I've come into some 6883B tubes (12V 6146B), so I decided to make a pair of monoblocks using them as a output tube. From the spec sheet, looks like one can get close to 100 watts from a pair of them.
To ensure reliability, will go with a separate screen supply at slightly under 250V. Plan to make plate voltage around 600 volts. Will get transformers from Heyboer for power and output. Reckon a 100 watt at 6.6K CT should do the trick. Going to use 6BL7 tubes for the input/phase splitter.
Anyone else try making an amp with these? Looking for some feedback/input on this. The 6883B seems like a good candidate for output tube usage. It can provide as much power as a 6550, and since they are NOS, no worries about questionable quality and high prices of current 6550's.
To ensure reliability, will go with a separate screen supply at slightly under 250V. Plan to make plate voltage around 600 volts. Will get transformers from Heyboer for power and output. Reckon a 100 watt at 6.6K CT should do the trick. Going to use 6BL7 tubes for the input/phase splitter.
Anyone else try making an amp with these? Looking for some feedback/input on this. The 6883B seems like a good candidate for output tube usage. It can provide as much power as a 6550, and since they are NOS, no worries about questionable quality and high prices of current 6550's.
These are pretty much the same tubes. I have seen some tubes with both numbers on them.
http://www.diyaudio.com/forums/tubes-valves/259279-8552-finals.html
The 6146 types can on paper provide as much power as a 6550. This may be possible in a HiFi amp where full power peaks are rare and infrequent. This is not the case if the tubes are expected to produce full power continuously.
The original Ampeg SVT guitar amp used 6 X 6146's to produce 300+ watts. Most of these amps suffered a flaming death. The amp was redesigned to use 6 X 6550 and they are plenty of them still playing today. Ampeg violated the screen grid dissipation in the 6146 causing tube arcs and blown OPT's.
I have squeezed nearly 100 watts from a pair of 6146'a in Pete Millett's Engineers Amp. Treat the 6146's like a sweep tube, and they will live. Feed the screens 400 volts and they fry.
http://www.diyaudio.com/forums/tubes-valves/259279-8552-finals.html
The 6146 types can on paper provide as much power as a 6550. This may be possible in a HiFi amp where full power peaks are rare and infrequent. This is not the case if the tubes are expected to produce full power continuously.
The original Ampeg SVT guitar amp used 6 X 6146's to produce 300+ watts. Most of these amps suffered a flaming death. The amp was redesigned to use 6 X 6550 and they are plenty of them still playing today. Ampeg violated the screen grid dissipation in the 6146 causing tube arcs and blown OPT's.
I have squeezed nearly 100 watts from a pair of 6146'a in Pete Millett's Engineers Amp. Treat the 6146's like a sweep tube, and they will live. Feed the screens 400 volts and they fry.
Thanks for the feedback. The plan is to employ a separate supply for the screens, which is going to be no more than 250 volts. I have a pair of quad 1625 amps that uses much the same approach, and they sound outstanding.
What would you recommend setting the bias for each tube?
Last edited:
Yes, limit the screen dissipation at all costs.
The only problem is that if you jack the plate voltage up, the screens are sitting somewhat low... so not quite the same as a sweep tube, I guess?
Thing is that you could take a quad of 1625 (807s) and get (iirc) almost 200 watts out of them, I think the plate voltage used was 700vdc. I had a schematic of an italian design that did this driving them in AB2.
There's a reason they have not been popular as an audio tube...
I use 3 x 6146Bs in my Johnson Viking Valiant II transmitter! 😀
(that's ~75watt carrier with >100% modulation on top of that - pair of 6DQ5s in the modulator)
The only problem is that if you jack the plate voltage up, the screens are sitting somewhat low... so not quite the same as a sweep tube, I guess?
Thing is that you could take a quad of 1625 (807s) and get (iirc) almost 200 watts out of them, I think the plate voltage used was 700vdc. I had a schematic of an italian design that did this driving them in AB2.
There's a reason they have not been popular as an audio tube...
I use 3 x 6146Bs in my Johnson Viking Valiant II transmitter! 😀
(that's ~75watt carrier with >100% modulation on top of that - pair of 6DQ5s in the modulator)
Most sweep tubes are designed to run with relatively low screen voltage compared to the plate so the recommendations are very typical of sweep tube usage. George is the leading expert in this area imo.
That said unless you've got plenty of these tubes on hand it might just make sense to sell these on eBay - someone will want them and get some current production 6550 or KT88.
I've never heard an amp I could live with that used 6146 family tubes.. Just my opinion FWIW, and YMMV.
That said unless you've got plenty of these tubes on hand it might just make sense to sell these on eBay - someone will want them and get some current production 6550 or KT88.
I've never heard an amp I could live with that used 6146 family tubes.. Just my opinion FWIW, and YMMV.
I wouldn't think the 6883B would be any more of a challenge than using the 1625 tubes. Previously had a 1625 quad amp made, and the only real challenge was installing plate stoppers (a suggestion from Miles Prower) to prevent amp oscillation. I actually think the 1625 amp sounds better than most 6550 amps I've heard or owned, especially with the current Russian/Chinese offerings.
So, I'm hoping to achieve the same results with these. I found a Acrosound schematic that gets 100 watts from a pair with 6SN7 and a T-350 transformer per side. Instead, will use a separate screen supply to keep the voltage at 250V.
According to research, the 6146B is a totally redesigned tube, and is reported to be much more reliable than the 6146 or 6146B. Some transmitters had issues with the tube, but do not see a problem with using this in a audio application.
Time will tell. 😀
The 6146B was redesigned to improve mechanical stability. The 6146 types and their little brother, the 2E26 were the "go to" tubes for most of the low band (30 to 50 MHz) and VHF (136 to 174 MHz) mobile radios made with glass octal tubes. Their life expectancy is say a taxi cab in NYC needed some improvement.
I wired some 6146's (not "B's") into Pete's red board and played with them for a few days. I had the board set up with variable power supplies for everything. I had the ability to go up to 550 volts on the screens, 650 volts on the plates (1.7 amp capabilities) and -100 volts of negative bias. I had OPT's for 1250, 1650, 2500, 3300, 5000, and 6600 ohms. I ran probably 50 different types of tubes through that setup over a year long period.
The setup and all of Tubelab was recently packed up and moved 1200 miles and will not be back together for nearly a year. I don't even know which box all of my notes are in right now.
The procedure for all tubes was the same. Pick a load impedance, start at a plate voltage that I think will work for that tube and load (start low and work your way up). I pick a bias that puts the tubes at 50 to 60% of max dissipation at idle, and turn down the current (more negative bias) as the plate voltage is increased to keep the dissipation constant. I could also adjust the screen voltage, and output tube plate to driver plate feedback resistors.
The object was to find a set of operating conditions that extracted the most power from a given set of tubes with good distortion characteristics. I wanted very low overall THD at low power levels (tenths of a percent below 1 watt, and under one percent at 10 watts mostly 2H and 3H, avoiding 5H and higher). The distortion must increase monotonically up to clipping with more 3H than any other harmonic. Output power was the power level where the distortion level rose to 5% with 5H below 1%. I wrote down the numbers for each load impedance and plate voltages in 50 volt increments in a notebook somewhere.
I tested tubes that produced anywhere from 25 watts (6GF5) per pair to 250 watts (6LW6). I don't remember the details on the 6146, but I would start with 450 volts or so and 5000 ohms, or 550 volts and 6600 ohms Set the idle current for about 12 to 15 watts per tube, and tweak from there. Most sweep tubes likes 150 to 175 screen volts. There were a few exceptions and the 6146 might have been one, I don't remember.
The 1625, 807 and 6BG6 were ALL derived from the 6L6GA and contain 6L6 type internals. They topped out at about 50 MHz. The 6146 was designed be a transmitting tube capable of reaching 175 MHz. It was the basis for several later tubes including a couple of Compactrons.
I wired some 6146's (not "B's") into Pete's red board and played with them for a few days. I had the board set up with variable power supplies for everything. I had the ability to go up to 550 volts on the screens, 650 volts on the plates (1.7 amp capabilities) and -100 volts of negative bias. I had OPT's for 1250, 1650, 2500, 3300, 5000, and 6600 ohms. I ran probably 50 different types of tubes through that setup over a year long period.
The setup and all of Tubelab was recently packed up and moved 1200 miles and will not be back together for nearly a year. I don't even know which box all of my notes are in right now.
The procedure for all tubes was the same. Pick a load impedance, start at a plate voltage that I think will work for that tube and load (start low and work your way up). I pick a bias that puts the tubes at 50 to 60% of max dissipation at idle, and turn down the current (more negative bias) as the plate voltage is increased to keep the dissipation constant. I could also adjust the screen voltage, and output tube plate to driver plate feedback resistors.
The object was to find a set of operating conditions that extracted the most power from a given set of tubes with good distortion characteristics. I wanted very low overall THD at low power levels (tenths of a percent below 1 watt, and under one percent at 10 watts mostly 2H and 3H, avoiding 5H and higher). The distortion must increase monotonically up to clipping with more 3H than any other harmonic. Output power was the power level where the distortion level rose to 5% with 5H below 1%. I wrote down the numbers for each load impedance and plate voltages in 50 volt increments in a notebook somewhere.
I tested tubes that produced anywhere from 25 watts (6GF5) per pair to 250 watts (6LW6). I don't remember the details on the 6146, but I would start with 450 volts or so and 5000 ohms, or 550 volts and 6600 ohms Set the idle current for about 12 to 15 watts per tube, and tweak from there. Most sweep tubes likes 150 to 175 screen volts. There were a few exceptions and the 6146 might have been one, I don't remember.
The 1625, 807 and 6BG6 were ALL derived from the 6L6GA and contain 6L6 type internals. They topped out at about 50 MHz. The 6146 was designed be a transmitting tube capable of reaching 175 MHz. It was the basis for several later tubes including a couple of Compactrons.
Thanks for the feedback. Some great insight provided. Will be sure to post updates as this chugs along.
I'm probably wrong, but did not the 807 pre-date the 6L6?
Or did they make the 807 for WWII service?
_-_-
Or did they make the 807 for WWII service?
_-_-
I had understood perhaps incorrectly that the 6L6 actually came first, and the 807 was developed for RF applications and applications where high plate voltages were required, and probably shortly before WWII. (1940 or so)
Progress in tube design was very rapid in the 1930s, the metal envelope versions of octal tubes started to appear around 1936.. (I believe the original 6L6 appeared in 1936) - equivalent G and GT versions appeared within a few years as the metal types proved rather expensive to make.. Loctals as well from Sylvania..
More from right here at diyA: http://www.diyaudio.com/forums/blogs/miles-prower/952-featured-vacuum-tube-807.html
Progress in tube design was very rapid in the 1930s, the metal envelope versions of octal tubes started to appear around 1936.. (I believe the original 6L6 appeared in 1936) - equivalent G and GT versions appeared within a few years as the metal types proved rather expensive to make.. Loctals as well from Sylvania..
More from right here at diyA: http://www.diyaudio.com/forums/blogs/miles-prower/952-featured-vacuum-tube-807.html
The 6L6 metal tube appeared to the public in 1939. Some designers may have gotten prototypes earlier. The glass 6L6G came late in 1939 as well.
As war approached a new tube was needed for HF (2 to 30 MHz) transmitters. The 6L6GA had too much capacitance, needed some better shielding, and more space between the plate and grid pins, so the 807 and 1625 were made from 6L6GA parts. Two TV sweep tubes were also created with 6L6GA guts too, the 6BG6G and the 6BD5.
As war approached a new tube was needed for HF (2 to 30 MHz) transmitters. The 6L6GA had too much capacitance, needed some better shielding, and more space between the plate and grid pins, so the 807 and 1625 were made from 6L6GA parts. Two TV sweep tubes were also created with 6L6GA guts too, the 6BG6G and the 6BD5.
The 807/1625 tubes are a great buy, and make a excellent audio tube. I'm hoping to have the same level of success with the 6883B tube.
Well, it took awhile, but amps are finished, and sounding mighty good. My buddy built these for me, as he does outstanding work.
I'm very impressed with the sound from these. All power and screen supplies have individual windings and regulation/filtering, The results are that the amp exhibits outstanding low level detail and dynamics, along with being dead quiet. This is the second amp I have had made that sports pentode operation, and I have to say that the sonic character from both of them seems subjectively better than the ultra-linear amps I have owned over the years.
The 6883B makes a excellent audio output tube, and the 6BL7 is nice input/phase splitter tube.
Got the idea for this from the Acrosound 1955 manual, which has a 90 watt 6146 schematic. The mastero amp from 1951 also sports a 100 watt 6146 amp design.
I would much rather use a tube like the 6883B over current 6550/KT88 tubes, for a variety of obvious reasons.
I'm very impressed with the sound from these. All power and screen supplies have individual windings and regulation/filtering, The results are that the amp exhibits outstanding low level detail and dynamics, along with being dead quiet. This is the second amp I have had made that sports pentode operation, and I have to say that the sonic character from both of them seems subjectively better than the ultra-linear amps I have owned over the years.
The 6883B makes a excellent audio output tube, and the 6BL7 is nice input/phase splitter tube.
Got the idea for this from the Acrosound 1955 manual, which has a 90 watt 6146 schematic. The mastero amp from 1951 also sports a 100 watt 6146 amp design.
I would much rather use a tube like the 6883B over current 6550/KT88 tubes, for a variety of obvious reasons.
Attachments
Last edited:
The amp continues to improve as they break in. Used Jupiter copper bypass caps, which sound pretty good once they get some hours on them. They sound a little closed in for the first 20 to 30 hours, at which point they start to open up nicely.
Anyone else make an amp with 6146/6883 output tubes?
Anyone else make an amp with 6146/6883 output tubes?
I made a triode connected 6146 amp; but my Driver circuit ran out of drive, {used 5687}even with a higher voltage for the drivers
The 6146 as a triode puts out a lot of power at 400 volts going by the Tube data from memory 25 watts.
Do you have a circuit of your amp? maybe the 6bl7 could drive them.
Phil
The 6146 as a triode puts out a lot of power at 400 volts going by the Tube data from memory 25 watts.
Do you have a circuit of your amp? maybe the 6bl7 could drive them.
Phil
Sorry, the specific front end is proprietary, provided by a industry professional for private use only. 🙁
I can tell you that the front end is pretty close to the Mastero amp from the early 50's, and the power stage is close to the Acrosound TO-350 schematic. There is a MOS-FET attached to the first driver tube to provide more drive to the phase splitter. The plates have 600V, fixed bias with separate power supplies for the bias and screens.
I can tell you that the front end is pretty close to the Mastero amp from the early 50's, and the power stage is close to the Acrosound TO-350 schematic. There is a MOS-FET attached to the first driver tube to provide more drive to the phase splitter. The plates have 600V, fixed bias with separate power supplies for the bias and screens.
6L6 came first, as one of the first beamers to beat the Phillips patent on power pentodes. The original 6L6s used the then new Octal base and a metal envelope, and was intended as an audio power final. As such, it put the highest voltage pin right next to the lowest, making for a nasty flash-over problem that limited its plate voltage, but that wasn't a problem for AF duty anyway. For sweep tube and RF duty, it's a bit of a problem, though 807s were used by Dumont for TV HD duty when B & W screens were small, and HD deflection coils not so current hungry.
The 807 and its variants came later as an RF power final in the then standard five pin format, and a plate cap connection to get the high voltage away from low voltage pins. There's an added shield cage, and it can stand up to higher voltages for more Class B/C RF power. The 6L6-oids (807, 1625, 6BG6, and probably others) can sound very good, but need a bit of extra work: local NFB (parallel or cathode) and the necessary gNFB to take the remaining edge off. They do tend to produce more of those nasty, higher order, harmonics, and need more NFB to clean 'em up, but once that's done, they perform with sonic excellence. The only other thing you need to look out for is snivets. A 1K5 screen stopper should prevent that.
Beam Power Tubes / Schade
STC 807 Application Report
Lots of good information to be found here.
Some great stuff, indeed.
I had these made in late 2013. They sound great. Output power is 110 watts for each mono block.
BTW, I had oscillation problems with them, and found the fix from a post you made here in DIY Audio:
" Yeah. 807s are notorious for this. What you have there is a damped RF oscillation that's being triggered by plate current cutoff (Barkhausen oscillation). I had the same thing happen with a project that used 807s. What you need is a really good plate stopper, and screen stoppers.
In this design, the plate stoppers (L1, R31 & L2, R32) consist of ten turns of #18 wire; ID= 7/16"; space wound. Install 100R, 2W, C-comp resistors inside the coil. Install right at the plate connector. (If you can't get 100R / 2W C-comps, use four, 470R / 0.5W C-comps in parallel.)
Install 1K5 / 0.5W C-comp or C-film grid stoppers right at the screen grid connection at the socket.
This should get the final stable."
Thanks!
Thought I would post an update about this project. The amps sound great. The 6883B makes an excellent audio output tube. I dropped the output bias to 30ma to maintain some longevity with the tubes. Had a couple with the same 80's date code go bad early. However, between dropping the bias and not using that specific date code, the amps have been reliable and sound great.
One other tweak that was done was to change the coupling caps to Jupiter Copper Foils. Made a big difference with transparency and clarity with the midrange and treble.
I had read in one of the old Vacuum Tube Valley articles that the 6146 was originally the high powered tube that was going to be used for audio, but the cost to make the tube is what drove the design of the 6550. They were looking for a high power low cost alternative.
One other tweak that was done was to change the coupling caps to Jupiter Copper Foils. Made a big difference with transparency and clarity with the midrange and treble.
I had read in one of the old Vacuum Tube Valley articles that the 6146 was originally the high powered tube that was going to be used for audio, but the cost to make the tube is what drove the design of the 6550. They were looking for a high power low cost alternative.
- Status
- Not open for further replies.