Also, if you can get 275V supply you could go for back-bias just using a 220 ohm resistor between the first and the second PSU capacitors. The cathode of the second cap is grounded while the cathode of first cap and the bridge are leaft floating. With the full current of the amp flowing through that 220 ohm resistor (say, 100 mA) you will get apprx. your 250V anode voltage and enough bias to play with.
Another solution could be 7x1.5V batteries in series.
Cheers,
45
P.S.
All the 6V6 I have, when put on the Hickok 539C as from manual hit the top of the gm scale at 6 mA/V and in some cases I must switch to 15 mA/V scale...!!!
Another solution could be 7x1.5V batteries in series.
Cheers,
45
P.S.
All the 6V6 I have, when put on the Hickok 539C as from manual hit the top of the gm scale at 6 mA/V and in some cases I must switch to 15 mA/V scale...!!!
Last edited:
I tried back bias and while it did work I felt it was needlessly complex given the intent of this design and I removed it. The current 12V batteries (A23 type) are very inexpensive, extremely compact and will last practically "forever." (3 - 5yrs until replacement) They're pretty ideal IMHO for this application as I really don't want much cathode degeneration. There is just a modicum of cathode bias generated by the 10 ohm cathode resistor which I also use to monitor the operating point. (It is quite stable and around 48mA - 50mA total per tube)
The amplifier is actually running and largely complete - I will replace the 200 ohm resistor in the supply with a choke as I need more ripple suppression. Any other changes will focus mainly on improving the driver stage, feedback implementation and concerns about stability. I may ultimately spring for a better set of OPTs - these are very good, but I suspect I need more primary inductance because of the relatively high rp of the 6V6 even with primary side fbk applied.
Mostly I need to calm down and do some measurements..
My 6V6GTs are U.S. made 1940s vintage Kenrad and NU (National Union) and may differ significantly from modern (foreign) production. They are very close to the stated conditions in the RC-12 tube manual when running at 250V and effectively -12.5V of bias.
I have a Hickok 539B.. Very nice tester..
Last edited:
OK, understood. I have tried a pentode driver for a 2A3 amp (Yamamoto design) with good results. An all pentode amp may be too much for my taste
(triode fanatic here) but you never know until you try. 6J1 sounds good in triode mode, don't know about the 6J7 - in case you want to triode connect that one later.
I have a quad of JAN-RCA 6V6GTY 1958 with grey glass and true micanol base (very nice, indeed!
) and a pair of Sylvania VT-107 1940's coke bottle and grey glass.The Europeans test the same, they are Italian Fivre 6V6GT with black glass from the early 70's. They have a slightly bigger bulb than the American versions but it is just this production. In fact the older 1950's type was very close to the RCA in every detail...
All these test around 5.5 to 6 mA/V and 40-42 mA.
Then I have some Russians which test unusually strong up to 7.5 mA/V at 45 mA!
Cheers,
45
OK, understood. I have tried a pentode driver for a 2A3 amp (Yamamoto design) with good results. An all pentode amp may be too much for my taste
Definitely in the triode fanatic camp myself, which is particularly why I designed this amplifier. Good once in a while to re-evaluate one's stance about such strongly held beliefs.
What I am hearing is much better in some ways than I expected, but there are also significant audible deficits. As always a bunch of trade-offs are involved. It is certainly very dynamic and punchy in a way that belies the low power rating. Detail, imaging and depth would all fit the exceptional category. Tonal balance tends to the hard, almost but not quite solid state sounding end of the spectrum. Bass is insufficient, but what there is sounds right enough. It is a little pushy on some material, and fatiguing to listen to on others. (In fairness that particular material was not much better sounding with an all triode amp.)
Makes me think that combining the colorations and strengths of the two might be a winning combination - combining strengths and offsetting weaknesses. Something I will look at after I complete the next amp which is a IT coupled GM70.
Last edited:
I have a quad of JAN-RCA 6V6GTY 1958 with grey glass and true micanol base (very nice, indeed!
The Europeans test the same, they are Italian Fivre 6V6GT with black glass from the early 70's. They have a slightly bigger bulb than the American versions but it is just this production. In fact the older 1950's type was very close to the RCA in every detail...
All these test around 5.5 to 6 mA/V and 40-42 mA.
Then I have some Russians which test unusually strong up to 7.5 mA/V at 45 mA!
Cheers,
45
How do those Russian ones sound?
Definitely in the triode fanatic camp myself, which is particularly why I designed this amplifier. Good once in a while to re-evaluate one's stance about such strongly held beliefs.
It is called now "Fix your knowledge"
By the way, 6J7 and 6SJ7 are very nice long forgotten tubes.
6П6С is a fair 6V6 copy, but has wide spread of parameters like all Russian tubes.
They sound fine. As a power tube I have only listened to guitar amps. In HiFi equipment I have once listened to a fully battery operated preamp of a friend of mine! In fact you can run them nicely at very low plate voltage. That worked at around 30V in triode connection (bias about -1.5V for 3.5 mA) 10K load (LC coupled).
Cheers,
45
Last edited:
I have been in the triode ONLY camp for like forever, except for guitar amps (pentodes just scream better). I was in search of a 100 WPC triode only P-P amp, which was found during the development of the amp discussed in the 6L6GC in AB2 thread.
Then I bought one of Pete Millett's red boards. It contains NO TRIODES, none, nada, zip. I can't stop playing with it. It just rocks better than most other amps. Of course I have hot rodded it into stratospheric power levels (250 WPC) but it sounds great as delivered. It is a different sound than a SET for sure, but a welcome change especially for loud or dynamic music. Don't dismiss the pentode until you try it for yourself.
My first Pyramid amp used triode input, triode Concetina splitter, double triode LTP driver, pentode output.
Second version used pentode-triode (pentode input, triode Concertina splitter), triode LTP driver, pentode output.
The last version uses almost all pentodes: pentode input, triode Concertina, pentode LTP driver, pentode output.
No difference except optimal amplification factor between all 3 versions. All sound as very good PP amps.
Second version used pentode-triode (pentode input, triode Concertina splitter), triode LTP driver, pentode output.
The last version uses almost all pentodes: pentode input, triode Concertina, pentode LTP driver, pentode output.
No difference except optimal amplification factor between all 3 versions. All sound as very good PP amps.
It is called now "Fix your knowledge"
By the way, 6J7 and 6SJ7 are very nice long forgotten tubes.
6П6С is a fair 6V6 copy, but has wide spread of parameters like all Russian tubes.
Good to know, I'm thinking I'd like to try some 6П6С at some point and as you can guess I am hard at work on the "Fix your knowledge" professional self improvement activity.
Good to know, I'm thinking I'd like to try some 6П6С at some point and as you can guess I am hard at work on the "Fix your knowledge" professional self improvement activity.
Each economical crisis in the history forced to fix some knowledges... I hope, a knowledge how to sew SMD rugs to replace the same rugs made half an year before will be "fixed", while good old knowledge how to make goods that last forever will be restored...
Progress (not the rocket)
Made some fairly minor changes to the design today, mostly adding the global feedback from the secondary I mentioned in a previous post.
Turns out the OPTs are phased incorrectly from the factory which is a minor disappointment that was easily dealt with on the secondary side. The fact that the gain increased by a significant number of dB above the LF corner of the transformer was a clue, and it dropped radically below it.. No buzz reduction either. The stability margin was adequate to prevent oscillation despite the addition of a lot of positive feedback from the output: gratifying.
I did some measurements which corroborated all of my off the cuff calculations, basically the gain with the internal loop closed was exactly 18 or 25dB as I had previously calculated (and may have mis-posted in a previous post) The open loop gain is extremely high - you can see hints in the little notes on the schematic. In any case I felt that a closed loop gain of around 14dB is about right for the supporting components in my system. This left me with 11dB margin which reduced the output buzz to about 2.0mVrms and lowered the output z by more than a factor of 3..
Incidentally the amplifier bandwidth measures at -3dB at 40Hz and 55kHz.
I boosted the plate voltage to roughly 270V and the measured cathode current is ~56mA - no red plates. Sounds slightly more dynamic.
The HCC are not very efficient with 5dB of BSC kicking in starting at 600Hz so with the feedback struggling to correct the falling bass response the amplifier now clips audibly on low bass driving these speakers at anything more than 92dB or so.. At reasonable volumes it sounds quite good - transformers need to have more primary inductance, otherwise they are quite good.
Here is the latest schematic, and another picture showing the amplifier with a GZ32 installed.
Note that the only change of significance is the addition of a 2K feedback resistor from the secondary back the 6J7 feedback node.
When the choke comes in I will install that as it will reduce the need for feedback from the secondary of the OPT if I wish to slightly increase the gain. Higher levels of feedback are possible which will extend the LF response and push down the buzz still further, but with the penalty of the amplifier clipping harder and more often on LF material - a trade off that is already on the wrong side of the tipping point IMO..
Made some fairly minor changes to the design today, mostly adding the global feedback from the secondary I mentioned in a previous post.
Turns out the OPTs are phased incorrectly from the factory which is a minor disappointment that was easily dealt with on the secondary side. The fact that the gain increased by a significant number of dB above the LF corner of the transformer was a clue, and it dropped radically below it..
No buzz reduction either. The stability margin was adequate to prevent oscillation despite the addition of a lot of positive feedback from the output: gratifying. I did some measurements which corroborated all of my off the cuff calculations, basically the gain with the internal loop closed was exactly 18 or 25dB as I had previously calculated (and may have mis-posted in a previous post) The open loop gain is extremely high - you can see hints in the little notes on the schematic. In any case I felt that a closed loop gain of around 14dB is about right for the supporting components in my system. This left me with 11dB margin which reduced the output buzz to about 2.0mVrms and lowered the output z by more than a factor of 3..
Incidentally the amplifier bandwidth measures at -3dB at 40Hz and 55kHz.
I boosted the plate voltage to roughly 270V and the measured cathode current is ~56mA - no red plates. Sounds slightly more dynamic.
The HCC are not very efficient with 5dB of BSC kicking in starting at 600Hz so with the feedback struggling to correct the falling bass response the amplifier now clips audibly on low bass driving these speakers at anything more than 92dB or so.. At reasonable volumes it sounds quite good - transformers need to have more primary inductance, otherwise they are quite good.
Here is the latest schematic, and another picture showing the amplifier with a GZ32 installed.
Note that the only change of significance is the addition of a 2K feedback resistor from the secondary back the 6J7 feedback node.
When the choke comes in I will install that as it will reduce the need for feedback from the secondary of the OPT if I wish to slightly increase the gain. Higher levels of feedback are possible which will extend the LF response and push down the buzz still further, but with the penalty of the amplifier clipping harder and more often on LF material - a trade off that is already on the wrong side of the tipping point IMO..
Attachments
My wish too, but unfortunately unlikely to be fulfilled. I am worried about the "rug industry" in this country and my future in it. Seems unlikely to continue even as it is for much longer thanks to Wall St.
You can use an optocompressor trick that I use in my push-pull amps that have lot of feedback for hard clipping. Measure screen grid currents by a LED that is coupled to an input attenuator that uses a light controlled resistor. No clipping, just hard compression instead. Easy, but effective.
Next will try with the Onkens - they're 10dB more efficient than the HCC and may sound ok with this amp now. I'll report back soon!!
The amp is pretty musical with the HCC and given their flaws may sound better with the Onkens..
Edit: Sounds quite decent driving the Onkens, the only small amp I have to do so.. The 2A3 amp that used to live around here and with ~ the same output power couldn't do it all.. My impression is a generally flatter, less bloomy/organic presentation, and obviously not as much bass extension. It sounds just a little dull - surprising actually considering the very wide bandwidth. The hint of pentode hardness I mentioned previously is noticeable but not in your face. Overall it is a bit subtractive of the music overall compared to my reference 300B DHT amp. (Fixed bias, fully regulated supply and 8Wrms per channel)
Incidentally this amplifier cannot do more than a couple of watts rms without running into serious, and I mean serious linearity problems. I'm a little surprised that's not closer to 4 W rms, however I do note in the tube manuals they typically measure the SE output power at 8 - 10% THD for full power.
Since this is intended as an amplifier for a secondary system the performance goals are not as lofty.. It cost more than I intended because I used a lot of good parts in its construction including a fairly expensive chassis. Overall however I spent less than $300 total including the yet to be installed choke. There were a significant number of expensive parts already on hand, so from scratch one could probably build this for under $400 with some care. Better transformers would I think make a significant difference - for what I paid they are excellent.
One major plus is that this is a very small amplifier, one of the most compact I have yet designed. Anyone notice how small it actually is, and not very heavy either.. Power consumption is less than a 60W light bulb too.. As tube amps go it's fairly green..
The amp is pretty musical with the HCC and given their flaws may sound better with the Onkens..
Edit: Sounds quite decent driving the Onkens, the only small amp I have to do so.. The 2A3 amp that used to live around here and with ~ the same output power couldn't do it all.. My impression is a generally flatter, less bloomy/organic presentation, and obviously not as much bass extension. It sounds just a little dull - surprising actually considering the very wide bandwidth. The hint of pentode hardness I mentioned previously is noticeable but not in your face. Overall it is a bit subtractive of the music overall compared to my reference 300B DHT amp. (Fixed bias, fully regulated supply and 8Wrms per channel)
Incidentally this amplifier cannot do more than a couple of watts rms without running into serious, and I mean serious linearity problems. I'm a little surprised that's not closer to 4 W rms, however I do note in the tube manuals they typically measure the SE output power at 8 - 10% THD for full power.
Since this is intended as an amplifier for a secondary system the performance goals are not as lofty..
It cost more than I intended because I used a lot of good parts in its construction including a fairly expensive chassis. Overall however I spent less than $300 total including the yet to be installed choke. There were a significant number of expensive parts already on hand, so from scratch one could probably build this for under $400 with some care. Better transformers would I think make a significant difference - for what I paid they are excellent.One major plus is that this is a very small amplifier, one of the most compact I have yet designed. Anyone notice how small it actually is, and not very heavy either.. Power consumption is less than a 60W light bulb too.. As tube amps go it's fairly green..
Last edited:
Clever, and just that's what's ordered in the sort of applications you run them in.. I need to learn to turn the volume down anyway..
I need to learn to turn the volume down anyway..
So, it is enough to put a bright red LEDs in screen grids.
I put some crappy 0.47uF film caps in parallel with the 100uF main supply caps, and it is audible. Those 100uF caps are premium grade Panasonic snap-in 450V caps. The main improvement seems to be a greater sense of dynamics and a little more image width.
As I think about it the treble also sounds a lot more open and extended.. Hmm..given the previous measurements that does not seem to make much sense, but there you have it.
Slight amount of 120Hz is audible with the GZ32.. That choke will be a nice improvement. Despite the tight quarters there does not appear to be an issue with magnetic coupling from the power transformer into the OPTs - hopefully the choke will not couple into the front end. I will mock it up electrically prior to drilling any holes and mounting it permanently as there are other ways to address this issue.
As I think about it the treble also sounds a lot more open and extended.. Hmm..given the previous measurements that does not seem to make much sense, but there you have it.
Slight amount of 120Hz is audible with the GZ32.. That choke will be a nice improvement. Despite the tight quarters there does not appear to be an issue with magnetic coupling from the power transformer into the OPTs - hopefully the choke will not couple into the front end. I will mock it up electrically prior to drilling any holes and mounting it permanently as there are other ways to address this issue.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.