On beauty and genius
Listened with the feedback loop in for a couple of days. Sounded okay, but closed in and bye bye magic..... Like watching a plain looking woman and trying to convince yourself she's beautiful (we're among men here, aren't we?). Or if you prefer: like talking to your neighbor and trying to convince yourself he's a true genius. Fatiguing.
Just removed it and am enjoying overwhelming beauty/genius again: effortless room-filling sound, engaging, immediacy, lovely textures, voices, decay of cymbals is a treat once again. Very happy with this build and even better times to come when the caps smooth out further.
The baby huey is probably even better, but this might just be the ticket if you have sand-phobia or want more simplicity.
Listened with the feedback loop in for a couple of days. Sounded okay, but closed in and bye bye magic..... Like watching a plain looking woman and trying to convince yourself she's beautiful (we're among men here, aren't we?). Or if you prefer: like talking to your neighbor and trying to convince yourself he's a true genius. Fatiguing.
Just removed it and am enjoying overwhelming beauty/genius again: effortless room-filling sound, engaging, immediacy, lovely textures, voices, decay of cymbals is a treat once again. Very happy with this build and even better times to come when the caps smooth out further.
The baby huey is probably even better, but this might just be the ticket if you have sand-phobia or want more simplicity.
Klimon,
Glad you are enjoying listening to your amp.
Question:
Based on Mona's schematic on Post # 23 . . .
When you removed the negative feedback network, did you do the following?
Disconnect R28 and C3 from the output transformer secondary.
Connect the disconnected ends of R28 and C3 to ground.
Connect a second 220uF bypass cap across R28 (same kind of cap as C12 is).
And then either retain the 100 Ohm between the cathodes, or take it out.
If you did not do that . . .
Then one input triode cathode sees 100 Ohms in series with the self bias 1500 Ohm paralleled with 220uF to ground. That is 100 Ohms un-bypassed in series with 1500 Ohms that is bypassed.
But the other triode cathode only sees 1500 Ohms that is bypassed to ground.
The shared 1500 Ohms will be at a higher bias voltage for both triodes than before when each cathode had its own 1500 Ohm resistor.
“Those who do not know history, do not know the meaning of Deja Vu”
Glad you are enjoying listening to your amp.
Question:
Based on Mona's schematic on Post # 23 . . .
When you removed the negative feedback network, did you do the following?
Disconnect R28 and C3 from the output transformer secondary.
Connect the disconnected ends of R28 and C3 to ground.
Connect a second 220uF bypass cap across R28 (same kind of cap as C12 is).
And then either retain the 100 Ohm between the cathodes, or take it out.
If you did not do that . . .
Then one input triode cathode sees 100 Ohms in series with the self bias 1500 Ohm paralleled with 220uF to ground. That is 100 Ohms un-bypassed in series with 1500 Ohms that is bypassed.
But the other triode cathode only sees 1500 Ohms that is bypassed to ground.
The shared 1500 Ohms will be at a higher bias voltage for both triodes than before when each cathode had its own 1500 Ohm resistor.
“Those who do not know history, do not know the meaning of Deja Vu”
Summer,
R28 of the upper triode is connected to ground and isn't bypassed (nor by C3, which I omitted, nor by a 220uf cap). The 100 ohm resistor is in place.
What does that tell you? Splitter imbalance and more 2nd order harmonics?
Should I add a 220uF?
Thanks,
Simon
R28 of the upper triode is connected to ground and isn't bypassed (nor by C3, which I omitted, nor by a 220uf cap). The 100 ohm resistor is in place.
What does that tell you? Splitter imbalance and more 2nd order harmonics?
Should I add a 220uF?
Thanks,
Simon
If you only ground the 1k5 there is still the 100Ω in the cathode circuit of the upper triode but the value is to small to make much difference.
If you take the 100Ω out the upper triode get a bit of local (cathode)feedback.
Mona
If you take the 100Ω out the upper triode get a bit of local (cathode)feedback.
Mona
My whole point was:
Each cathode had about 1500 Ohms to ground for self bias.
Call the self bias voltage X.
When you take out the 1500 Ohm feedback resistor, then both cathodes share a single 1500 Ohm resistor to ground.
Now, the self bias voltage has increased to about 1.5 or 2 times X.
That changes the operating point, and the dynamic range, and the distortion of the input triodes.
I hope that explains it.
Each cathode had about 1500 Ohms to ground for self bias.
Call the self bias voltage X.
When you take out the 1500 Ohm feedback resistor, then both cathodes share a single 1500 Ohm resistor to ground.
Now, the self bias voltage has increased to about 1.5 or 2 times X.
That changes the operating point, and the dynamic range, and the distortion of the input triodes.
I hope that explains it.
That is a 'paraphase' phase splitter, not a long-tailed pair.
The Valve Wizard -Paraphase
The imbalance in resistor values is intentional. The outputs from each plate of the phase splitter are inherently imbalanced, and the 220k/250k discrepancy in value is an attempt to balance those outputs -- not a typo.
If you make R20 and R21 the same value, you will make the performance of the phase splitter worse (higher 2nd order harmonic distortion).
--
Not really, but enjoying things too much now to worry about it. It sure doesn't sound 'broken'...
Some baby huey flavor
Would the plate to plate & shunt feedback loop of the baby huey 6v6 work with my build from post 23?
EL84 Amp - Baby Huey
= 39k from 6v6 plate to 13k shunt to 220k driver plate load
I'm anxious to try it, but better safe than sorry from an electrical point of view.
Would it change tonal balance? My system is a little too bright now. Tried GNFB before, but that took away too much life and highs.
Thanks!
Simon
Would the plate to plate & shunt feedback loop of the baby huey 6v6 work with my build from post 23?
EL84 Amp - Baby Huey
= 39k from 6v6 plate to 13k shunt to 220k driver plate load
I'm anxious to try it, but better safe than sorry from an electrical point of view.
Would it change tonal balance? My system is a little too bright now. Tried GNFB before, but that took away too much life and highs.
Thanks!
Simon
Simon,
Are you still preferring no GNF, after living with your amp for a while?
I just noticed you were using Toroidy iron. What primary impedance and did you get custom wound ones with 25% UL taps? I’m wondering if the results with feedback would be more satisfying if you tweeked it for your specific output transformer.
Learning from your experience since I also have that PCB and a standard Toroidy pair of the 10k, 40% UL output transformers.
(TTG-ECL86PP TOROIDY - Transformer: speaker | 25VA; O90x55mm; 0.01/43kHz; 150mA | TME - Electronic components)
Hi François,
I haven't listened to it for a while now, breaking-in speakers with an SS amp (bass boost😉) Ultimate Open Baffle Gallery
Still some 100 hours to go, but eager to get the amp's village-deep soundstage and palpable texture back... Will let you know how it performs on the Liis in a few weeks' time.
A little more complicated, but I'ld also consider a baby huey for those opts, or middle ground: http://www.dissident-audio.com/PP_ECL86/Schema.gif . Your pcbs might even serve if you cut some traces and add some wires.
I'm using the off-the-shelf TTG-EL84PP
Best
Simon
I haven't listened to it for a while now, breaking-in speakers with an SS amp (bass boost😉) Ultimate Open Baffle Gallery
Still some 100 hours to go, but eager to get the amp's village-deep soundstage and palpable texture back... Will let you know how it performs on the Liis in a few weeks' time.
A little more complicated, but I'ld also consider a baby huey for those opts, or middle ground: http://www.dissident-audio.com/PP_ECL86/Schema.gif . Your pcbs might even serve if you cut some traces and add some wires.
I'm using the off-the-shelf TTG-EL84PP
Best
Simon
Simon,
Thanks for the update. I will look forward to your future listening notes; with and without NFB.
I actually have two pairs of the Toroidy TTG-ECL86PP transformers. They are capable of 25watts output, but have 10k primaries, vs. 8k of the TTG-EL84PP, both 43% UL. Reportedly, 6V6s perform best with 10k primaries in PP.
The idea is to compare amp designs and plan to use pair for a 6V6 Baby Huey (using a 6SL6 input, like gingertube said he personally enjoyed best) and the one you built based on Boffin/Mona derivations of the Dynaco 6SL6/6V6 schematic. Perhaps that will also shine some light on your earlier question re. plate-to-plate and shunt feedback.
Thanks for the update. I will look forward to your future listening notes; with and without NFB.
I actually have two pairs of the Toroidy TTG-ECL86PP transformers. They are capable of 25watts output, but have 10k primaries, vs. 8k of the TTG-EL84PP, both 43% UL. Reportedly, 6V6s perform best with 10k primaries in PP.
The idea is to compare amp designs and plan to use pair for a 6V6 Baby Huey (using a 6SL6 input, like gingertube said he personally enjoyed best) and the one you built based on Boffin/Mona derivations of the Dynaco 6SL6/6V6 schematic. Perhaps that will also shine some light on your earlier question re. plate-to-plate and shunt feedback.
That would be an interesting comparison, François - I hope you make it.
After this loudspeaker chapter (bitumen drying on subs + fullrangers breaking in), I'm back into amp building as well. Low power SE thingies.
That'll allow me to better evaluate the 6v6 pp's performance as well.
After this loudspeaker chapter (bitumen drying on subs + fullrangers breaking in), I'm back into amp building as well. Low power SE thingies.
That'll allow me to better evaluate the 6v6 pp's performance as well.
I've had the amp hooked up on my broken-in Liis for a couple of days now and am REALLY enjoying it, even without the bass support the speakers need. Super detailed (discovering 'quiet sounds' and overdubs album after album), involving and transparent sound. Deep soundstage. All that compared to my Onkyo A9110 workhorse. I have no reserve whatsoever to recommend this design. No nfb needed in my case 😉
Simon,
Thanks for the update. Great news! Congratulations again.
I hope to getting around building my Boffin/Mona 6V6 amp late this summer.
Enjoy the music,
Francois
Thanks for the update. Great news! Congratulations again.
I hope to getting around building my Boffin/Mona 6V6 amp late this summer.
Enjoy the music,
Francois
Hi François,
For the record 😉 You might very well need some kind of feedback in this design.
After living with it for a couple of weeks, it's too bright. Might be amended with a speaker zobel in my case (impedance rises significantly in the Liis), but I'll let it rest for a while until I can do distortion measurements or compare with other low damping factor tube amps in the works.
Have fun experimenting
Simon
For the record 😉 You might very well need some kind of feedback in this design.
After living with it for a couple of weeks, it's too bright. Might be amended with a speaker zobel in my case (impedance rises significantly in the Liis), but I'll let it rest for a while until I can do distortion measurements or compare with other low damping factor tube amps in the works.
Have fun experimenting
Simon
Good day Simon,
Thanks for the alert about the assessment of “brightness”. I will follow in you footsteps when I get to build my version. In my experience feedback seems to increase brightness, so I wonder why you think feedback will help in this case.
Kind regards,
Francois
Thanks for the alert about the assessment of “brightness”. I will follow in you footsteps when I get to build my version. In my experience feedback seems to increase brightness, so I wonder why you think feedback will help in this case.
Kind regards,
Francois
From experience: when I ran it with feedback, it sounded more tonally balanced (as does my transistor amp). Without feedback the highs are more prominent and extended.
The theory behind it, I believe, is that nfb increases damping factor, allowing to counter the rise in speaker impedance (and frequence response) from say 1khz to 20khz. That could also be tackled with a zobel on the speaker.
What I find odd is that on the one hand, without feedback the amp shows incredible detail and realism (overdubs of voices, richness of textures, sounds in the low dynamic range) and you get the impression of hearing "everything on the recording". Downside is that it's fatiguing, has a somewhat gritty tone to it and tonal balance is too bright. I guess too much energy in the presence range and above. Or 3rd order distortion, without 2nd order??
Ciao
Simon
The theory behind it, I believe, is that nfb increases damping factor, allowing to counter the rise in speaker impedance (and frequence response) from say 1khz to 20khz. That could also be tackled with a zobel on the speaker.
What I find odd is that on the one hand, without feedback the amp shows incredible detail and realism (overdubs of voices, richness of textures, sounds in the low dynamic range) and you get the impression of hearing "everything on the recording". Downside is that it's fatiguing, has a somewhat gritty tone to it and tonal balance is too bright. I guess too much energy in the presence range and above. Or 3rd order distortion, without 2nd order??
Ciao
Simon