R7-resistor or Choke instead? I say:Aviod Choke beacose Choke want to enhance sound negative(slightly `metalize` the sound,and want to pick up noise/hum if is not properly placed or designed)),like ewery Iron in `path` of sound(OPT special did that negative sound degeneration ).Actualy (Theoreticaly) best way for PSU is DC Batery PSU for all amp stages including filament suply(practicaly to Expencive).But we live in real world of AC110/220v main suply with all advantages/disvantages of that.BTW OTL Concept is to avoid the Iron(OPT) in sound `Path`(Choke is Iron Device),remember PSU stage is Integral part of that OTL Circlotron(High Resolution)Tube amps.Good luck
Last edited:
The power supply is set up so that we get a certain amount of current for V4- a choke would put the voltage a tad high. If properly bypassed, at any power level/signal condition all you get on the plate of V4 is a straight line, which is what you want. We've tried regulating this point, as it can make a difference, but as long as the amp is operating A2 regulation has no audible/measureable effect. OTOH if the amp is biased for AB (with a slightly higher output tube B+), regulating the plate of V4 can have a profound effect on how good the amp sounds.
We use a choke for the V4 cathode resistors as we are trying to get all the voltage we can.
We use a choke for the V4 cathode resistors as we are trying to get all the voltage we can.
So Class-A2 operation allows a simpler power-supply topology while Class-AB operation was the motivation behind the rather costly B+ driver-stage regulator during the time period that you had a distributor requirement for Class-AB? In Class-AB mode, did differences in the output-stage grid-drive requirements place additional demands on the driver-stage B+? I'm just trying to understand the means by which Class-AB mode was modulating the driver-stage B+...
Thanks! I was curious about the mechanism by which the driver's B+ was being affected in Class-AB mode.
OK, this means that a word of caution is in order for those who were interested in enabling the Class-A/Class-AB switch; you'll need to address the need to regulate the driver B+ power-supply. The VB-408 three-pin regulator from ST Microelectronics that was included in the original M-60 signal-patch schematic is "non-stocked" at most of the DIY suppliers and was priced at a bit over $120/unit in small quantities.
An exclusively Class-A configuration is sounding better all the time...
Last edited:
This is the Reg the Atma guys told me to hunt for. they are unobtanium i guess. I have a pre-amp power supply board i need these for.
I have been thinking about Nelsons Pass's Super Symmetry System and wondering if the same techniques couldn't be applied to the Diff Pair of the Atma front end?
I have been thinking about Nelsons Pass's Super Symmetry System and wondering if the same techniques couldn't be applied to the Diff Pair of the Atma front end?
With the M-60 global NFB loop engaged, the M-60 essentially forms a Super-Symmetry topology, at least in my estimation...
According Nelson Pass' patent documentation (http://passdiy.com/pdf/articles/susy.pdf), "Super-Symmetry works by exploiting the complementary characteristics of precision matched balanced circuits to differentially reject distortion and noise, and extends this symmetry to make the distortion and noise virtually identical on each half of a balanced amplifying circuit. This gives as much as a 100:1 reduction in unwanted signal components without requiring the equivalent amount of negative feedback. It is simply much easier to tweak the two halves of the circuit into symmetry than to eliminate the distortion in each half of the circuit."
The M-60's two-stage topology, with just a single highly-optimized differential-pair as the input stage (cascoded and fed by a cascoded current-source), feeding a balanced Circlotronic cathode-follower output stage, certainly meets the "simple signal-path" criteria for maximizing the effectiveness of Super-Symmetry. The "light touch" global NFB in the M-60 design (only about 1-2dB of negative feedback) also aligns with the tenets of minimizing global NFB in order to preclude the generation of higher-order distortion components, as noted by Peter Baxandall and John Linsey-Hood in Nelson Pass' whitepaper on the interplay between distorion and feedback (http://passdiy.com/pdf/articles/distortion_feedback.pdf).
Of course, Ralph Karsten recommends running the M-60 "open-loop", noting the loss of subtle musical performance cues within the sound reproduction when the global NFB loop is engaged...
With the M-60 global NFB loop engaged, the M-60 essentially forms a Super-Symmetry topology, at least in my estimation...
According Nelson Pass' patent documentation (http://passdiy.com/pdf/articles/susy.pdf), "Super-Symmetry works by exploiting the complementary characteristics of precision matched balanced circuits to differentially reject distortion and noise, and extends this symmetry to make the distortion and noise virtually identical on each half of a balanced amplifying circuit. This gives as much as a 100:1 reduction in unwanted signal components without requiring the equivalent amount of negative feedback. It is simply much easier to tweak the two halves of the circuit into symmetry than to eliminate the distortion in each half of the circuit."
The M-60's two-stage topology, with just a single highly-optimized differential-pair as the input stage (cascoded and fed by a cascoded current-source), feeding a balanced Circlotronic cathode-follower output stage, certainly meets the "simple signal-path" criteria for maximizing the effectiveness of Super-Symmetry. The "light touch" global NFB in the M-60 design (only about 1-2dB of negative feedback) also aligns with the tenets of minimizing global NFB in order to preclude the generation of higher-order distortion components, as noted by Peter Baxandall and John Linsey-Hood in Nelson Pass' whitepaper on the interplay between distorion and feedback (http://passdiy.com/pdf/articles/distortion_feedback.pdf).
Of course, Ralph Karsten recommends running the M-60 "open-loop", noting the loss of subtle musical performance cues within the sound reproduction when the global NFB loop is engaged...
M60-two stage Amp,thats wrong,You forgot the DC coupled driver stage.Thats third(second actualy) stage.BTW there is posible to implement other topology in the` front end`like ordinary 2x cathode follower(one double triode splited in two CF coupled via 0,1uf to driver stage).This aproach is good only for XLR(balanced)input,( in that case there is no more phase inverter for RCA input).But Whole Amp is stell duall diferencial from input to output.(I experyment with that concept of input stage,sounds good).Good Luck
The DC driver stage has no voltage gain, which is what is usually referred to when someone talks about a gain stage, as it is the voltage gain that defines the input sensitivity of the amplifier. If you are going to include current gain as a gain stage, then the amp has 3 stages of gain: one voltage gain stage and 2 current/follower stages.
We developed a device called the PSVB408 that is a monolithic high voltage adjustable regulator IC, with half the dropout voltage of the VB408 and also lower noise. However we don't use it in the amps, since the AB feature has faded with time. Even with the regulator the amp in AB is only a sonic shadow of the amp in A2 mode.
Last edited:
I know sonic diference betwen A2 mode vs AB mode,but I use 4x6c33c-b in Multiphase Concept(full floathing 4 phase bridge) in output stage with separate 4 driver DC coupled(bipolar +- 350v suply) so there is litlebeat easyer for implementation of diferent input stage and to switch output stage to A2 mode easy,Ab mode is only good compromise to enhance output power and lower the heat generated from output stages tubes.dual diferencial cascode gain stage work OK100%,but I try during experimentation to implement alternative solution for input stage(Why not to try), CF input stage sounds OK when I drive the input with tube preamp via XLR (diferencial),I try also for input stage 2xtriode Grounded Grid configuration with 2x Fet in cathodes-Sounds amazing good,and some other configuration.Good Luck
Last edited:
If you take out V2. What are the minimum speaker requirements that will work?
I have never found chokes in the power supply to have any negative effects in audio. In fact I like using choke input in the power supply. Even in solid state amps I find they work very well. They are best if use in a class A amps. If any thing a large choke makes the amp smoother and have a greater sense of ease. The choke act like a large fly wheel that stabilizes the power supply. It increases regulation. At the same time heavily filters the mains. After a choke input you do not measure a crazy wicked saw tooth ripple wave form. It takes stress off the power transformer and diodes. They run cooler. And slows in rush current. Extending the life of you capacitors. As with any power supply you must take care were you place your components. Off course if you want to minimize the influence of your power supply then your going to have put it in a separate chassis a few feet away. Or add some shielding around it if its all in one chassis.
I have never found chokes in the power supply to have any negative effects in audio. In fact I like using choke input in the power supply. Even in solid state amps I find they work very well. They are best if use in a class A amps. If any thing a large choke makes the amp smoother and have a greater sense of ease. The choke act like a large fly wheel that stabilizes the power supply. It increases regulation. At the same time heavily filters the mains. After a choke input you do not measure a crazy wicked saw tooth ripple wave form. It takes stress off the power transformer and diodes. They run cooler. And slows in rush current. Extending the life of you capacitors. As with any power supply you must take care were you place your components. Off course if you want to minimize the influence of your power supply then your going to have put it in a separate chassis a few feet away. Or add some shielding around it if its all in one chassis.
V2 have no relationship with speaker isue,V2 tube is `grounded` grid based tube and is the integral part of cascoded gain stage.It is posible to remove V2 for less sensitivity isue (gain),but I think in major cases this is not necesarity.(and why to `mess`so perfect dual diff.cascode gain stage/inverter) About chokes:Sometime chokes have resonant frequency with filter and produce subsonic `Wave`in DC line,this afect reproducing mechanism negative(special in output stage under peak curent condition in asociated DC line(phase).CRCRCCC filter configuration is far better acording My opinion.BTW My Amps use MULTIPHASE Concept(four phase balanced bridge OTL),each phase have separate suply,thats is actualy only disvantage of this Concept (150V-0,2-0,3A idlle per phase depending in which class want to run output stage).So there is some advantage beacose idlle (and peak) curent per each four phase is lower than in two phase Concept(traditional Circlotron),I use only 3100uf capacitor/360v=,with 0,1uf and uf5408 paralled per phase,this Fast SS uf5408 diode have good impact in reproduction mechanism(improve AC power transfer thru the each phase.For input/drive stage I dont use choke but for M60 maybe is necesarity.I use CRC configuration and have no `problemo`.In the mid time I try to colect parts for 8phase Full floathing balanced bridge Circlotron(8x6c33c-b). 1000v thats `dangerous` be carefull . Good Luck
Last edited:
I am trying to do a kind of S60 or 2x60W estereo amp (8x6as7+46sn7) per channel for it see for the next trafos....
Fils 6sa7 1x8412 (4x12V16A)
Fils and + -B's of 6sn7 2t230 (2x230V0,37A+2x6.3V4A)
6as7 bridges's 2x6450(100V5,5A)
According to my calculations the current will be sufficient for 120W, is ok?
Some modification in the PS or in another side?
Fils 6sa7 1x8412 (4x12V16A)
Fils and + -B's of 6sn7 2t230 (2x230V0,37A+2x6.3V4A)
6as7 bridges's 2x6450(100V5,5A)
According to my calculations the current will be sufficient for 120W, is ok?
Some modification in the PS or in another side?
Interestingly enough, the folks at Atma-Sphere Music Systems actually built a relevant demonstration project in which the 300B was incorporated into the Circlotronic output stage (Atma-Sphere Music Systems - Papers). The 300B-based prototype illustrates that directly-heated triodes can be used within the M-60 topology and it would be interesting to see what you hear with a 211-based M-60 OTL implementation...
Keep in mind that an OTL bypasses the usual output transformer, so the output-stage vacuum-tubes themselves must interface directly with the loudspeaker load. The 6AS7 has been popular in OTL designs due to its inherent ability to provide suitable output power to the loudspeaker when reasonably paralleled; the venerable 211 exhibits a substantially different set of power-delivery characteristics, so there will likely be challenges in optimizing the 211 in a Circlotronic OTL output stage.
Here we are back at the title of this thread
the 211 is not suitable for an OTL as it will not operate well with only 150V or so on the plate and handle any great deal of current without saturation. The 300b was challenging enough, in that application the tubes were really 6300bs, a graphite plate version of the tube, and the grids were biased at +15V to make the tube conduct enough to be worthwhile. The 300b actually still has some of its curve left to use in this mode, but 4 tubes only made about 10 watts, where 4 6AS7Gs will make about 20. Rp plays a **huge** role in an OTL!
Got it!
I had no plans of using 150V on a 211
I will try the 6AS7 tube first.
What is the the typical range the 50K pots should be set to start for the 6AS7 tube? And what are we looking for the meters?
I would not use a 50K pot- it will give you too much range. That's why you see the 5K and 25K pots in the schematic. If you are wanting to use a tube with greater mu (6C33, EL509), the thing to do is to change the value of the resistor that goes to ground from the bias control.