Wow i can't believe i have missed this Thread! first off Hi Ralph!
I am building a set of Atma-Clones myself. But I plan on building a set of 16-20 tube mono blocks. I hope to achieve 200 watts per which should be easily accomplished as the 20 tube atma makes 200 watts.
For the output HV i found some NICE toroid transformers. I hate to give this secret away but Apex jr. has some Basler 16-16431 500w transformers with dual 52V secondaries that when wired in series would give 104Vac which should provide 147V rectified. and the best part is...Apex is selling them for $25 each! that's damn cheap! you will need a pair but still $50 is a great deal. And these transformers were designed for an audio amp to start with so there pretty nice.
I have considered a SMPS for my amps as well. but not for the HV. for the 50+ amps of filament current i will need per mono block!!! I have managed to acquire all the parts i need for my amps surplus thus far EXCEPT the dang filament transformers. (I keep hoping to find a stash of vintage Triump Motorcycle parts i can trade to Ralph for some trafo's), However Ralph told me he recommends running AC on the filaments so that sort of ruled out my thinking for SMPS. however i do have some really nice monster sized 200 amp mosfets from an industrial sized UPS system that i bet i could drive with a oscillator circuit to re-AC the DC from a SMPS but...were getting off into the deep end there.
I need to figure what i am going to do for the filament supplies and then i can start on the chassis work.
I have had 2 old Atma prototypes and they sound great. the first one was something like 30 years old. the second something like 20 years old. and it still runs great and sounds awesome!
I look forward to seeing how this develops. I am quite interested.
Zc
I am building a set of Atma-Clones myself. But I plan on building a set of 16-20 tube mono blocks. I hope to achieve 200 watts per which should be easily accomplished as the 20 tube atma makes 200 watts.
For the output HV i found some NICE toroid transformers. I hate to give this secret away but Apex jr. has some Basler 16-16431 500w transformers with dual 52V secondaries that when wired in series would give 104Vac which should provide 147V rectified. and the best part is...Apex is selling them for $25 each! that's damn cheap! you will need a pair but still $50 is a great deal. And these transformers were designed for an audio amp to start with so there pretty nice.
I have considered a SMPS for my amps as well. but not for the HV. for the 50+ amps of filament current i will need per mono block!!! I have managed to acquire all the parts i need for my amps surplus thus far EXCEPT the dang filament transformers. (I keep hoping to find a stash of vintage Triump Motorcycle parts i can trade to Ralph for some trafo's), However Ralph told me he recommends running AC on the filaments so that sort of ruled out my thinking for SMPS. however i do have some really nice monster sized 200 amp mosfets from an industrial sized UPS system that i bet i could drive with a oscillator circuit to re-AC the DC from a SMPS but...were getting off into the deep end there.
I need to figure what i am going to do for the filament supplies and then i can start on the chassis work.
I have had 2 old Atma prototypes and they sound great. the first one was something like 30 years old. the second something like 20 years old. and it still runs great and sounds awesome!
I look forward to seeing how this develops. I am quite interested.
Zc
There is an interesting page that details building a M-60 kit that has some good info and descriptions etc.
Building the Atma-Sphere M-60 Mk II
Building the Atma-Sphere M-60 Mk II
I just had carpal-tunnel surgery, so I'm short a hand for a few weeks.
I'm just about done with a revised M-60 signal-path schematic that includes the manual-adjustment bias/DC-offset circuit; I've altered the circuit to provide sufficient adjustment range to accommodate the 6AS7's as well as a number of the other output-tube options being discussed in the thread (6C33's, etc.). The automatic bias/DC-offset schematics will have to wait for my hand to recover from the carpal-tunnel surgery...
6FQ7/6CG6 are considered "9 pin 6SN7". I used them in what started out as a SUN 2A3 "clone" (with 6B4G)
dave
Last edited:
M-60 OTL Revised Signal-Path Schematic (v1.1)
Attached is the revised M-60 signal-path schematic, incorporating an alternative manual output-stage biasing circuit. The new circuit elements enable:
The output-stage biasing circuit is sufficiently low in impedance that it should not introduce any potential for modulation effects on the driver stage.
The automatic output-stage bias circuit and automatic output-stage DC-balance circuit additions that I had previously offered to cook-up for the group will have to wait until I've recovered from my recent carpal-tunnel surgery and caught-up on my day-job workload; a one-handed engineer faces a number of productivity challenges...
Attached is the revised M-60 signal-path schematic, incorporating an alternative manual output-stage biasing circuit. The new circuit elements enable:
* Use of Alternative Output Tubes - the bias-set potentiometers enable a broad range of output-tube bias voltages, enabling the use of a variety of output tubes. Allowing for the additional 6-8VDC of negative bias necessary to support the driver tube bias needs, the effective bias range for the output tubes is -40VDC to -87VDC
* Line-independent Bias Supply - the use of the Zener diodes to establish the bias-voltage range makes the bias-supply largely independent of any AC-line voltage variations; I may add a solid-state current-source to further isolate the bias-voltage circuit from AC-line disturbances
* Class-A/AB Switch - since a number of thread members are interested in supporting the ability to switch between Class-A and Class-AB output-stage operations, I've included that functionality in the signal-path schematic
* Line-independent Bias Supply - the use of the Zener diodes to establish the bias-voltage range makes the bias-supply largely independent of any AC-line voltage variations; I may add a solid-state current-source to further isolate the bias-voltage circuit from AC-line disturbances
* Class-A/AB Switch - since a number of thread members are interested in supporting the ability to switch between Class-A and Class-AB output-stage operations, I've included that functionality in the signal-path schematic
The output-stage biasing circuit is sufficiently low in impedance that it should not introduce any potential for modulation effects on the driver stage.
The automatic output-stage bias circuit and automatic output-stage DC-balance circuit additions that I had previously offered to cook-up for the group will have to wait until I've recovered from my recent carpal-tunnel surgery and caught-up on my day-job workload; a one-handed engineer faces a number of productivity challenges...
Attached is the revised M-60 signal-path schematic, incorporating an alternative manual output-stage biasing circuit. The new circuit elements enable:
* Use of Alternative Output Tubes - the bias-set potentiometers enable a broad range of output-tube bias voltages, enabling the use of a variety of output tubes. Allowing for the additional 6-8VDC of negative bias necessary to support the driver tube bias needs, the effective bias range for the output tubes is -40VDC to -87VDCThe output-stage biasing circuit is sufficiently low in impedance that it should not introduce any potential for modulation effects on the driver stage.
* Line-independent Bias Supply - the use of the Zener diodes to establish the bias-voltage range makes the bias-supply largely independent of any AC-line voltage variations; I may add a solid-state current-source to further isolate the bias-voltage circuit from AC-line disturbances
* Class-A/AB Switch - since a number of thread members are interested in supporting the ability to switch between Class-A and Class-AB output-stage operations, I've included that functionality in the signal-path schematic
The automatic output-stage bias circuit and automatic output-stage DC-balance circuit additions that I had previously offered to cook-up for the group will have to wait until I've recovered from my recent carpal-tunnel surgery and caught-up on my day-job workload; a one-handed engineer faces a number of productivity challenges...
WOW!!!!!!!!!!!!!!!! Thank you!
How is your recovery progressing? Oh yeah....the great catch up tango..
Good Job Mullard!
I have a question and a comment. 1- the coupling caps C2 and C3 are critical in this circuit. there value determines the low freq cut off point and the type of cap used here and the cap type makes a big difference. Atma uses Teflon caps and they are very expensive, something like $50 each so use the best quality cap you can afford here.
2- Why did you choose to increase the cathode resistors for the output tubes? I have had this discussion with the Atma guys previously. as i understand it increasing those resistors will actually lower distortion a bit. but why? Now i can sort of understand that it would help promote current sharing amongst tube sections similar to Emitter resistors in a SS amp. But the output impedance if these amps is already pretty high, wouldn't we want to use the lowest value resistors we can here to try and drop the output impedance as much as possible?
I have a question and a comment. 1- the coupling caps C2 and C3 are critical in this circuit. there value determines the low freq cut off point and the type of cap used here and the cap type makes a big difference. Atma uses Teflon caps and they are very expensive, something like $50 each so use the best quality cap you can afford here.
2- Why did you choose to increase the cathode resistors for the output tubes? I have had this discussion with the Atma guys previously. as i understand it increasing those resistors will actually lower distortion a bit. but why? Now i can sort of understand that it would help promote current sharing amongst tube sections similar to Emitter resistors in a SS amp. But the output impedance if these amps is already pretty high, wouldn't we want to use the lowest value resistors we can here to try and drop the output impedance as much as possible?
tympani1d,
You're quite welcome! I'm very interested in seeing this project through to completion. I'm continuing to gather all of the necessary components and parts; now that I have most of the major physical components, I've begun the process of engineering the physical implemention of the circuit (chassis, cooling, etc.). With all of the "big iron" pieces (power transformers, power-supply filter chokes, etc.) and the need to manage a significant heat load, the chassis design will be large-scale in nature.
With regards to the hand surgery, the post-op cast comes off next week, so that's when I'll be able to start the massive "catch-up" effort at work; then I can finally get back to performing a bit of audio-engineering...
Zero Cool,
Agreed, in such an elegantly simple signal-path as the M-60 design, all of the components traversed by the signal have a significant impact on the sound. The V-Cap TFTF (Teflon Film Tin Foil) capacitors have been the current high-watermark for coupling-capacitor performance, but V-Cap has recently introduced their CFTF (Copper Foil Teflon Film) capacitor as a new assault on state-of-the-art capacitor design (unfortunately, at about twice the price per capacitor...). My initial protoboard auditions (single pair of 6AS7 output tubes via balanced Sennheiser HD-600 headphones) demonstrated that the V-Cap TFTF capacitors sound considerably better than the other film capacitors that I have on-hand (Solen polypropylene, Jantzen Z-Superior polypropylene and AuriCap Teflon), but that's an opinion...
While the V-Cap TFTF caps are the current optional capacitors at Atma-Sphere, I'd be interested to know if any one has compared them with the new V-Cap CFTF capacitors.
That design choice wasn't mine; the increased value of the output-tube cathode resistors is the recommendation of Ralph himself and represents the current production M-60 configuration. As you noted, the increased resistor value promotes more effective current-sharing among the paralleled output tubes as well as resulting in lower overall distortion levels. It's true that the higher value of cathode resistors will increase the effective output impedance of the M-60 power-amplifier, but that's a design trade-off that's balanced against the reduced distortion (and improved sound, according to members of the M-60 owners' forum). The Zero autoformers offer a practical means of addressing the need to mate the M-60's with lower-impedance loudspeakers.
This is DIYaudio, so feel free to build-out the circuit with alternative output-tube cathode resistor values and determine which set of trade-offs best match the needs of your audio system.
Well that certainly explains why the resistor change sounded so familiar! Ralph has told me the same thing but...I was curious why you choose to do it as well hahahah!
I am actually building a pair of MA2 type 4 chassis mono blocks. 20 6AS7's each channel. but much of the front end is the same so I am following this thread closely. I have collected all the parts with the exception of the filament transformers...I need 6.3V at 50++ amps just to tun the 20 output tubes. and I have not found an affordable option thus far.
Zc
I am actually building a pair of MA2 type 4 chassis mono blocks. 20 6AS7's each channel. but much of the front end is the same so I am following this thread closely. I have collected all the parts with the exception of the filament transformers...I need 6.3V at 50++ amps just to tun the 20 output tubes. and I have not found an affordable option thus far.
Zc
Zc,
I'm also trying to include any "lessons learned", particularly those acquired during the multi-decade-long M-60 production run...
Wow! That has to be one massive pile of parts lying about in your workshop! Very cool!
As to filament power-supply transformers, I certainly understand and empathize with the difficulty in locating high-current 6.3VAC power transformers, but a rational/not-insane number of paralleled transformers meets the heater needs of the M-60. But you're facing a situation that might make it reasonable to consider having custom toroid power-transformers fabricated. I've had excellent results with Toroid Corporation of Maryland (Toroidal Transformers for power supplies and isolation in medical, instrument, audio and control panel transformer applications.), so I'd certainly recommend them.
The MA-2 tube complement includes six 6SN7's (in addition to the phalanx of twenty 6AS7's); would I be correct in assuming that the extra pair of 6SN7's comprise additional paralleled driver stages to handle the fan-out to the 20 6AS7's?
Here is a link to a chassis that may work for a 4 tube 6C33B output tube.
Design Build Listen Ltd.
And a sub chassis that can hold 8 octal output tubes JJ EL509 or 6AS7G. Also, looks like plenty of room on this frame to attach other parts. This link below takes you to the page. And then click on "valve mounting panel" to see this part.
More Information
I have ordered parts from the before and the service is good.
Design Build Listen Ltd.
And a sub chassis that can hold 8 octal output tubes JJ EL509 or 6AS7G. Also, looks like plenty of room on this frame to attach other parts. This link below takes you to the page. And then click on "valve mounting panel" to see this part.
More Information
I have ordered parts from the before and the service is good.
Given that the production M-60 employs AC-powered filaments, I assume that Ralph has experiences that have informed that position. However, most of my vacuum-tube pre-amplifier designs have benefited from the use of DC filament power-supplies (both from a sonic standpoint as well as empirical measurements). So, I'm not planning to exclude DC power-supplies from consideration for the M-60 small-signal tubes (differential-pair and driver tubes); I want to evaluate the circuit with both flavors of filament power-supplies.
In my experiences, I have found it difficult to partner SMPS switching power-supplies with analog circuits. The primary problem has been keeping all of the high-frequency switching noise from negatively impacting the analog signal path. Even my experiments with Class-D PWM power-amplifier modules have yielded far more musical results when powered via linear power-supplies (and external clock signals to time-sync the Class-D modules).
So, there's three 6SN7 driver tubes fanning-out to 10 pairs of 6AS7's? Is there a "secret handshake" mapping of driver tubes to output tubes that keeps everything symmetric?
I have not heard Teflon caps. But, I do like Jensen copper and oil for coupling caps. Also, I don't like metal film caps in audio. Even in the power supply. I have had great results with Takman carbon films. My bias is for smooth, coherent and proper tone in music. I can't tolerate anything that is thin, glares and squawks or is dead, dark and veiled. I have heard plenty of tube amps and speakers that do both.
The M-60 driver stage has been a particular point of focus in the refinement of the product design. Ralph Karsten has been iteratively refining the M-60 power-supply for the driver stage over several decades and the current power-supply configuration reflects that ongoing, multiple-decade product-voicing effort. As a student of the audio-engineering art, I'm quite interested to hear the composite voicing of all of those years of production experience.
Of course, this is the DIYaudio forum, so we are individually free to experiment with design alternatives and reach our own conclusions within the context of our individual music systems. Personally, I plan to start with the "as-is" power-supply configuration, but I'm also interested in the sound (and measured performance) with a suitable choke in place of R7...
As noted a bit earlier in the thread, the incredibly elegant and simple M-60 signal-path topology really magnifies the sonic contribution of all components (passive and active). As the only capacitors in the M-60 signal-path, the coupling capacitors between the differential-pair and the driver/follower tubes significantly affect the overall voicing of the design. As a result, these passives afford a massive opportunity to tune/tailor the overall musical presentation of the M-60. You can certainly "season to taste"...
I have also employed Jensen copper-and-oil capacitors to good effect in a number of designs over the years; the Jensen's are absolutely a "classic" that consistently delivers the musical goods. But don't "throw the baby out with the bath-water" with regards to film capacitors; these V-Cap Teflons are often a delightful revelation. The new V-Cap CFTF (Copper Foil and Teflon Film) might be just the Teflon capacitor to align with your musical preferences...