will offer up my opinion in case anyone wants it.
Saturation sounds bad on the typical low or mid powered amp. Core saturation on an amp of this magnitude can cause rapid failure. An amp capable of 200 WPC will need a power supply and output tubes capable of an amp of peak current. When the OPT saturates its inductance drops to near zero. This puts the fully conducting output tube and the OPT's DCR directly across the power supply. Something has got to give!
I have been experimenting with big powered amps for a while, and have come to one conclusion. SWEEP TUBES! I have a set of 1250 ohm Plitron OPT's rated for 400 Watts from 20 Hz to 20 KHz. They were intended for use with 6 or 8 KT88's per channel. I tried 6 EH KT88's (all I have) and it works, but I liked the way sweep tubes work better. Why, the sweep tube can pull the plate down to near zero, while the KT88's strain to make 50 volts. This makes the sweep tubes more efficient, and capable of more power from a given B+ than any of the typical audio tubes. They have a more dynamic sound too.
So, which sweep tube????? Answer, it doesn't make that much of a difference. For this power level you will need 500 to 550 volts for the plates of the output tubes, 150 to 250 for the screens depending on the tubes, and power for the drivers depending on what you use.
Any of the real sweep tubes capable of this power have maximum screen grid ratings of 175 to 275 volts. I use a mosfet regulator and adjust the voltage for best efficiency. This makes UL impossible.
Case 1) If you intend to listen to music even at the edge of clipping, the AVERAGE power level will be less than 20 watts, due to the "crest factor" of most music. The PEAK power can be 150 to 250 watts.
Case 2) If you might be tempted to say plug in a guitar or bass, crank beyond 11 and rip out a 12 minute solo, then the AVERAGE power might be closer to 50 watts, and the amp must be expected to run in clipping for as much as 50% of the time. Peak power can be 300 watts of more.
Case 3) If you are going to perform continuous sine wave testing for long periods of time the average power, can be equal to the peak power and will be whatever you design to, say 200 watts. This is worse case, but not realistic in the real world.
Bob Carver made his career designing amplifiers for the first case. I had one of his early 400 watt "magnetic field amplifier." The manual explained the design criteria, and warned the user not to use it for any use other than home HiFi. Of course I plugged my guitar into it, and it would play quite loudly, but shut off completely for several seconds if you hit clipping hard enough.
I design and test ALL of my amplifiers for the second case unless I know it will be used harder. The third case is rarely needed even for a dance floor subwoofer. This decision is needed to determine the number of tubes, and the power supply requirements.
You also need to look at your loudspeakers. If your OPT's are rated for an 8 ohm load, they will reflect 1500 ohms to the tubes only if the load is exactly 8 ohms. If your speakers have significant dips below 8 ohms, especially in the bass region, then the tubes will see less than 1500 ohms. This drives the peak tube current up. Each tube will see 1/4 the total OPT impedance, or 375 ohms. With a 550 volt supply, the peak tube current will be 1.46 AMPS! This is why you need so many KT88's. A single 6LW6 is rated for 1.4 amps, as are several big sweep tubes. I am assuming a class AB1 amp here. Class AB2 is not needed with sweep tubes, but may be needed to reach the needed power and efficiency level with conventional audio tubes.
I have found that to design for the second case the power supply can be designed to supply all the heater and driver continuously. The screen supply must be capable of supplying the sum of all the maximum screen currents continuously. Screen current is usually quite low, if the plate voltage is high. The plate supply needs to be sized at least equal to the total power output of the amp, but capable of supplying 130% of that value for brief periods (bass transients). A large low ESR capacitor can provide the reserve capacity for peaks. The average power value (= to power output) can be adjusted up or down depending on use. So for a 200 WPC amp using a 500 volt supply, you need 800 mA. This is quite conservative and the power transformer will run cool even if the amp is played loud for hours.
Now, back to the tubes.....which ones and how many? Any true sweep tube can eat a 500 volt supply without a problem. There are two criteria, peak current and dissipation. One pair of BIG, 35 or 40 watt sweep tubes can hit 1.5 amps, which may be enough. Two pair of 22 to 30 watt sweep tubes should have plenty of peak current capability, and 2 or 3 pair of small 17 watt tubes may be OK.
The efficiency of a typical sweep tube amp can be around 80% AT THE ONSET OF CLIPPING. This means that at 200 watts output, the power supply will furnish 250 watts. 200 will be delivered to the load, and 50 watts will be burned up in the output tubes. This will be divided by the number of tubes. At zero power output (idle) the dissipation is dependent on the idle current. Most sweep tubes work well at 30 to 35 mA. This is 17.5 watts per tube at 500 volts. Peak dissipation occurs at somewhere between 1/3 and 2/3 of maximum power, where the efficiency can be 60%. Fortunately the amp doesn't spend much time in this range. Sweep tubes are designed to run continuously at full power in a TV set. The dissipation ratings are conservative.
So, some examples.
I get 250 WPC from a single pair of 6LR6's with 2500 ohms on 650 volts. This IS pushing things a bit too far.
I and several other builders have built a hot rodded version of Pete Milletts "engineers amplifier" at the 125 WPC level. We used a 550 volt supply, a 3300 ohm OPT and one pair of 6HJ5 tubes per channel. I have used mine for a guitar amp, and cranked it loud for hours. We all used a surplus power transformer we got for $15 on Ebay. It gets rather warm with this abuse, but works OK, and runs cool for "normal" use. 4 tubes per channel would deliver 250 WPC on 550 volts with a 1650 ohm load. They should be OK for 200 WPC on 500 volts with a 1500 ohm load.
A few of us have been testing different tubes using Pete's new driver board. I have seen over 100 watts from a single pair of 13GB5's. 150 watts just brought the onset of plate glow. This was using a 3300 ohm OPT on 600 volts. I am building an amp using these cheap little tubes at the 100 WPC level on about 525 volts. It will be used for guitar amp duty. Two pair should do 150 to 200 WPC without a problem.
Saturation sounds bad on the typical low or mid powered amp. Core saturation on an amp of this magnitude can cause rapid failure. An amp capable of 200 WPC will need a power supply and output tubes capable of an amp of peak current. When the OPT saturates its inductance drops to near zero. This puts the fully conducting output tube and the OPT's DCR directly across the power supply. Something has got to give!
I have been experimenting with big powered amps for a while, and have come to one conclusion. SWEEP TUBES! I have a set of 1250 ohm Plitron OPT's rated for 400 Watts from 20 Hz to 20 KHz. They were intended for use with 6 or 8 KT88's per channel. I tried 6 EH KT88's (all I have) and it works, but I liked the way sweep tubes work better. Why, the sweep tube can pull the plate down to near zero, while the KT88's strain to make 50 volts. This makes the sweep tubes more efficient, and capable of more power from a given B+ than any of the typical audio tubes. They have a more dynamic sound too.
So, which sweep tube????? Answer, it doesn't make that much of a difference. For this power level you will need 500 to 550 volts for the plates of the output tubes, 150 to 250 for the screens depending on the tubes, and power for the drivers depending on what you use.
Any of the real sweep tubes capable of this power have maximum screen grid ratings of 175 to 275 volts. I use a mosfet regulator and adjust the voltage for best efficiency. This makes UL impossible.
Case 1) If you intend to listen to music even at the edge of clipping, the AVERAGE power level will be less than 20 watts, due to the "crest factor" of most music. The PEAK power can be 150 to 250 watts.
Case 2) If you might be tempted to say plug in a guitar or bass, crank beyond 11 and rip out a 12 minute solo, then the AVERAGE power might be closer to 50 watts, and the amp must be expected to run in clipping for as much as 50% of the time. Peak power can be 300 watts of more.
Case 3) If you are going to perform continuous sine wave testing for long periods of time the average power, can be equal to the peak power and will be whatever you design to, say 200 watts. This is worse case, but not realistic in the real world.
Bob Carver made his career designing amplifiers for the first case. I had one of his early 400 watt "magnetic field amplifier." The manual explained the design criteria, and warned the user not to use it for any use other than home HiFi. Of course I plugged my guitar into it, and it would play quite loudly, but shut off completely for several seconds if you hit clipping hard enough.
I design and test ALL of my amplifiers for the second case unless I know it will be used harder. The third case is rarely needed even for a dance floor subwoofer. This decision is needed to determine the number of tubes, and the power supply requirements.
You also need to look at your loudspeakers. If your OPT's are rated for an 8 ohm load, they will reflect 1500 ohms to the tubes only if the load is exactly 8 ohms. If your speakers have significant dips below 8 ohms, especially in the bass region, then the tubes will see less than 1500 ohms. This drives the peak tube current up. Each tube will see 1/4 the total OPT impedance, or 375 ohms. With a 550 volt supply, the peak tube current will be 1.46 AMPS! This is why you need so many KT88's. A single 6LW6 is rated for 1.4 amps, as are several big sweep tubes. I am assuming a class AB1 amp here. Class AB2 is not needed with sweep tubes, but may be needed to reach the needed power and efficiency level with conventional audio tubes.
I have found that to design for the second case the power supply can be designed to supply all the heater and driver continuously. The screen supply must be capable of supplying the sum of all the maximum screen currents continuously. Screen current is usually quite low, if the plate voltage is high. The plate supply needs to be sized at least equal to the total power output of the amp, but capable of supplying 130% of that value for brief periods (bass transients). A large low ESR capacitor can provide the reserve capacity for peaks. The average power value (= to power output) can be adjusted up or down depending on use. So for a 200 WPC amp using a 500 volt supply, you need 800 mA. This is quite conservative and the power transformer will run cool even if the amp is played loud for hours.
Now, back to the tubes.....which ones and how many? Any true sweep tube can eat a 500 volt supply without a problem. There are two criteria, peak current and dissipation. One pair of BIG, 35 or 40 watt sweep tubes can hit 1.5 amps, which may be enough. Two pair of 22 to 30 watt sweep tubes should have plenty of peak current capability, and 2 or 3 pair of small 17 watt tubes may be OK.
The efficiency of a typical sweep tube amp can be around 80% AT THE ONSET OF CLIPPING. This means that at 200 watts output, the power supply will furnish 250 watts. 200 will be delivered to the load, and 50 watts will be burned up in the output tubes. This will be divided by the number of tubes. At zero power output (idle) the dissipation is dependent on the idle current. Most sweep tubes work well at 30 to 35 mA. This is 17.5 watts per tube at 500 volts. Peak dissipation occurs at somewhere between 1/3 and 2/3 of maximum power, where the efficiency can be 60%. Fortunately the amp doesn't spend much time in this range. Sweep tubes are designed to run continuously at full power in a TV set. The dissipation ratings are conservative.
So, some examples.
I get 250 WPC from a single pair of 6LR6's with 2500 ohms on 650 volts. This IS pushing things a bit too far.
I and several other builders have built a hot rodded version of Pete Milletts "engineers amplifier" at the 125 WPC level. We used a 550 volt supply, a 3300 ohm OPT and one pair of 6HJ5 tubes per channel. I have used mine for a guitar amp, and cranked it loud for hours. We all used a surplus power transformer we got for $15 on Ebay. It gets rather warm with this abuse, but works OK, and runs cool for "normal" use. 4 tubes per channel would deliver 250 WPC on 550 volts with a 1650 ohm load. They should be OK for 200 WPC on 500 volts with a 1500 ohm load.
A few of us have been testing different tubes using Pete's new driver board. I have seen over 100 watts from a single pair of 13GB5's. 150 watts just brought the onset of plate glow. This was using a 3300 ohm OPT on 600 volts. I am building an amp using these cheap little tubes at the 100 WPC level on about 525 volts. It will be used for guitar amp duty. Two pair should do 150 to 200 WPC without a problem.
Hey, I'd agree that sweep tubes are an option. But likely you'll have to go with some with odd filament voltages, as most of the prime sweep tubes have long ago been sucked up for other uses...
They are capable of significant short term peak currents.
Not long term sine wave testing at the power levels that they can do on peak. This is why they were very popular for cheap 1kW class SSB (ham radio & illegal CB) linear amplifiers. 4 tubes would do the 1kW PEP.
If you go with sweep tubes, go for spares right away, they'll only get harder to find over time. They are rugged and last a long time though.
Otoh, the 6C33C has a very low plate Z and handles a boatload of current. Not sure how much power in AB2 can be squeezed out. But it likely will not run out of current...
I have seen 4 x 807 with 750v on the plate in AB2 do 200 watts... but not with a 1500 plate impedance...
_-_-
They are capable of significant short term peak currents.
Not long term sine wave testing at the power levels that they can do on peak. This is why they were very popular for cheap 1kW class SSB (ham radio & illegal CB) linear amplifiers. 4 tubes would do the 1kW PEP.
If you go with sweep tubes, go for spares right away, they'll only get harder to find over time. They are rugged and last a long time though.
Otoh, the 6C33C has a very low plate Z and handles a boatload of current. Not sure how much power in AB2 can be squeezed out. But it likely will not run out of current...
I have seen 4 x 807 with 750v on the plate in AB2 do 200 watts... but not with a 1500 plate impedance...
_-_-
The big advantage of sweep tubes is the high gm and thus low driving voltage.
Then it is possible to build a 200 W output stage with a simple cathodyne phase splitter.
My experience is that UL is not offering any remarkable advantage compared to pentode connection.
UL requires essentially bigger driving level and still produces lower maximum output power.
Pentode connected output stage can be designed to be as linear as UL.
The higher output impedance of pentode stage can be easily lowered with GNFB.
Then it is possible to build a 200 W output stage with a simple cathodyne phase splitter.
My experience is that UL is not offering any remarkable advantage compared to pentode connection.
UL requires essentially bigger driving level and still produces lower maximum output power.
Pentode connected output stage can be designed to be as linear as UL.
The higher output impedance of pentode stage can be easily lowered with GNFB.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.