Some 20 years ago in order to mimic a mesh anode I thought the parallel wires of g2 could be used. This worked right away and sounded very good. A friend of mine gave the schematic to Jonathan Weiss who made a commercial version of it under the name parallax amp which still can be found on his oma- website.
The g2 dissipation can be 12 watts without getting red. At the moment I use an ecc83; Rk becomes 2k2 and everything else stays the same. I wanted a little more gain as mu g2g1 of the 519 is only about 3,6. I hope many of you will try this too; you won't regret it.
The g2 dissipation can be 12 watts without getting red. At the moment I use an ecc83; Rk becomes 2k2 and everything else stays the same. I wanted a little more gain as mu g2g1 of the 519 is only about 3,6. I hope many of you will try this too; you won't regret it.
You will acheive about 5Watts like that with heavy compression as the G2 heats up. 7W peak.
Not exactly an improvement on a standard triode amplifier.
Not exactly an improvement on a standard triode amplifier.
I estimate the output power to be about 3 wats. I have quite a few amps ( pp, se, both directly and indirectly heated) and this is one of the best sounding.
That is a cute circuit, but . . .
1. The 390 Ohm cathode resistor in the PL519 is un-bypassed.
That will cause a large loss of voltage gain;
And that will cause a large increase of the screen resistance, rs, that drives the 3.5k output transformer primary.
So . . .
Low gain
Low damping factor
Limited frequency response (high screen resistance drives the primary inductance at low frequencies; and high screen resistance drives the distributed capacitance of the primary).
By any chance, is the schematic incorrectly drawn?
Perhaps the 20uF oil cap should be connected directly to the cathode.
Try that, and see if you like it.
2. On a different circuit . . .
I once built a 4-65A true tetrode amplifier. But . . .
The screen drove the primary of the output transformer.
The Screen maximum dissipation is 10 Watts.
The Plate Cap was connected directly to ground. No plate current.
The screen is built like a Bird Cage. But the plate hides the view of the screen.
The 4-65A is a very pretty tube.
1. The 390 Ohm cathode resistor in the PL519 is un-bypassed.
That will cause a large loss of voltage gain;
And that will cause a large increase of the screen resistance, rs, that drives the 3.5k output transformer primary.
So . . .
Low gain
Low damping factor
Limited frequency response (high screen resistance drives the primary inductance at low frequencies; and high screen resistance drives the distributed capacitance of the primary).
By any chance, is the schematic incorrectly drawn?
Perhaps the 20uF oil cap should be connected directly to the cathode.
Try that, and see if you like it.
2. On a different circuit . . .
I once built a 4-65A true tetrode amplifier. But . . .
The screen drove the primary of the output transformer.
The Screen maximum dissipation is 10 Watts.
The Plate Cap was connected directly to ground. No plate current.
The screen is built like a Bird Cage. But the plate hides the view of the screen.
The 4-65A is a very pretty tube.
Last edited:
The unbypassed part of the cathode resistor gives only a small difference in gain but gives a little improvement of sound quality. The circuit is not incorrectly drawn; this is how I made it and it sounds very good (J. Weiss asks $ 31.500 for it the only difference being the use of a stabilized pwr supply with a 6080).
A friend of mine also built a 4-65a amp screen as anode and that one sounds very good too.
A friend of mine also built a 4-65a amp screen as anode and that one sounds very good too.
Interesting, but I wonder why. What are you gaining in the process of burning up the PL519 for its limited screen dissipation, while leaving idles those 33+ watt plates? I’m just trying to understand why this “sounded good”, presumably better that a triode connected PL519. Good PL519s are also scarcer and more expensive now than 20 yrs ago.
Kees,
Thanks again. I saw your response on Lenco Heaven. I'm now driving the 509/6KG6 with a 6A6 via Hammond 126C interstage which I'm really liking. And will probably do a White/monkey style next. I've tried 6HJ5 G2 as plate but haven't been very successful and am looking around in my stash for other options. I'm really liking the G2 sound.
I'm 300V on G2, 60V on cathode over 1.2K resistor so far. Plate and G3 unconnected.
Matt
Thanks again. I saw your response on Lenco Heaven. I'm now driving the 509/6KG6 with a 6A6 via Hammond 126C interstage which I'm really liking. And will probably do a White/monkey style next. I've tried 6HJ5 G2 as plate but haven't been very successful and am looking around in my stash for other options. I'm really liking the G2 sound.
I'm 300V on G2, 60V on cathode over 1.2K resistor so far. Plate and G3 unconnected.
Matt
Kees,
Thanks!
I guess the lower gain and lower damping factor was worth trading off to have lower distortion.
After all, the sound is what it is all about.
Regarding the un-bypassed 390 resistor:
I did calculations with the un-bypassed 390 Ohm resistor.
With a screen (acting as plate) load of 3500 Ohms, and a g1g2 u = 3.6:
3500 Ohms / (U + 1) = 3500 / 4.6 = 761 Ohms. That is the impedance of the cathode.
The signal on the control grid gets divided between the cathode impedance and the 390 Ohm un-bypassed resistor.
761 + 390 = 1151 Ohms
761 / 1151 = 0.66
20 Log 0.66 = -3.6 dB. That is just a moderate amount of negative feedback.
The increased screen impedance (acting as plate) is u x 390 Ohms.
3.6 x 390 Ohms = 1404 Ohms. The screen impedance is increased by 1404 Ohms.
All this about a good sounding amplifier, almost makes me want to build a 4-65A amplifier again, with the screen acting as plate, and the plate grounded.
(People really wondered about the grounded plate that was in plain view).
Low power is usually all I need, I mostly do near-field listening, 2 1/2 to 3 feet.
Thanks!
I guess the lower gain and lower damping factor was worth trading off to have lower distortion.
After all, the sound is what it is all about.
Regarding the un-bypassed 390 resistor:
I did calculations with the un-bypassed 390 Ohm resistor.
With a screen (acting as plate) load of 3500 Ohms, and a g1g2 u = 3.6:
3500 Ohms / (U + 1) = 3500 / 4.6 = 761 Ohms. That is the impedance of the cathode.
The signal on the control grid gets divided between the cathode impedance and the 390 Ohm un-bypassed resistor.
761 + 390 = 1151 Ohms
761 / 1151 = 0.66
20 Log 0.66 = -3.6 dB. That is just a moderate amount of negative feedback.
The increased screen impedance (acting as plate) is u x 390 Ohms.
3.6 x 390 Ohms = 1404 Ohms. The screen impedance is increased by 1404 Ohms.
All this about a good sounding amplifier, almost makes me want to build a 4-65A amplifier again, with the screen acting as plate, and the plate grounded.
(People really wondered about the grounded plate that was in plain view).
Low power is usually all I need, I mostly do near-field listening, 2 1/2 to 3 feet.
Last edited:
Thanks Francois, Matt, Pyka and 6A3s for your reactions,
I very much liked the sound produced by a few pre-war mesh plate tubes I happened to have but these are very rare (and expensive) and therefore I thought the g2 wires could possibly used as such. I had collected a large number of 509/519's for my 10 tube per channel otl's. The otl's require careful selection of matched tubes so I bought these tubes where ever I could find them (I have a few hundred of them now).
I also found the best sounding output tubes to be low-mu types so the low-mu g2g1 of the 509/519 came out handy. The fact that these tubes are very sturdy and I had many of those made the choice easy. Next of course was the question how much could g2 dissipate? It appeared most could stand 15w without getting red so I decided to go a little lower than that. Besides these tubes are so strong when used this way they last a lifetime. A friend of mine used them in his working place on a daily basis for 4 years and when tested they measured the same as in the beginning.
I cannot answer the question why the tube used this way sounds that transparent and indeed more transparent than used as ordinary triode plate and g2 connected. Actually I wanted a cheap alternative for the 300B and that's what I got. I can only say: Try it for yourself!
I very much liked the sound produced by a few pre-war mesh plate tubes I happened to have but these are very rare (and expensive) and therefore I thought the g2 wires could possibly used as such. I had collected a large number of 509/519's for my 10 tube per channel otl's. The otl's require careful selection of matched tubes so I bought these tubes where ever I could find them (I have a few hundred of them now).
I also found the best sounding output tubes to be low-mu types so the low-mu g2g1 of the 509/519 came out handy. The fact that these tubes are very sturdy and I had many of those made the choice easy. Next of course was the question how much could g2 dissipate? It appeared most could stand 15w without getting red so I decided to go a little lower than that. Besides these tubes are so strong when used this way they last a lifetime. A friend of mine used them in his working place on a daily basis for 4 years and when tested they measured the same as in the beginning.
I cannot answer the question why the tube used this way sounds that transparent and indeed more transparent than used as ordinary triode plate and g2 connected. Actually I wanted a cheap alternative for the 300B and that's what I got. I can only say: Try it for yourself!
Just trying to explain: Because one can look through a mesh plate and see the grids, the filamentary cathode etc., the tube, of course, sounds more transparent.
Well, just kidding. It's just another snake oil.
Best regards!
Hi Pyka,
In your situation you have 60v at the cathode; as Vb is 265v you get about 200v across the tube and Pg2 is 6 watts. No doubt it will work. But you can easily go higher than that. See Matt's reaction. Vb 300, Rk 1k2 and 60v Vk means 50ma Ik; Now you have 240 v across the tube and Pg2= 12 w. This should be possible. I have only experience with European tubes (Philips, Telefunken, Siemens) and they never gave problems but I believe these Russian tubes are sturdy as well. In case g2 gets a little red add say a 100 ohms to Rk. You don't have to be afraid the tube will die so soon. I once had a leaky coupling cap and 80 v on g1 for an hour and after replacing the cap the amp played like before.
In your situation you have 60v at the cathode; as Vb is 265v you get about 200v across the tube and Pg2 is 6 watts. No doubt it will work. But you can easily go higher than that. See Matt's reaction. Vb 300, Rk 1k2 and 60v Vk means 50ma Ik; Now you have 240 v across the tube and Pg2= 12 w. This should be possible. I have only experience with European tubes (Philips, Telefunken, Siemens) and they never gave problems but I believe these Russian tubes are sturdy as well. In case g2 gets a little red add say a 100 ohms to Rk. You don't have to be afraid the tube will die so soon. I once had a leaky coupling cap and 80 v on g1 for an hour and after replacing the cap the amp played like before.
I got good results by using 6L6 and EL503 triode connected at 1/10 of the rated power or even less. My hypotesis is that the large cathode and the low internal resistence of the tube are the main reasons of the good sound. On this context, I see the unconnected anode as a excellent solution to the top cap issue. Has anyone tried the trick on a ceramic external anode triode such as 8167 / 4CX300A ? It does have a even larger cathode emission, screen dissipation is 12W, and it is pitched as vibration resistent (low microphonics). Unfortunately, there are no published triode curves of this tube family connected with G2 as anode. If it works, it should be possible to bolt the anode to a large chassis heatsink, avoiding the special (and expensive) socket and the forced cooling. Heater supply requirements and size are also reasonable.
I once had 0V on the grid of a 12GT5 and an orange plating tube for about 2 minutes. The tube still "worked" but now instead of -60V bias it only needed -12V... Tube was used up in 2 minutes 🙂
It sounded fine though - at least until the tube ran out of current to give...
Kees,
I also saw your post on the Lenco forum and since I had most to the parts required, I tried it as well. My amp is similar but different. I am using EL509/6KG6 at 10 W plate dissipation. (285V on the anode, 62V at the cathode, Rk 1361). I have a regulated power supply with variable B+ voltage and that is where I settled out. Once it sounded good, I stopped and haven't adjusted it. I want to live with it awhile.
I will confirm that is a nice sounding setup. I would like to get measurements on frequency response and distortion, but I don't have the equipment. If there is anyone in DFW that could help with that, I would appreciate it.
I also saw your post on the Lenco forum and since I had most to the parts required, I tried it as well. My amp is similar but different. I am using EL509/6KG6 at 10 W plate dissipation. (285V on the anode, 62V at the cathode, Rk 1361). I have a regulated power supply with variable B+ voltage and that is where I settled out. Once it sounded good, I stopped and haven't adjusted it. I want to live with it awhile.
I will confirm that is a nice sounding setup. I would like to get measurements on frequency response and distortion, but I don't have the equipment. If there is anyone in DFW that could help with that, I would appreciate it.
Hi Decramer,
You can certainly parallel 509/519's. I have one myself just opposite of where I sit right now. Give each one it's own cathode resistors; matched tubes are not necessary as they are not operating in pp. I use these big TME toroidal outputs and that's a good combination. A friend of mine even used 4 tubes in parallel,
again each with it's own Rk and a 1k output transformer.
Todd: As you use a variable regulated pwr supply it's possible you find a position you think sounds best. If so it would be interesting for the forum members. I have to admit I built quite a few of these amps each with different voltages and each one sounded good but I never took the trouble to find which one sounded best.
You can certainly parallel 509/519's. I have one myself just opposite of where I sit right now. Give each one it's own cathode resistors; matched tubes are not necessary as they are not operating in pp. I use these big TME toroidal outputs and that's a good combination. A friend of mine even used 4 tubes in parallel,
again each with it's own Rk and a 1k output transformer.
Todd: As you use a variable regulated pwr supply it's possible you find a position you think sounds best. If so it would be interesting for the forum members. I have to admit I built quite a few of these amps each with different voltages and each one sounded good but I never took the trouble to find which one sounded best.