Hi folks,
Has anybody tried out the 814 pentode as triode ?
There are very nice curves on Pete Millet's site: http://www.tubebooks.org/file_downloads/814%20curves.gif
Thing is, anode voltage (and thus G2 voltage) goes well beyound 1kV. Furthermore the drawed Operating Point is at about 700V.
But: In the Datasheet they give a max. of 400V Ug2.
The reason might be that the datasheet only gives RF operating data. Max Ug2 for AF operating is usually much higher. But I'm still in doubt. Anyone has some facts ?
Has anybody tried out the 814 pentode as triode ?
There are very nice curves on Pete Millet's site: http://www.tubebooks.org/file_downloads/814%20curves.gif
Thing is, anode voltage (and thus G2 voltage) goes well beyound 1kV. Furthermore the drawed Operating Point is at about 700V.
But: In the Datasheet they give a max. of 400V Ug2.
The reason might be that the datasheet only gives RF operating data. Max Ug2 for AF operating is usually much higher. But I'm still in doubt. Anyone has some facts ?
Actually I do have the same question as I plan to build a PP-Amp at 700V...any new views on this ?
The data sheet gives pentode mode data. Then high g2 current occur at the moment when plate voltage goes lower than Ug2.
At triode mode this does not happen since plate and g2 are at the same potential.
At triode mode this does not happen since plate and g2 are at the same potential.
...this would mean that there should be no issue at 700V ?
I was wondering as even Ale (Bartola) got nervous at 600V...he suggested to tie together g2 and g3 to the anode using an 1K resistor to make it more rough/reliable...does this makes sense, normally I would connect g3 to ground, no ?
I was wondering as even Ale (Bartola) got nervous at 600V...he suggested to tie together g2 and g3 to the anode using an 1K resistor to make it more rough/reliable...does this makes sense, normally I would connect g3 to ground, no ?
Last edited:
The high voltage is not the problem, but too high screen current would be.
In triode connection the possibility of overheating the g2 is mostly avoided.
Normally, means in pentode connection, it is obvious to connect g3 to cathode, but there is no sense to do so if triode connection is the target.
So all plate, g3 and g2 together.
In triode connection the possibility of overheating the g2 is mostly avoided.
Normally, means in pentode connection, it is obvious to connect g3 to cathode, but there is no sense to do so if triode connection is the target.
So all plate, g3 and g2 together.
Thx for your thoughts...I just found an old thread which I read years ago:
EL34 Triode Mode - G3 where?
to summarize:
- Normally, connecting g3 to the cathode will decrease distortion a bit (see measurements for the El34), theory is that electrons get better controlled.
- one has to distinguish between g3 as a control grid and g3 as a beam forming plate. When a beamforming plate, no choice: Needs to be connected to the cathode, otherwise the beam forming plates becoming the opposite of beam forming: They become anodes.
- For non beam tubes with normal grid at g3, some people claimed, that only connecting g3 to the anode makes up a true triode behavior, but when doing so it is advised to give g2 a lower grid stopper value like 200ohms vs. 1k for g3
...I have measured a 4p1l in both connections and indeed rp, mu and trans. are changing. Not dramatically, but like rp 1k5 to 1k3 and mu 8.3 va. 9 if I recall right.
So, in the case of the 814: I see that Ale worked well with connecting g3 to the plate, so it will work. Therefore I guess I will simply test and listen both scenarios.
The only thing which would have made me believe that there is only one correct way of connecting g3 (to cathode) are the three hints inthe data sheet https://frank.pocnet.net/sheets/049/8/814.pdf
1. The title: This is a beam power tube...so how is the beam control generated if g3 is not at 0V ?
2. There are at least four application examples where g3-voltages is defined as 0V with the footnote to connect g3 to either the mid-point of an AC-Heater or the neg end of a DC-Heater
3. In the title of the grid current curves they talk about beam PLATES...which are defined there to be at 0V...so no grids, PLATES !
EL34 Triode Mode - G3 where?
to summarize:
- Normally, connecting g3 to the cathode will decrease distortion a bit (see measurements for the El34), theory is that electrons get better controlled.
- one has to distinguish between g3 as a control grid and g3 as a beam forming plate. When a beamforming plate, no choice: Needs to be connected to the cathode, otherwise the beam forming plates becoming the opposite of beam forming: They become anodes.
- For non beam tubes with normal grid at g3, some people claimed, that only connecting g3 to the anode makes up a true triode behavior, but when doing so it is advised to give g2 a lower grid stopper value like 200ohms vs. 1k for g3
...I have measured a 4p1l in both connections and indeed rp, mu and trans. are changing. Not dramatically, but like rp 1k5 to 1k3 and mu 8.3 va. 9 if I recall right.
So, in the case of the 814: I see that Ale worked well with connecting g3 to the plate, so it will work. Therefore I guess I will simply test and listen both scenarios.
The only thing which would have made me believe that there is only one correct way of connecting g3 (to cathode) are the three hints inthe data sheet https://frank.pocnet.net/sheets/049/8/814.pdf
1. The title: This is a beam power tube...so how is the beam control generated if g3 is not at 0V ?
2. There are at least four application examples where g3-voltages is defined as 0V with the footnote to connect g3 to either the mid-point of an AC-Heater or the neg end of a DC-Heater
3. In the title of the grid current curves they talk about beam PLATES...which are defined there to be at 0V...so no grids, PLATES !
Last edited:
The beam forming at pentode mode is necessary to avoid anode current kink, as we all know,
but when operated in triode mode there is no need for any "beam forming" because there is no such element as screen grid that originally generates the need to use beam forming plates.
It would be possible that if the beam forming plates are used as a part of anode, the total anode dissipation is increased substantially.
For sure the Gm is increased, if beam forming plates are connected to plate.
Last edited:
Hi,
it took much longer than expected, but now I am ready to get my 814 working in SE mode.
With all the respect to Ale, I read the data sheet he posted...especially regarding Class B mode:
http://www.bartola.co.uk/valves/wp-content/uploads/2013/01/VT-154_GL-814-Tetrode-datasheet.pdf
Ok...
814 SE A2 Amplifier – Bartola(R) Valves
I am not sure though I follow his design approach.
His design approach:
- Seems to be not to exceed 600V on g2...which is not an issue actually in triode mode as discussed here and elsewhere.
- class A2...even though his psu made 800V...only to use cathode bias ??? Instead of fixed bias...?
- 100mA from the cathode ??? The data sheet specs 60mA as max...so, the old rule: high voltage, low current and the tube will live longer
Open question is how much g2 power is allowed...I would read the data sheet with 6.7Watt max...no? So, the resistor in front of g2 needs to be adjusted to regulate current if necessary...no ?
So, if my understanding is right (please correct if Igot something wrong), I will
- get HV high to 800V+
- class a1
- max 60mA
- max 50W anode dissipation...800V@60mA..48Watt.
- https://i0.wp.com/www.bartola.co.uk/valves/wp-content/uploads/2014/05/814-SE-class-A1.png
- so...-Ug should be around -76V...leading to a voltage swing of approx. 900V...
- 900Vpp=318Vrms into a 7K primary and 6 ohm load on the secondary is a ratio of 34:1, leading to 9.35Vrms on the secondary. 9.35V/6ohm=1.56A...lead to 14.6W in class A1..
- transformer coupled...so anything above 150Vpp from the driver gives us transient reserves...in Class A2
Any flaws in that plan ?
it took much longer than expected, but now I am ready to get my 814 working in SE mode.
With all the respect to Ale, I read the data sheet he posted...especially regarding Class B mode:
http://www.bartola.co.uk/valves/wp-content/uploads/2013/01/VT-154_GL-814-Tetrode-datasheet.pdf
Ok...
814 SE A2 Amplifier – Bartola(R) Valves
I am not sure though I follow his design approach.
His design approach:
- Seems to be not to exceed 600V on g2...which is not an issue actually in triode mode as discussed here and elsewhere.
- class A2...even though his psu made 800V...only to use cathode bias ??? Instead of fixed bias...?
- 100mA from the cathode ??? The data sheet specs 60mA as max...so, the old rule: high voltage, low current and the tube will live longer
Open question is how much g2 power is allowed...I would read the data sheet with 6.7Watt max...no? So, the resistor in front of g2 needs to be adjusted to regulate current if necessary...no ?
So, if my understanding is right (please correct if Igot something wrong), I will
- get HV high to 800V+
- class a1
- max 60mA
- max 50W anode dissipation...800V@60mA..48Watt.
- https://i0.wp.com/www.bartola.co.uk/valves/wp-content/uploads/2014/05/814-SE-class-A1.png
- so...-Ug should be around -76V...leading to a voltage swing of approx. 900V...
- 900Vpp=318Vrms into a 7K primary and 6 ohm load on the secondary is a ratio of 34:1, leading to 9.35Vrms on the secondary. 9.35V/6ohm=1.56A...lead to 14.6W in class A1..
- transformer coupled...so anything above 150Vpp from the driver gives us transient reserves...in Class A2
Any flaws in that plan ?
Hi,
Not sure if you understood the circuit topology. Check on Michael Koster's design and there is a lot about it here in this forum.
It's fixed bias as supplies are stacked, the HT supply is 600V, not 800V, hence the limitation.
At 100mA iddle current in triode works fine and valves measured same after few years of you. If you want to be more conservative and have higher HT, it does make sense.
Good luck
Cheers
Ale
Not sure if you understood the circuit topology. Check on Michael Koster's design and there is a lot about it here in this forum.
It's fixed bias as supplies are stacked, the HT supply is 600V, not 800V, hence the limitation.
At 100mA iddle current in triode works fine and valves measured same after few years of you. If you want to be more conservative and have higher HT, it does make sense.
Good luck
Cheers
Ale
So, it took a lottle longer than expected, as I went on to compare the 300B family with Svetlana 300B, EML 300mesh, Kron 842, Kron T-100, EmL 520 and lately Shuguang 845c (with Metal Anodes).
I prefer the 845c by far sofar, followed by the Kron 842. I like the fine and fluent resolution of the TT filament...Thomas Mayer has there really a point (Elrog tubes).
So, now I am ready for the 814 as I will will compare all the TT filament candidates (814,813,Gm70,845,211,4242).
I might start a little higher with around 840V and 60mA...if I got it right g2 is able to take 6.7Watt input, which means 6.7/840=8mA of the 60mA, right ?
I have put a 150 ohm resistor between Anode and g2, so I need to measure a voltage smaller than 1,2V (150*0.008) and we are good to go, correct ?
I prefer the 845c by far sofar, followed by the Kron 842. I like the fine and fluent resolution of the TT filament...Thomas Mayer has there really a point (Elrog tubes).
So, now I am ready for the 814 as I will will compare all the TT filament candidates (814,813,Gm70,845,211,4242).
I might start a little higher with around 840V and 60mA...if I got it right g2 is able to take 6.7Watt input, which means 6.7/840=8mA of the 60mA, right ?
I have put a 150 ohm resistor between Anode and g2, so I need to measure a voltage smaller than 1,2V (150*0.008) and we are good to go, correct ?
I wonder as well because this https://tubedata.tubes.se/sheets/049/8/814.pdf tells us, its a 50W device in CCS with 6.7W on g2...So, if you give it the full 50W, at 1400V, we are down to 35mA...what is your application ?
using a pair of RCA-813 beam power tubes , as SE audio amp , driven by passive exiter gain stage, not a small signal tube preamp, or any FET preamp. just two power tubes , that's it.........
Ok, got my 814 working this weekend...very nice. very.
I am running currently Philips NOS, build like a tank. I currently use it with 920V@54mA...after a couple of hours it fall to 50mA, I need to see if I am destroying them already with such a setting or is this is more HV fluctuation as I run them
- fixed bias with regs. coleman regs
- DC filament coleman regs
- g3 grounded
- g2 over 150 ohm resistor to Anode (with 0,6v across so 4mA goes over g2)
- 7k4 primary
I am running currently Philips NOS, build like a tank. I currently use it with 920V@54mA...after a couple of hours it fall to 50mA, I need to see if I am destroying them already with such a setting or is this is more HV fluctuation as I run them
- fixed bias with regs. coleman regs
- DC filament coleman regs
- g3 grounded
- g2 over 150 ohm resistor to Anode (with 0,6v across so 4mA goes over g2)
- 7k4 primary
That's a shame, I presume you measured the HT and it's not changed. Unless you have some form of regulation it will fluctuate with the mains supply in your area; 230Vac +/-10%
Is it the same situation for both tubes ?
The gas in the tube could have reduced as it's a NOS tube, but I dunno if that could be a cause.
Will be interested to know how you get on.
Is it the same situation for both tubes ?
The gas in the tube could have reduced as it's a NOS tube, but I dunno if that could be a cause.
Will be interested to know how you get on.