Some dwimmer knocked that sign down and ran. It lay there for a week until my daughter figured I would want it and dragged it around to my shop.
It went really well. It is the best sounding amp I have built. I did make some small changes from ideas in this thread. I am on the road just now. I plan to post a schematic when I get home. I built it on a large frame so it's easy to make alterations and I have swapped several parts making small changes. The interesting part is that the amp is very robust, so small changes to values haven't made much in the way of differences.
Cathode resistance 880ohm 25w, isn't it a lot?
What potentiometer value (if placing one) would be correct? I know the standard is 100k, but is it the best for CD players? in my case (220ohm rotel)of course
What potentiometer value (if placing one) would be correct? I know the standard is 100k, but is it the best for CD players? in my case (220ohm rotel)of course
A 25 Watt self bias resistor that is dissipating 3.3 Watts, just like in this 2A3 Amp, runs very cool (the large surface area of the resistor).
If there is very little real estate (room) under the chassis, then you can lay a 105 degrees-C bypass capacitor directly on the resistor.
I hate it when I have to put parts too close to a Hot resistor.
Instead, use a physically large [cool] resistor.
$0.03
Adjusted for Inflation
If there is very little real estate (room) under the chassis, then you can lay a 105 degrees-C bypass capacitor directly on the resistor.
I hate it when I have to put parts too close to a Hot resistor.
Instead, use a physically large [cool] resistor.
$0.03
Adjusted for Inflation
I used a 100k log taper pot and it works well with a Marantz DVD player and a 600 ohm signal generator. I don't see any reason to change it. If you have any problems check the grounding, it needs to be done right!
This 6C6 in pentode mode connection is a big problem in that amplifier.
The gain is huge, yhus the dynamic capacitances are large. with 600ohm input impedance of the source it giving phase shift of about -35deg@20KHz. with 1M pot it will be disaster with low levels when a part of the pot adding resistance to the source...
.
Also output has very high output impedance of about 68K@1KHz which is not enough to frive 2A3...
Distortion is about 3.2% in most cases.
-A=80-90 depending on Rload.
.
The input sensitivity is about 1Vp-p for aprox. 90Vp-p.
Some standard value of digital sources is 2Vrms something smaller than 6Vp-p !!!
So this amp with 6C6 as input tube deserve in start 6:1 attenuation
If we link the preamp at the input it will be total disaster, pot will be at the minimum for max 2A3 power.
As in s many tube designs - the main issue is too sensitive inputs...
.
This tube is not to be in the input stage for 2A3 or most of the power tubes in my opinion.
.
There is a good model in the tolic spice models. Please downloard and simulate.
.
6C6 spice models
.
I made 2 settings values with almost the same thd. But these one with lower RL value is little better because of the smaller amplification.
.
If we can say that about 90Vp-p is need for 2A3, and we have output from digital devices as 6Vp-p
then 90Vp-p/6Vp-p=15 times amplification is enough for input tube. And settings to -4Vug as safe margin.
With amplification factor of the tube as 20-22. And lower Ri to be useful as driver.
That is opposite of 6C6 in pentode mode.
.
But in triode mode it will be much better as input in contemporary tube amp?
🙂
The gain is huge, yhus the dynamic capacitances are large. with 600ohm input impedance of the source it giving phase shift of about -35deg@20KHz. with 1M pot it will be disaster with low levels when a part of the pot adding resistance to the source...
.
Also output has very high output impedance of about 68K@1KHz which is not enough to frive 2A3...
Distortion is about 3.2% in most cases.
-A=80-90 depending on Rload.
.
The input sensitivity is about 1Vp-p for aprox. 90Vp-p.
Some standard value of digital sources is 2Vrms something smaller than 6Vp-p !!!
So this amp with 6C6 as input tube deserve in start 6:1 attenuation
If we link the preamp at the input it will be total disaster, pot will be at the minimum for max 2A3 power.
As in s many tube designs - the main issue is too sensitive inputs...
.
This tube is not to be in the input stage for 2A3 or most of the power tubes in my opinion.
.
There is a good model in the tolic spice models. Please downloard and simulate.
.
6C6 spice models
.
I made 2 settings values with almost the same thd. But these one with lower RL value is little better because of the smaller amplification.
.
If we can say that about 90Vp-p is need for 2A3, and we have output from digital devices as 6Vp-p
then 90Vp-p/6Vp-p=15 times amplification is enough for input tube. And settings to -4Vug as safe margin.
With amplification factor of the tube as 20-22. And lower Ri to be useful as driver.
That is opposite of 6C6 in pentode mode.
.
But in triode mode it will be much better as input in contemporary tube amp?
🙂
That's super interesting and I will try triode mode. However, I must point out that the original designers(s) knew what they were doing as the design as I built it (schematic when I return home) sounds terrific. In addition, it is quite robust in that small changes to component values have little impact on the sound.
Tip of the hat to Joe Roberts.
Tip of the hat to Joe Roberts.
I made more acurate model for triode connected 6C6 and 6J6 tubes.
It is a good tube and it could be used in triode mode for many designs.
BUT the Vb is very high for a best perfprmace with Rload, about +550V, to much dissipation on the Rload...
.
With a L Load it is lower Vb BUT again Lp has to be super high of minimum 150Hy for good phase and transfer in LF.
Distortion is very low for large output in p-p.
.
Can accept 10K source input impedance as generator.
Still capacitances are a bit too havy...
.
It is a good tube and it could be used in triode mode for many designs.
BUT the Vb is very high for a best perfprmace with Rload, about +550V, to much dissipation on the Rload...
.
With a L Load it is lower Vb BUT again Lp has to be super high of minimum 150Hy for good phase and transfer in LF.
Distortion is very low for large output in p-p.
.
Can accept 10K source input impedance as generator.
Still capacitances are a bit too havy...
.
Code:
**********************************************
* Created on 10/08/2024 11:19 using paint_kit.jar 3.1
* www.dmitrynizh.com/tubeparams_image.htm
* Plate Curves image file: 6C6 triode connection GE
* Data source link: 6C6 PDF Gen Elec.
*----------------------------------------------------------------------------------
.SUBCKT 6C6_GE_triode 1 2 3 ; Plate Grid Cathode
+ PARAMS:
+ MU=21.43
+ KG1=2475
+ KP=156
+ KVB=300
+ VCT=0.2
+ EX=1.472
*--------------- added 0.7pF for each socket connection
+ CCG=3.7P
+ CGP=2.7P
+ CCP=11.2P
*--------------- capacitances from datas in triode mode
* + CCG=3P
* + CGP=2P
* + CCP=10.5P
*---------------
+ RGI=2000
*----------------------------------------------------------------------------------
* Vp_MAX=700 Ip_MAX=20 Vg_step=2.5 Vg_start=2.5 Vg_count=15
* Rp=50000 Vg_ac=4.5 P_max=1.75 Vg_qui=-5 Vp_qui=200
* X_MIN=67 Y_MIN=51 X_SIZE=941 Y_SIZE=675 FSZ_X=1821 FSZ_Y=815 XYGrid=true
* showLoadLine=y showIp=y isDHT=n isPP=n isAsymPP=n showDissipLimit=y
* showIg1=n gridLevel2=n isInputSnapped=n
* XYProjections=y harmonicPlot=y dissipPlot=y
*----------------------------------------------------------------------------------
E1 7 0 VALUE={V(1,3)/KP*LOG(1+EXP(KP*(1/MU+(VCT+V(2,3))/SQRT(KVB+V(1,3)*V(1,3)))))}
RE1 7 0 1G ; TO AVOID FLOATING NODES
G1 1 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG1}
RCP 1 3 1G ; TO AVOID FLOATING NODES
C1 2 3 {CCG} ; CATHODE-GRID
C2 2 1 {CGP} ; GRID=PLATE
C3 1 3 {CCP} ; CATHODE-PLATE
D3 5 3 DX ; POSITIVE GRID CURRENT
R1 2 5 {RGI} ; POSITIVE GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS 6C6_GE_triode
*$what about the screen resistor and screen cap in the original schematic?
Where do you put it in the 6c6 connected in triode mode?
If I take them out I have no gain
I am using the model you passed in LTSpice
Where do you put it in the 6c6 connected in triode mode?
If I take them out I have no gain
I am using the model you passed in LTSpice
Zoran,
Thanks for your comments and models. I'll have lots to study about this design when I get home.
In the first page of this thread I posted about my SE2A3 amp using a plate choke loaded 6A6 driver circuit. Click here for details and schematic.
Here's the triode connected 6C6 version loaded by 200H nickel plate chokes by dave slagle of Intact Audio.
This one uses a 7193/2C22 triode + discontinued James JS4120N 200H nickel plate choke + 2A3 and Hashimoto H203S OPT which I use in my mono hifi rig.
Actually, the linestage of my mono rig also uses a James 4120N plate choke + triode connected 6C6 voltage amp.
I learned about choke-capacitance coupling from reading the Radiotron Designers Handbook third edition from the 1930s, when engineers only had slide rulers and listened to live unamplified musicians to aid their design efforts.
A Juilliard-trained pianist who visited my man cave actually preferred the pentode 6C6 driven 2A3 amp over the plate loaded/triode'd 6C6 version. 🙂
Even if I can justify spending for dave slagle's nickel plate chokes, I highly recommend experimenting with the affordable Hammond 156C 150H@8mA plate choke. Just keep your current below the 8mA max to maintain good fidelity.
Computer simulations are great tools but perhaps there's no substitute to old school building and listening to find out which circuit enhances one's musical pleasure.
As I type I'm listening to my Altec 755Cs in Silbatone boxes driven by my Radiotron SE2A3 amp with Finemet FM6WS.
Happy listening!
This one uses a 7193/2C22 triode + discontinued James JS4120N 200H nickel plate choke + 2A3 and Hashimoto H203S OPT which I use in my mono hifi rig.
Actually, the linestage of my mono rig also uses a James 4120N plate choke + triode connected 6C6 voltage amp.
I learned about choke-capacitance coupling from reading the Radiotron Designers Handbook third edition from the 1930s, when engineers only had slide rulers and listened to live unamplified musicians to aid their design efforts.
A Juilliard-trained pianist who visited my man cave actually preferred the pentode 6C6 driven 2A3 amp over the plate loaded/triode'd 6C6 version. 🙂
Even if I can justify spending for dave slagle's nickel plate chokes, I highly recommend experimenting with the affordable Hammond 156C 150H@8mA plate choke. Just keep your current below the 8mA max to maintain good fidelity.
Computer simulations are great tools but perhaps there's no substitute to old school building and listening to find out which circuit enhances one's musical pleasure.
As I type I'm listening to my Altec 755Cs in Silbatone boxes driven by my Radiotron SE2A3 amp with Finemet FM6WS.
Happy listening!
Last edited:
I omit them in triode mode.
I connect the screen grid to the plate via a 100 Ohm resistor.
If You read the text in Radiotron book You will realise thatin that time totaly diferent electronic standards was valid. Amp has very limited BW just little abowe 10KHz and probably 50-100Hz LF -3db. That is by tubes and RC coupling filters onnly - if we add Output transformer it will be worse.
...
And if we take in account this huge value pots, that in low levels (making additionak R for the source) with dynamic capacitances creating huge HL filter...
.
I dont believe based on the exact known parameters of 6C6 that it is a good input for 2A3. And Simulations with spice models are confirming this.
.
And I simple dont believe that some device with huge phase shift at the eges of BW can sound good...
.
By the way:
Hashimoto H203S OPT 9Hy (60mA, 3.5K load) - very insuficient inductance for even decent bass, big cutt-off that could not be compensated with bigger value of Ck and bigger value of C coupling... 🙁
.
Tamura F-7002 25Hy, (100mA, 3.5K), a lot better in terms of primary inductance and much friendly to LF (bass) with 2A3
I will try to emulate for these 2 OT and the results will be pretty obvious...
.
That is just for LF because Lp is the 1st factor. No other datas and no measurements of transferand phase. But other a bit complicated measurements are present...
Behaviour in the HF is as usual, determined with the value of pot. These values that people are using of 100K or even more are completely without taking in a count dynamic capacitances of the input tubes...
THeis is for lower levels of sygnal, part of pot that is serial to the source, becoming a part of RC HF cut/off filter... ;(
Please calculate this value for each specific tube.
.
Computer sims are just for clarify the things that already present in the listening. For me. After all, I made few tube amps and mount few OTs before computer sims, when in the whole internet was 10-20 tube audio sites, without freat PDF tubedatasheet bases, etc.
Just xerox-copied books, measurements, paper and pen calculations, and listening sessions with freaks doing the same...
.
For instance, it was a dream to construct reactive load line by current in transformer. Now it is a easy and pretty cousy to have visual presentation of that effect in the LF as in the HF...
Last edited:
I have RCA LC-1A 4 pieces in folded Woigt horn, central driver is Iskra 14in, HF units are Westrex London 2390A with acoustic lenses. I am planing to add 4 Westrex 2080E. Amps are Quad II (adopted to KT88, I gave my 4 x vintage GEC KT66 to a friend...). For now.
I am thinking to build one new 2A3 SE amp too 🙂. I mount all the iron parts, have PIO C and tubes for more than decade 🙁
zoran,
You said:
"And I simple dont believe that some device with huge phase shift at the eges of BW can sound good..."
Please define huge phase shift.
At 1 kHz, 45 degrees phase shift = 3 inches in air
At 10kHz, 45 degrees phase shift - 0.3 inches in air
I can hear 10kHz.
Are the distances from the left loudspeaker to your left ear, and from the right loudspeaker to your right ear, equal to each other with a 0.3 inch accuracy (such as +0.15 inch on left, and -0.15 inch on right).
Well, If you build a non-negative feedback SE amplifier, the output transformer low frequency response and related phase at low frequency are
One Pole responses:
-1dB: Phase is 26 degrees Leading
And . . .
-3dB: Phase is 45 degrees Leading
Whatever low frequency is -3dB, the -1dB frequency is 2 times higher (one octave higher than the -3dB frequency).
Example -3dB @ 20Hz, is -1dB @ 40Hz.
For the output transformer's high frequency response and phase shift, it is a 2 Pole response:
The phase shifts are even larger, and the frequency for -1dB and -3dB are much closer than 2:1 (<< one octave).
I pity those who can not Enjoy Listening to simple amplifiers that have such "bad" frequency responses and such "bad" phase shifts.
A playback "System" includes amplifiers and Loudspeakers.
For more data on frequency response and phase response of loudspeakers, we all need to investigate the loudspeaker threads of diyAudio.
And then there is the interaction of the loudspeaker impedance versus frequency, and the loudspeaker phase-of-impedance versus frequency . . and how those reactances aid or reduce the effects of the output transformer's secondary inductance, secondary-to-primary leakage reactance, and the primary distributed capacitance that is effectively "multiplied" at the secondary (according to the primary turns / secondary turns).
I was pleased to be at Eduardo B. DeLima's presentation at an early VSAC in Silverdale, WA.
Eduardo also wrote up the concepts in an early "Valve" magazine, published by Doc Bottlehead.
If I worried about all of those 'problems' in this Post # 57, I might not enjoy listening very much; but I very much enjoy listening.
I will only mention the words "Group Delay".
I gave up measuring that when I retired from my job that had all kinds of such measurement equipment.
It is more than I care to do at home . . . measure and calculate manually. And Software and I are the greatest of enemies,
so please do not provide me a link to the software solution.
Just like the bumblebee, our mono and stereo systems keep "flying".
It took engineers decades before they could understand the bumblebee' flight.
And, it has taken engineers decades to understand how our ears, and our highly powerful brain's processing made listening to our mono and stereo music systems "sound" great!
$0.03
Adjusted for Inflation
You said:
"And I simple dont believe that some device with huge phase shift at the eges of BW can sound good..."
Please define huge phase shift.
At 1 kHz, 45 degrees phase shift = 3 inches in air
At 10kHz, 45 degrees phase shift - 0.3 inches in air
I can hear 10kHz.
Are the distances from the left loudspeaker to your left ear, and from the right loudspeaker to your right ear, equal to each other with a 0.3 inch accuracy (such as +0.15 inch on left, and -0.15 inch on right).
Well, If you build a non-negative feedback SE amplifier, the output transformer low frequency response and related phase at low frequency are
One Pole responses:
-1dB: Phase is 26 degrees Leading
And . . .
-3dB: Phase is 45 degrees Leading
Whatever low frequency is -3dB, the -1dB frequency is 2 times higher (one octave higher than the -3dB frequency).
Example -3dB @ 20Hz, is -1dB @ 40Hz.
For the output transformer's high frequency response and phase shift, it is a 2 Pole response:
The phase shifts are even larger, and the frequency for -1dB and -3dB are much closer than 2:1 (<< one octave).
I pity those who can not Enjoy Listening to simple amplifiers that have such "bad" frequency responses and such "bad" phase shifts.
A playback "System" includes amplifiers and Loudspeakers.
For more data on frequency response and phase response of loudspeakers, we all need to investigate the loudspeaker threads of diyAudio.
And then there is the interaction of the loudspeaker impedance versus frequency, and the loudspeaker phase-of-impedance versus frequency . . and how those reactances aid or reduce the effects of the output transformer's secondary inductance, secondary-to-primary leakage reactance, and the primary distributed capacitance that is effectively "multiplied" at the secondary (according to the primary turns / secondary turns).
I was pleased to be at Eduardo B. DeLima's presentation at an early VSAC in Silverdale, WA.
Eduardo also wrote up the concepts in an early "Valve" magazine, published by Doc Bottlehead.
If I worried about all of those 'problems' in this Post # 57, I might not enjoy listening very much; but I very much enjoy listening.
I will only mention the words "Group Delay".
I gave up measuring that when I retired from my job that had all kinds of such measurement equipment.
It is more than I care to do at home . . . measure and calculate manually. And Software and I are the greatest of enemies,
so please do not provide me a link to the software solution.
Just like the bumblebee, our mono and stereo systems keep "flying".
It took engineers decades before they could understand the bumblebee' flight.
And, it has taken engineers decades to understand how our ears, and our highly powerful brain's processing made listening to our mono and stereo music systems "sound" great!
$0.03
Adjusted for Inflation
Last edited:
Hello @6A3sUMMER!
While you were typing your post above, which I agree with 100%, especially the part regarding enjoying music, my wife and I were having fine Indian cuisine with the guy who wrote, "I never met met a 2a3 amp I didn't like."
The long evening was a whirlwind discussion of music, art, museums and life.
Below is a picture of us having an after dinner brew.
Cheers!
JE
While you were typing your post above, which I agree with 100%, especially the part regarding enjoying music, my wife and I were having fine Indian cuisine with the guy who wrote, "I never met met a 2a3 amp I didn't like."
The long evening was a whirlwind discussion of music, art, museums and life.
Below is a picture of us having an after dinner brew.
Cheers!
JE
A pic for the ages. I wonder what his thoughts on 2A3 amps are nowadays.
@dubadub - we didn't talk much about audio gear but he mentioned that he might use the Altec/Peerless A100 clone he published in the last SP issue when he gets his pair of WE757A clone speakers. I think that's a PP2A3 amp.