KT77 isn't too shabby in triode mode. Beats EL34.
https://www.audiomatica.com/tubes/kt77.htmI've tested some JJ KT77 on the curve tracer, and they seem to match the triode curves shown.
https://www.audiomatica.com/tubes/kt77.htmI've tested some JJ KT77 on the curve tracer, and they seem to match the triode curves shown.
smoking-amp, this datasheets appears to be a copy of an EL34's. Even the picture clearly shows a real pentode instead of a beam power tube, hence resembles an EL34..
Best regards!
Best regards!
Indeed! The tube pics look like similar pentode internals. I wonder what goes on there. Looks like it is marked NATIONAL on the KT77 tube, with Made in Germany marked on it.
I do see a difference in the pin-out diagrams between the two. I checked the JJ KT77 here and they are beamers as they are supposed to be. The triode curves on the curve tracer here look like the ones shown. Which aren't too bad. The KT77 triode curves are what attracted me to get some to trace .
I do see a difference in the pin-out diagrams between the two. I checked the JJ KT77 here and they are beamers as they are supposed to be. The triode curves on the curve tracer here look like the ones shown. Which aren't too bad. The KT77 triode curves are what attracted me to get some to trace .
Last edited:
What about PP GU50 in triode? Not the greatest triode ever, but relatively inexpensive and 40W anode dissipation.
Here is another source indicating not the greatest triode curves for the EL34.
http://www.dissident-audio.com/Yves/EL34-Trioded.gif Looks like it came from a datasheet, and it looks to agree with the earlier Audiomatica EL34 curves (different vertical current scales).
The JJ KT77 curves I see on the curve tracer here agree with the earlier Audiomatica KT77 curves. The kind of triode curves I like for $19.
Although certainly not up to 300B or 2A3 curves. $$$
I have a tube emulation scheme for those kind of curves. A 12HL7 driver to a power tube and N Fdbk to the 12HL7 screen grid, makes the power tube emulate the 12HL7 internal triode: ( for $1 )
http://www.dissident-audio.com/Yves/EL34-Trioded.gif Looks like it came from a datasheet, and it looks to agree with the earlier Audiomatica EL34 curves (different vertical current scales).
The JJ KT77 curves I see on the curve tracer here agree with the earlier Audiomatica KT77 curves. The kind of triode curves I like for $19.
Although certainly not up to 300B or 2A3 curves. $$$
I have a tube emulation scheme for those kind of curves. A 12HL7 driver to a power tube and N Fdbk to the 12HL7 screen grid, makes the power tube emulate the 12HL7 internal triode: ( for $1 )
Last edited:
When talking about amplifier power output, do not forget to take into account the insertion loss of the output transformer.
Some manufacturers actually rate the insertion loss of their output transformers.
Some of those are telling the truth; others of those are telling marketing lies.
I have carefully measured a lot of output transformers. Many of them are closer to 1dB insertion loss, than they are to 0.5dB insertion loss.
1 dB of insertion loss will take 20 Watts from the output tubes, and 'convert it' to only 15.9 Watts.
Surprised?
Example: Consider a 5k Ohm to 8 Ohm single ended transformer, which has 250 Ohms primary DCR, and 0.4 Ohms secondary DCR.
That transformer has 1 dB of insertion loss!
Surprised?
The same principles apply to push pull transformers. DCR eats up power.
Comments?
Some manufacturers actually rate the insertion loss of their output transformers.
Some of those are telling the truth; others of those are telling marketing lies.
I have carefully measured a lot of output transformers. Many of them are closer to 1dB insertion loss, than they are to 0.5dB insertion loss.
1 dB of insertion loss will take 20 Watts from the output tubes, and 'convert it' to only 15.9 Watts.
Surprised?
Example: Consider a 5k Ohm to 8 Ohm single ended transformer, which has 250 Ohms primary DCR, and 0.4 Ohms secondary DCR.
That transformer has 1 dB of insertion loss!
Surprised?
The same principles apply to push pull transformers. DCR eats up power.
Comments?
Last edited:
6336A I love that tube. ordered one (£65 all in) on ebay from the states, it made it halfway across the country and then made its way back to the seller. full refund offered no explanation, but I got the feeling I was trying to export some thing that I shouldn't.
I did start rebuilding a Quad ii with 807s once. 211 845 a bit scary volts for this build, and I have a 845/GM70 project on the go already.
I have tried triode connection on EL34 and GU50s, Wasn't convinced by the GU50, 34 sounded pretty sweet though
I think I am sold on the KT77 idea at the moment.(£49 for 4 from UK seller 🙂 and they are the right height to fit my box.
I did start rebuilding a Quad ii with 807s once. 211 845 a bit scary volts for this build, and I have a 845/GM70 project on the go already.
I have tried triode connection on EL34 and GU50s, Wasn't convinced by the GU50, 34 sounded pretty sweet though
I think I am sold on the KT77 idea at the moment.(£49 for 4 from UK seller 🙂 and they are the right height to fit my box.
That's a nice box - looks like a 19" subrack, 3U or 4U. Very useful kind of chassis for tube amps. I use 19" subrack parts for all my builds.
I saw that deal on four JJ KT77s on ebay.uk. Nice price!
I saw that deal on four JJ KT77s on ebay.uk. Nice price!
Weird about the 6336A shipping problem. Maybe the seller wrote "JAN xxxx" in big letters on the box or shipping label. Can't imagine an old V Reg. tube being an issue though. I plan to try one out in the tube emulator scheme, give it near perfect 12HL7 curves. Turn it into a low Voltage, high current 211.
The high 2nd Harmonic (EL34) likely gives it a nice "sweet" sound in SE use. That would all disappear in a P-P setup though. Lower 2nd H in the KT77 curves should also give it lower higher harmonics as well in P-P. It will be a different sounding amp than a SE one.
The high 2nd Harmonic (EL34) likely gives it a nice "sweet" sound in SE use. That would all disappear in a P-P setup though. Lower 2nd H in the KT77 curves should also give it lower higher harmonics as well in P-P. It will be a different sounding amp than a SE one.
Last edited:
I have a pair of Philips EL6400 PA amps rebuilt as hifi monoblocks. MkI used PL36 in class AB1 (Ultralinear IIRC), MkII had triode wired 6P21S pushed into class A2 @ 11W by a balanced, choke loaded 6N6P cathode follower a´la Altec 1570. Don´t know about filament longevity, I could´nt get rid of the hum so I did som minor rewiring and changed to 6L6G output tubes. OPT ~5,6k p-p if I remember correctly.
With EL34 and KT77 being essentially interchangeable, one could try the P-P Amp with EL34 on one side and KT77 on the other side. Might give a little SE sound effect (in class A triode mode), if there is no additional N Fdbk circuitry.
Letting the KT77 run in pentode mode, and the el34 run in triode mode should give an even stronger SE effect (adjusting bias's for DC balance). (or KT77 in triode and EL34 in pentode or KT77 triode with KT77 pentode .... you need some individual triode/pentode/UL switches for the output tubes!!)
Maybe some suggestions around here on which EL34 brand is best for SE use?
Letting the KT77 run in pentode mode, and the el34 run in triode mode should give an even stronger SE effect (adjusting bias's for DC balance). (or KT77 in triode and EL34 in pentode or KT77 triode with KT77 pentode .... you need some individual triode/pentode/UL switches for the output tubes!!)
Maybe some suggestions around here on which EL34 brand is best for SE use?
Last edited:
It is extremely hard to run a KT77 in pentode mode . . .
The KT77 is a Kinkless Tetrode, otherwise known as a Beam Power Tube. Think: Beam Formers.
An EL34 is a Pentode. Think: Real Suppressor Grid.
Both KT77, and EL34 will run well in Ultra Linear mode, and will run well in Triode Wired mode.
Just Saying . . .
I often mis-type or mis-describe my answers too.
The KT77 is a Kinkless Tetrode, otherwise known as a Beam Power Tube. Think: Beam Formers.
An EL34 is a Pentode. Think: Real Suppressor Grid.
Both KT77, and EL34 will run well in Ultra Linear mode, and will run well in Triode Wired mode.
Just Saying . . .
I often mis-type or mis-describe my answers too.
Oh, of course. I was mainly suggesting by "pentode" mode that one output tube could function similarly (roughly) to a CCS on one P-P side, so that SE overall effects could be obtained in P-P class A. The GEC/MOV KT77 datasheet does emphasize that the tube was designed specifically for UL mode, but also works well in triode mode, and "reluctantly" states that beam mode can be used too.
A better CCS function could be obtained by also disconnecting the AC drive to that tube, Ra = 23K Ohm at 110 mA, but only half the class A P-P power output obtainable without the AC drive on one side, actually even less than half the power out with the finite 23K Ohm "CCS"
It might be possible to get even closer to a CCS function by using a -reduced- AC drive on one side (along with the DC screen V for "pentode" mode). Just enough drive to make up for the 23K Ra loading.
The rounded knees of the KT77 curves make me think that this tube has a larger than usual width slot in the beam plate, so as to get intermediate results between real pentode and beam pentode/tetrode. It certainly does not have the sharp knees seen on most beam pentode/tetrodes. Check out some TV "sweeps" if you want to see "square" knees for maximum switching efficiency.
A better CCS function could be obtained by also disconnecting the AC drive to that tube, Ra = 23K Ohm at 110 mA, but only half the class A P-P power output obtainable without the AC drive on one side, actually even less than half the power out with the finite 23K Ohm "CCS"
It might be possible to get even closer to a CCS function by using a -reduced- AC drive on one side (along with the DC screen V for "pentode" mode). Just enough drive to make up for the 23K Ra loading.
The rounded knees of the KT77 curves make me think that this tube has a larger than usual width slot in the beam plate, so as to get intermediate results between real pentode and beam pentode/tetrode. It certainly does not have the sharp knees seen on most beam pentode/tetrodes. Check out some TV "sweeps" if you want to see "square" knees for maximum switching efficiency.
Last edited:
@ 6A3summer. I am at the moment more of a trial and error when it comes to conceptualising. My inner rule of thumb for push pull with pentodes is power output should be about half the sum of the anodes dissipation, all else being equal and well sorted, for A1 operation. I guess that doesn't work so well for triode.
(several hours later)
Am thinking about output transformers now (for those that may have missed it, I am going to use KT77s), and a very happy thing has happened 🙂 🙂 🙂 it is fairly safe to assume what I will be building is in effect a triode wired EL34 push pull amplifier, not that revolutionary, and almost 100 years of prior art to plunder.
Even more exciting EL34 data sheet suggests 430v and a 10 k anode to anode TX in triode mode for 14W and less than 1% distortion. this sounds thoroughly worthy, low distortion is a fetish of mine at the moment.
http://www.r-type.org/pdfs/el34.pdf
I even better I have a 20 watt el84 transformer, as well as the one pictured above which will be lower z, in my stash, so am pretty confident to start cutting holes 🙂
(several hours later)
Am thinking about output transformers now (for those that may have missed it, I am going to use KT77s), and a very happy thing has happened 🙂 🙂 🙂 it is fairly safe to assume what I will be building is in effect a triode wired EL34 push pull amplifier, not that revolutionary, and almost 100 years of prior art to plunder.
Even more exciting EL34 data sheet suggests 430v and a 10 k anode to anode TX in triode mode for 14W and less than 1% distortion. this sounds thoroughly worthy, low distortion is a fetish of mine at the moment.
http://www.r-type.org/pdfs/el34.pdf
I even better I have a 20 watt el84 transformer, as well as the one pictured above which will be lower z, in my stash, so am pretty confident to start cutting holes 🙂
smoking-amp,
Good points!
Sorry I mis-read what you were suggesting in your Post # 34.
1. The best way to build a CCS EL34 or KT77 is with an un-bypassed self bias resistor, and Pentode mode, or Kinkless Tetrode/Beam Power mode.
Just match the plate current to the other tube that does have signal applied to it.
Then, the CCS tube can either be without signal applied to it, or it can have signal applied to it. Either way, there will be some 2nd harmonic distortion (and some 2nd order Intermodulation distortion too). The "gain" of push and pull will be unequal, whether the CCS tube's g1 is not driven, or if it is driven (the un-bypassed self bias resistor of the CCS will assure that).
Of course, a lot of global negative feedback will get rid of most of the 2nd order distortion that we introduced in 1. above.
So, in order to have a damping factor higher than unity, with one of the tubes in CCS mode, the other tube needs to be in either Ultra Linear mode, or in Triode Wired mode (or at least the driven tube needs some form of Schade feed back to the driver plate, or the driver cathode).
with no global negative feedback or Schade negative feedback, the damping factor would be far less than Unity (<1).
Good points!
Sorry I mis-read what you were suggesting in your Post # 34.
1. The best way to build a CCS EL34 or KT77 is with an un-bypassed self bias resistor, and Pentode mode, or Kinkless Tetrode/Beam Power mode.
Just match the plate current to the other tube that does have signal applied to it.
Then, the CCS tube can either be without signal applied to it, or it can have signal applied to it. Either way, there will be some 2nd harmonic distortion (and some 2nd order Intermodulation distortion too). The "gain" of push and pull will be unequal, whether the CCS tube's g1 is not driven, or if it is driven (the un-bypassed self bias resistor of the CCS will assure that).
Of course, a lot of global negative feedback will get rid of most of the 2nd order distortion that we introduced in 1. above.
So, in order to have a damping factor higher than unity, with one of the tubes in CCS mode, the other tube needs to be in either Ultra Linear mode, or in Triode Wired mode (or at least the driven tube needs some form of Schade feed back to the driver plate, or the driver cathode).
with no global negative feedback or Schade negative feedback, the damping factor would be far less than Unity (<1).
A pair of 6080's using cathode bias on each tube is good for 25W.
