I am not an expert in this so i guess you are right, usually gain goes down... After much testing i think i finished the build.....
At 50w i get 0.3 - 0,4% THD at 600v plate voltage with 10db of NFB, i can go higher more than 20db now after some changes i made but i don't think it is needed. Basically what made the amplifier very stable was a 5,6Kohm 5w resistor in series with a 2,2nf capacitor from one anode to the other.
I also measured the frequency response with Feedback and without. Without NFB i get 0db @ 1khz, +3db at 20 Hz, +6 db from 2khz to 4khz, +10db up to 7-8 khz approximately and +3db at 20khz after halving the power stage's coupling capacitor otherwise i get ~+13db as i said before. With NFB the frequency response curve is almost a flat line with 0db at 1khz -2db at 20Hz and -2db at 20Khz.
Chris
At 50w i get 0.3 - 0,4% THD at 600v plate voltage with 10db of NFB, i can go higher more than 20db now after some changes i made but i don't think it is needed. Basically what made the amplifier very stable was a 5,6Kohm 5w resistor in series with a 2,2nf capacitor from one anode to the other. I also measured the frequency response with Feedback and without. Without NFB i get 0db @ 1khz, +3db at 20 Hz, +6 db from 2khz to 4khz, +10db up to 7-8 khz approximately and +3db at 20khz after halving the power stage's coupling capacitor otherwise i get ~+13db as i said before. With NFB the frequency response curve is almost a flat line with 0db at 1khz -2db at 20Hz and -2db at 20Khz.
Chris
I'll be using toroids as OPTs so I might not need that much. I'm basically dropping these into a Williamson KT88 design I already have. With each tube with it's plates in parallel (through 10R sharing resistors). I considered doing a triode connected design but the internal cap between screen and cathode might interfere? I dunno why it's even there, but they are RF tubes...
Will not work as a triode because 225v is the absolute maximum for the screens. The capacitor is too small for the audio range i think so it is not a problem. I was thinking of using a toroid as an output transformer but i wasn't sure if it will work ok, have you any experience with toroidal transformers?
Chris
Chris
You might be able to use the 829B Beam Power Tube in triode mode at more than 225V screen + plate; but only if you use both sides of the tube in parallel (there is only 1 screen, so two plates and one screen must all be in parallel).
Screen voltage is limited in Beam Power mode:
In beam power mode, with 225V on the screen, when the control grid goes to 0V, the plate voltage goes far below 225V. As a result, the screen draws way too much current (and that current times 225V cooks the screen).
Screen voltage limit may be higher in Triode mode:
In triode mode, when the control grid goes to 0V, the plate goes far far below 225V, but since the screen is connected to the plate (suggested to be through a 100 Ohm resistor), the screen voltage also goes far far below 225V. As a result, the screen does not draw nearly as much current.
Other screen voltage limitations:
The insulation of the screen to the other elements may be an issue, as well as the screen to cathode bypass capacitor voltage rating.
The 5894 has similar limitations, with maximum screen rating of 250V, and triode mode must have two plates and one screen in parallel. Might it work at 300V? I don't know.
An example of (successful) increased maximum screen voltage rating is the 807. The 807 in Beam Power mode has a maximum screen voltage rating of 300V. But the same 807 Beam Power Tube in triode mode has a maximum screen voltage rating of 400V. That resulted in 33% more voltage rating. With 33% more voltage swing into the load you might get 78% more power (+2.5 dB) versus the old maximum screen voltage rating of 300V.
Screen voltage is limited in Beam Power mode:
In beam power mode, with 225V on the screen, when the control grid goes to 0V, the plate voltage goes far below 225V. As a result, the screen draws way too much current (and that current times 225V cooks the screen).
Screen voltage limit may be higher in Triode mode:
In triode mode, when the control grid goes to 0V, the plate goes far far below 225V, but since the screen is connected to the plate (suggested to be through a 100 Ohm resistor), the screen voltage also goes far far below 225V. As a result, the screen does not draw nearly as much current.
Other screen voltage limitations:
The insulation of the screen to the other elements may be an issue, as well as the screen to cathode bypass capacitor voltage rating.
The 5894 has similar limitations, with maximum screen rating of 250V, and triode mode must have two plates and one screen in parallel. Might it work at 300V? I don't know.
An example of (successful) increased maximum screen voltage rating is the 807. The 807 in Beam Power mode has a maximum screen voltage rating of 300V. But the same 807 Beam Power Tube in triode mode has a maximum screen voltage rating of 400V. That resulted in 33% more voltage rating. With 33% more voltage swing into the load you might get 78% more power (+2.5 dB) versus the old maximum screen voltage rating of 300V.
Last edited:
Very well explained, thank you! I don't know why someone would like to use a pentode as a triode because the sound that comes out from the 829 is of excellent clarity and quality.
BTW after many tests i decided against NFB but i will fit a switch and a potentiometer in the amplifier's face for adjustments. I am now certain after hearing many cd (wav not mp3) that there is information lost or altered with feedback, the open loop sound is far better and clear even at 3% distortion. I can even hear the recording's bad spots easily where with NFB everything sounds dull (excellent sound again but a bit dull) and this is not because of the frequency response because i fixed that.
Chris
BTW after many tests i decided against NFB but i will fit a switch and a potentiometer in the amplifier's face for adjustments. I am now certain after hearing many cd (wav not mp3) that there is information lost or altered with feedback, the open loop sound is far better and clear even at 3% distortion. I can even hear the recording's bad spots easily where with NFB everything sounds dull (excellent sound again but a bit dull) and this is not because of the frequency response because i fixed that.
Chris
Last edited:
hendrixgr,
I am glad your 829B amp sounds great.
In general amps with Beam Power Output Tubes have a low damping factor. How this interacts with different loudspeakers varies. Global Negative feedback increases the amp's damping factor.
Global Negative feedback can lower the distortion amplitude of the low order harmonics. But it also adds some upper order harmonics that were not there before, even though their amplitudes are far below the amplitudes of the original low order harmonics.
I would rather hear your amplifier and speakers, than to question the theory or the results.
Happy listening!
I am glad your 829B amp sounds great.
In general amps with Beam Power Output Tubes have a low damping factor. How this interacts with different loudspeakers varies. Global Negative feedback increases the amp's damping factor.
Global Negative feedback can lower the distortion amplitude of the low order harmonics. But it also adds some upper order harmonics that were not there before, even though their amplitudes are far below the amplitudes of the original low order harmonics.
I would rather hear your amplifier and speakers, than to question the theory or the results.
Happy listening!
I understand what you are saying very well. In the end i decided with my ears and my subjective brain's audio processing. All measurements with 20db of global NFB are better than without it, SNR is 98 db instead of 70, THD is 0.2-0.3% instead of 2-2.5%, everything...
When i use my ears with global NFB drums are soft, the singer's voice is softer like it lost it's sharpness and the background instruments like bass, piano etc sound like lost in the distance. I am not saying that the sound is not good but it does start to sound like my SS amplifier....
Chris.
When i use my ears with global NFB drums are soft, the singer's voice is softer like it lost it's sharpness and the background instruments like bass, piano etc sound like lost in the distance. I am not saying that the sound is not good but it does start to sound like my SS amplifier....
Chris.
Chris,
You are bringing out what I often tell others, this is about a System, not just an amplifier. Negative feedback is not always a Panacea. Each part in the system interacts with others in the system (hopefully Positive Synergy). Recording Equipment and Technique, Recording Media, Signal Source, Amp, Loudspeakers, Room, and Ears.
If it sounds good, it probably is.
You are bringing out what I often tell others, this is about a System, not just an amplifier. Negative feedback is not always a Panacea. Each part in the system interacts with others in the system (hopefully Positive Synergy). Recording Equipment and Technique, Recording Media, Signal Source, Amp, Loudspeakers, Room, and Ears.
If it sounds good, it probably is.
one more question about tubes
Hi. I am ready to proceed to the actual build but i think i will go crazy because i can't decide what tubes to use for the PI and the preamplifier so i need a little extra push I can't decide between E88CC and 6SN7 for the preamplifier and PI stage, my goal is the absolute minimum THD without global NFB. Of course each tube type will have it's own design parameters bias and load, i am not talking about replacing one with the other.
Also i can't decide if it is useful to use a 6SN7 after the Cathodyne phase splitter ala Williamson because the final power stage may requires only 20Vp (40Vpp) but if tomorrow i decide to use local feedback for the push pull power stage (schade feedback), an extra post PI 6SN7 stage might be useful.
Any opinion is welcomed!
Chris
Hi. I am ready to proceed to the actual build but i think i will go crazy because i can't decide what tubes to use for the PI and the preamplifier so i need a little extra push
I can't decide between E88CC and 6SN7 for the preamplifier and PI stage, my goal is the absolute minimum THD without global NFB. Of course each tube type will have it's own design parameters bias and load, i am not talking about replacing one with the other. Also i can't decide if it is useful to use a 6SN7 after the Cathodyne phase splitter ala Williamson because the final power stage may requires only 20Vp (40Vpp) but if tomorrow i decide to use local feedback for the push pull power stage (schade feedback), an extra post PI 6SN7 stage might be useful.
Any opinion is welcomed!
Chris
Last edited:
Then build one of each and evaluate the result !Hi. I am ready to proceed to the actual build but i think i will go crazy because i can't decide what tubes to use for the PI and the preamplifier so i need a little extra push
Also i can't decide if it is useful to use a 6SN7 after the Cathodyne phase splitter ala Williamson because the final power stage may requires only 20Vp (40Vpp) but if tomorrow i decide to use local feedback for the push pull power stage (schade feedback), an extra post PI 6SN7 stage might be useful.
Any opinion is welcomed!
Chris
Feedback, both internally within an amp and externally when observing the result of something AND make choices according to the experience is the "scientific method".
After much testing i decided to use the 6SN7GT. It has lower distortion and it sounds well WITH FEEDBACK OR NOT! At last because this dilemma was wearing me out but now i am thinking if it is a good idea to also fit an LTP after the cathodyne like the Williamson topology, any opinions?
I only use 10db of feedback and the distortion at 30w output is 0.2% max and it doesn't go much up even at preset full output of 60w! SNR is 98db and there is no noise at all even without feedback. I also positioned the leds underneath the output tubes just to check the effect which is very nice but i am still not satisfied because the 6SN7 has a key and i can't use the socket hole for injecting light in to the tube. Finally i need to add some blue leds in to the led mix...
Chris
I only use 10db of feedback and the distortion at 30w output is 0.2% max and it doesn't go much up even at preset full output of 60w! SNR is 98db and there is no noise at all even without feedback. I also positioned the leds underneath the output tubes just to check the effect which is very nice but i am still not satisfied because the 6SN7 has a key and i can't use the socket hole for injecting light in to the tube. Finally i need to add some blue leds in to the led mix...
Chris
Attachments
Last edited:
If you absolutely must have LED lit tubes, you can use the 6CG7/6FQ7, they are identical to the 6SN7, just in a nine-pin base, and the 6CG7 has an internal shield between the plates
I've got piles of them, as they are often fairly priced NOS, and very useful little guys.
Williamson style topology wouldn't be a bad idea, 6SN7 concertina into a 6SN7 LTP is one of my favorite topologies for swinging lotsavolts. My preference is to leave the cathode of the concertina unbypassed, gives distortion that is low and primarily 2nd harmonic. I say try it and see
I've got piles of them, as they are often fairly priced NOS, and very useful little guys.
Williamson style topology wouldn't be a bad idea, 6SN7 concertina into a 6SN7 LTP is one of my favorite topologies for swinging lotsavolts. My preference is to leave the cathode of the concertina unbypassed, gives distortion that is low and primarily 2nd harmonic. I say try it and see
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.