> what are the main derivatives/equivalents of the 807?
All the main power pentodes tended to converge in the 1950s.
> 807 is a lot like 6L6. It's even closer to 1625.
Indeed the 807 and the original 6L6 are about the same cathode, similar structure, but the 807's plate will take more voltage and power. 807 is limited use (who needs anything that big?), uses a plate-cap because the plate voltage is too high for a socket, and uses an older socket. 6L6 is cut-down to use cheaper plate-metal and the new octal socket, common in mass-produced electronics.
807/6L6 solved the problem of triodes (low current at low plate voltage, improved only by reducing gain) and reduced several problems in tetrodes (high screen current, very kinky when screen voltage is below plate voltage). It revolutionized big-amp design. It also established a patent position: the Beam Electrode construction was unique. One other point: the 807/6L6 has a lot of curvature: this causes large 2nd harmonic in single-ended use, but that cancels in push-pull, and the curvature improves the power sensitivity (driver voltage needed to reach a given output power).
6V6 is very nearly half a 6L6 with cheaper plate metal. 6AQ5 is a smaller 6V6.
By comparison, the 6F6 is less curved and has higher screen current (for the same plate current). This is non-Beam Electrode technology. In single-ended, a 6F6 makes less 2nd harmonic than a 6V6 (similar size). However the real reason the 6F6 lingered (see early H-P 200AB) is that after the "better" tubes came out, 6F6 was sold cheaper.
To dodge RCA's patent on Beam Electrode, and to fight the severe kink in the knee of 6L6-like tubes, someone in Europe did the "KT" (Kinkless Tetrode) design. This does roughly the same thing somewhat differently, and made a darn good power tube without infringing on RCA.
These tubes are different in detail but do about the same thing. They have to. The transconductance is limited by cathode area which is a function of price and heater power, and by grid spacing which is a function of production precision. While small tubes eventually grew "frame grid" which gave tighter spacing, power tubes can't be delicate. So at the same current, all these tubes give about the same Gm, or Gm rises somewhat between small and big cathode. And for good power sensitivity, the Mu (to G2) has to be much more than 5, but no higher than necessary: generally about 10 for these tubes (6L6 is 8, the fattest H-sweep tubes ran as low as 5 for more pulse current at low B+, some small power tubes have Mu like 20 to simplify/cheapen the driver).
Between the 807's power, the 6L6 convenient socket, and the KT66's soft kinkless linearity, lay a "universal power tube". Around 1950 there came the 6L6GC, which has much higher plate rating and less kinking than 6L6. While it works in any 6L6 socket, it can do much more. In fact it was too universal: for marketing they had to make cheap versions and heavy-duty versions. The 6550 is a lot like a 6L6 with a bigger cathode and only made with the best plate-metal so it has a 600V rating. 5881 is a low-profile 6L6GC with lower plate ratings. 7027 seems to be a skinny 6L6GC with different pinout so you can't stick in generic 6L6GCs. EL34 seems to be an older softer power pentode (not full Beam or KT detailing) but was introduced with a very high voltage rating and could make astonishing output for its heater power (if you fed it carefully; its main use is modest power without full Beam-tube harshness). And the misbegotten 8417, a last-gasp 6550 with too-close grid so it could use a cheaper driver and stave-off the transistor invasion, and 7591 which is also close-grid easy-drive but not to the point it fails (like some 8417s did) and otherwise a small 6L6GC.
Without dredging back to 1930, or tubes under 10 watts or over 45 watts, or WE or Euro designs: I think only the 6F6 is not kin to either 6L6 or the 6L6GC/KT-series flock. 8417 and 7591 have higher Gm/Ip than the others, some of the TV sweep tubes stray further from the trunk, but most power tubes do descend from the same general design.
By comparison, 2A3 makes limited power and is hard to drive, but is more linear than 6L6 and any triode is more tolerant of elliptical (speaker) loadlines than any pentode (and also gives damping). WE-300 is even more linear, being very carefully drawn and built for audio work (WE made a very similar tube with less linearity, better for RF use). Many of the TV V-sweep triodes are also very linear, because raster distortion is moderately obvious and not easily cured with NFB (not at TV set prices).
All the main power pentodes tended to converge in the 1950s.
> 807 is a lot like 6L6. It's even closer to 1625.
Indeed the 807 and the original 6L6 are about the same cathode, similar structure, but the 807's plate will take more voltage and power. 807 is limited use (who needs anything that big?), uses a plate-cap because the plate voltage is too high for a socket, and uses an older socket. 6L6 is cut-down to use cheaper plate-metal and the new octal socket, common in mass-produced electronics.
807/6L6 solved the problem of triodes (low current at low plate voltage, improved only by reducing gain) and reduced several problems in tetrodes (high screen current, very kinky when screen voltage is below plate voltage). It revolutionized big-amp design. It also established a patent position: the Beam Electrode construction was unique. One other point: the 807/6L6 has a lot of curvature: this causes large 2nd harmonic in single-ended use, but that cancels in push-pull, and the curvature improves the power sensitivity (driver voltage needed to reach a given output power).
6V6 is very nearly half a 6L6 with cheaper plate metal. 6AQ5 is a smaller 6V6.
By comparison, the 6F6 is less curved and has higher screen current (for the same plate current). This is non-Beam Electrode technology. In single-ended, a 6F6 makes less 2nd harmonic than a 6V6 (similar size). However the real reason the 6F6 lingered (see early H-P 200AB) is that after the "better" tubes came out, 6F6 was sold cheaper.
To dodge RCA's patent on Beam Electrode, and to fight the severe kink in the knee of 6L6-like tubes, someone in Europe did the "KT" (Kinkless Tetrode) design. This does roughly the same thing somewhat differently, and made a darn good power tube without infringing on RCA.
These tubes are different in detail but do about the same thing. They have to. The transconductance is limited by cathode area which is a function of price and heater power, and by grid spacing which is a function of production precision. While small tubes eventually grew "frame grid" which gave tighter spacing, power tubes can't be delicate. So at the same current, all these tubes give about the same Gm, or Gm rises somewhat between small and big cathode. And for good power sensitivity, the Mu (to G2) has to be much more than 5, but no higher than necessary: generally about 10 for these tubes (6L6 is 8, the fattest H-sweep tubes ran as low as 5 for more pulse current at low B+, some small power tubes have Mu like 20 to simplify/cheapen the driver).
Between the 807's power, the 6L6 convenient socket, and the KT66's soft kinkless linearity, lay a "universal power tube". Around 1950 there came the 6L6GC, which has much higher plate rating and less kinking than 6L6. While it works in any 6L6 socket, it can do much more. In fact it was too universal: for marketing they had to make cheap versions and heavy-duty versions. The 6550 is a lot like a 6L6 with a bigger cathode and only made with the best plate-metal so it has a 600V rating. 5881 is a low-profile 6L6GC with lower plate ratings. 7027 seems to be a skinny 6L6GC with different pinout so you can't stick in generic 6L6GCs. EL34 seems to be an older softer power pentode (not full Beam or KT detailing) but was introduced with a very high voltage rating and could make astonishing output for its heater power (if you fed it carefully; its main use is modest power without full Beam-tube harshness). And the misbegotten 8417, a last-gasp 6550 with too-close grid so it could use a cheaper driver and stave-off the transistor invasion, and 7591 which is also close-grid easy-drive but not to the point it fails (like some 8417s did) and otherwise a small 6L6GC.
Without dredging back to 1930, or tubes under 10 watts or over 45 watts, or WE or Euro designs: I think only the 6F6 is not kin to either 6L6 or the 6L6GC/KT-series flock. 8417 and 7591 have higher Gm/Ip than the others, some of the TV sweep tubes stray further from the trunk, but most power tubes do descend from the same general design.
By comparison, 2A3 makes limited power and is hard to drive, but is more linear than 6L6 and any triode is more tolerant of elliptical (speaker) loadlines than any pentode (and also gives damping). WE-300 is even more linear, being very carefully drawn and built for audio work (WE made a very similar tube with less linearity, better for RF use). Many of the TV V-sweep triodes are also very linear, because raster distortion is moderately obvious and not easily cured with NFB (not at TV set prices).
PRR
You know your tubes!
Diyhifisupply’s Joplin www.diyhifisupply.com/diyhs_joplin.htm has the flexibility to use 2a3/ 45/ 300b. I really like the idea of being able (with a little time) to change tube types, not just “roll”.
I wonder if it’s feasible in the same way to do 2a3/ 6550/ EL84?
You know your tubes!
Diyhifisupply’s Joplin www.diyhifisupply.com/diyhs_joplin.htm has the flexibility to use 2a3/ 45/ 300b. I really like the idea of being able (with a little time) to change tube types, not just “roll”.
I wonder if it’s feasible in the same way to do 2a3/ 6550/ EL84?
For one a 2A3 or DHT's typically have one half/ one third the distortion of triode wired pentodes.
The 300b, 2A3, 45 are simply the better known..there are plenty of other DHT's that obviously fall into this category. PX25, PX4, 10, WE205D, Ba,4019A and countless others.
So what should you get? If you want more power for your money get 6550's...
If you want low distortion get a tube that was designed before WW2
Yes..both measured and audibly the DHT's sound best.
The 300b, 2A3, 45 are simply the better known..there are plenty of other DHT's that obviously fall into this category. PX25, PX4, 10, WE205D, Ba,4019A and countless others.
So what should you get? If you want more power for your money get 6550's...
If you want low distortion get a tube that was designed before WW2
No experience with the pp-6550 plenty with the el-34/6ca7
I recently built this amp.
http://www2u.biglobe.ne.jp/~tossie/F2A3PP-E.html
I replaced the 6a6 with the 6N7G I really like this amp, a few on the Joe
List recommended it, they though it was very good also.
i liked it much better than the el-34 in triode.
Tony
I recently built this amp.
http://www2u.biglobe.ne.jp/~tossie/F2A3PP-E.html
I replaced the 6a6 with the 6N7G I really like this amp, a few on the Joe
List recommended it, they though it was very good also.
i liked it much better than the el-34 in triode.
Tony
SY said:
If you're talking about distortion just before clipping in Single Ended Class A, it's easy to see from looking at the triode plate curves. The curves are obviously much more uniform for a 300B than a triode connected 6550, for example. Do you have reason to doubt that the curves tell the story?
Jeff
Yes you can get Chinese 2A3s for very low prices, but they’re not the best 2A3s. I was thinking (at that time) a better 6550 would be better than an average 2A3.
You had compared 2A3s to 6550s in *NOS*.
I compared them in what I believe is a better *current production *brand – JJ (correct me if I’m wrong):
EL84 $28 a Matched Pair.
KT88 $98 a Matched Pair.
JJ 2A3s I believe are about $320 a Matched Pair!
Are NOS generally better than eg JJ (Tesla)?
But now I hear almost convincing evidence for the DHTs
(so now leaning back to 2A3s/ 300B)
Anyone know how many watts a two a side 2A3 or 300B could generate in ultralinear?
Yes you can get Chinese 2A3s for very low prices, but they’re not the best 2A3s. I was thinking (at that time) a better 6550 would be better than an average 2A3.
You had compared 2A3s to 6550s in *NOS*.
I compared them in what I believe is a better *current production *brand – JJ (correct me if I’m wrong):
EL84 $28 a Matched Pair.
KT88 $98 a Matched Pair.
JJ 2A3s I believe are about $320 a Matched Pair!
Are NOS generally better than eg JJ (Tesla)?
But now I hear almost convincing evidence for the DHTs
(so now leaning back to 2A3s/ 300B)
Anyone know how many watts a two a side 2A3 or 300B could generate in ultralinear?
Rick57,
You can't run triodes in Ultralinear - so "sorry" but your question about 2A3 and 300B in ultralinear does'nt make sense. Ultralinear involves driving the screen grid and is applicable to pentodes and beam tetrodes.
In order for us here at the forum to better advise you (i've seen most of your info gathering / learning stuff posts) perhaps you could advise us as to what speakers you are trying to drive, and in how big a room and perhaps even what sort of music you like to listen to.
From personal experience - driving VAF Research DC-X speakers.
My home built EL84 Ultralinear Push Pull (the Morgan Jones "Bervois Valley" design, using JJ EL84s which are stunning) doesn't take a single backward step compared to my Chinese "Music Angel" 845 SET (which has a few minor upgrades and new pre and driver tubes).
I tried the EL84 Amp in Triode Mode and did'nt like it.
I have rebuilt the MJ "Bervois Valley" Amp to a new circuit - (See Hey Planet 10 and Others thread) and am in final fine tune mode. Initial impressions are that it is even better again.
The reason for the questions at para 2 above is so we can think about issues like:
My VAF DC-Xs have VERY GOOD self damping and a REASONABLY flat impedance vs frequency curve which means I can use low damping factor amps without trouble. Your speakers may be quite different.
P.S. Do you definitely want to build something or are you happy to buy something. The reason I ask is that the guy who was bring the "Music Angels" into Oz is getting out of that line of amps (since you can order them yourself from HK) and I know he has 2 off 845 SETs left in stock which he is quiting at cost. Email me if you want the details.
Cheers,
Ian
You can't run triodes in Ultralinear - so "sorry" but your question about 2A3 and 300B in ultralinear does'nt make sense. Ultralinear involves driving the screen grid and is applicable to pentodes and beam tetrodes.
In order for us here at the forum to better advise you (i've seen most of your info gathering / learning stuff posts) perhaps you could advise us as to what speakers you are trying to drive, and in how big a room and perhaps even what sort of music you like to listen to.
From personal experience - driving VAF Research DC-X speakers.
My home built EL84 Ultralinear Push Pull (the Morgan Jones "Bervois Valley" design, using JJ EL84s which are stunning) doesn't take a single backward step compared to my Chinese "Music Angel" 845 SET (which has a few minor upgrades and new pre and driver tubes).
I tried the EL84 Amp in Triode Mode and did'nt like it.
I have rebuilt the MJ "Bervois Valley" Amp to a new circuit - (See Hey Planet 10 and Others thread) and am in final fine tune mode. Initial impressions are that it is even better again.
The reason for the questions at para 2 above is so we can think about issues like:
My VAF DC-Xs have VERY GOOD self damping and a REASONABLY flat impedance vs frequency curve which means I can use low damping factor amps without trouble. Your speakers may be quite different.
P.S. Do you definitely want to build something or are you happy to buy something. The reason I ask is that the guy who was bring the "Music Angels" into Oz is getting out of that line of amps (since you can order them yourself from HK) and I know he has 2 off 845 SETs left in stock which he is quiting at cost. Email me if you want the details.
Cheers,
Ian
Ian
can't run triodes in Ultralinear Might then double the number of output valves
You ask very good questions:
Speakers & room:
When an extension is built on my house (hopefully within 6 months) in a room 5.3* 6 * 3m, a pair of
Bass - two Peerless XLS a side, driven by Hypex, with active XO at c. 150-180 Hz to:
180- 1500 Hz: JBL 2202 sealed: Nominal impedance 8, 100 dB
> 1500 Hz: Aurum Cantus 1: Impedance flat around 7, 100 dB
The tube amp is to drive the latter two.
I listen to pop/ rock/ world music, so like ‘oomph’ (dynamics) but also ‘clarity’.
The idea of Ultralinear was so that I could also use the amp on other less sensitive speakers if I wanted, eg the mid panels on the SL Phoenix which is under – sensitivity not stated, but meant to be highish, though output impedance of the amp should be <0.4 ohm.
(I ordered the Morgan Jones book today)
How is self damping estimated or measured?
The "Music Angel”/ Rarefaction are probably really good value, but I’m after a power amp, with better quality parts.
can't run triodes in Ultralinear
Might then double the number of output valves You ask very good questions:
Speakers & room:
When an extension is built on my house (hopefully within 6 months) in a room 5.3* 6 * 3m, a pair of
Bass - two Peerless XLS a side, driven by Hypex, with active XO at c. 150-180 Hz to:
180- 1500 Hz: JBL 2202 sealed: Nominal impedance 8, 100 dB
> 1500 Hz: Aurum Cantus 1: Impedance flat around 7, 100 dB
The tube amp is to drive the latter two. I listen to pop/ rock/ world music, so like ‘oomph’ (dynamics) but also ‘clarity’.
The idea of Ultralinear was so that I could also use the amp on other less sensitive speakers if I wanted, eg the mid panels on the SL Phoenix which is under – sensitivity not stated, but meant to be highish, though output impedance of the amp should be <0.4 ohm.
(I ordered the Morgan Jones book today)
How is self damping estimated or measured?
The "Music Angel”/ Rarefaction are probably really good value, but I’m after a power amp, with better quality parts.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.