Hi folks.
New to this forum (hello everybody!
), I am seeking advice on a DAC analog output stage. It will be transformer coupled to a pair PCM63 DAC chips (resistor and stepup-tranny forming a current/voltage converter ensuring low voltage on the DAC outputs, low distortion), and I would like to put to use a pair of 6SN7 tubes I have laying around (a NOS pair of RCA 5692 CRC-JAN red base. So beautifully constructed they inspired me to make them glow 
).
I guess the book approach on this would be a cathode follower (all scematics i've found on tube DACs seems to use CF exept some 6S45P SE), but I am thinking it would be possible with push pull operation, where the stepup-transformer connected to DAC chip has a secondary centertap connected to GND, thus acting as phase splitter (Lundahl LL1636 amorphous core, wired 1:10+10), and using a push-pull preamp OPT perhaps with CT on secondary too for balanced lines.. The attached image shows a very basic circuit employing this push-pull idea.
I do not have very much experience in tube circuitry, but in my understanding the push pull topology would yield a much wider bandwidth than any SE or CF design, and it seems like I will avoid any capacitors in the signal path wich I think is a very good objective..
Can anyone comment on this design? Any thoughts or suggestions before I spend cash doing stupid mistakes?
New to this forum (hello everybody!
), I am seeking advice on a DAC analog output stage. It will be transformer coupled to a pair PCM63 DAC chips (resistor and stepup-tranny forming a current/voltage converter ensuring low voltage on the DAC outputs, low distortion), and I would like to put to use a pair of 6SN7 tubes I have laying around (a NOS pair of RCA 5692 CRC-JAN red base. So beautifully constructed they inspired me to make them glow ).I guess the book approach on this would be a cathode follower (all scematics i've found on tube DACs seems to use CF exept some 6S45P SE), but I am thinking it would be possible with push pull operation, where the stepup-transformer connected to DAC chip has a secondary centertap connected to GND, thus acting as phase splitter (Lundahl LL1636 amorphous core, wired 1:10+10), and using a push-pull preamp OPT perhaps with CT on secondary too for balanced lines.. The attached image shows a very basic circuit employing this push-pull idea.
I do not have very much experience in tube circuitry, but in my understanding the push pull topology would yield a much wider bandwidth than any SE or CF design, and it seems like I will avoid any capacitors in the signal path wich I think is a very good objective..
Can anyone comment on this design? Any thoughts or suggestions before I spend cash doing stupid mistakes?
Attachments
Hi C_L,
You might want to read Lynn Olson's ideas on push pull topologies. His Amity headphone preamp uses a common bias resistor to force balance between the two halves. Of course, you wouldn't need the attenuators in your application.
Hope this helps.
You might want to read Lynn Olson's ideas on push pull topologies. His Amity headphone preamp uses a common bias resistor to force balance between the two halves. Of course, you wouldn't need the attenuators in your application.
Hope this helps.
Thanks for your replies pedroskova, slowmotion.
At first glance I do not quite understand how Lynn Olson's circuit can balance differences betwen the plates with it's single resistor, but I like the simplicity very much. I will do some more reading on his stuff. My 5692 tubes are closely matched, so I guess a bias circuit for matching is not really that important in my first approach anyways...
Christian Rintelen's approach appears to contain some quite interresting elements, but he uses like 200 components where I wish to use just one or two..
CL
At first glance I do not quite understand how Lynn Olson's circuit can balance differences betwen the plates with it's single resistor, but I like the simplicity very much. I will do some more reading on his stuff. My 5692 tubes are closely matched, so I guess a bias circuit for matching is not really that important in my first approach anyways...
Christian Rintelen's approach appears to contain some quite interresting elements, but he uses like 200 components where I wish to use just one or two..
CL
I've been working on my digital stage the last week, so the analog output stage has been halted for a while as my digital stage gets in shape. I am however building the Lundahl 1:10+10 stepup transformer (mentioned in previous post) into the digital part (as DAC chip output transformer), so that the output stage will be fed by balanced lines on XLR connectors (yes, external output stage). I guess the eventual output stage will get simple 1+1:1+1 interstage-trannys at its inputs.
As I use no digital filters in my DAC, all filtering needs to be done in an analog manner with inductors (coils and trannys) and capacitors. My previous filter setup for my Crystal CS8412 digital reciever SPDIF input used a 0.22uF coupling capacitor and an 1:1 input transformer. I have now changed this and adopted the principle of balanced lines to my SPDIF signal lines (very experimental), by adding another identical 1:1 transformer, and wiring them 1+1:2. A new XLR connector replaces the old 2-pole SPDIF connector. My hope is that balancing the SPDIF line will lower the noise enough to allow me to remove the 0.22uF capacior from the circuit. Time will show.
Needless to say, a balanced SPDIF line needs an unbalanced to balanced converter in the transmitting end of the cable. This adapter also uses a simple 2:1+1 transformator conversion from 2-pole to 3-pole.
That's all for now. I am kept busy with studies at the moment (every moment), so progress is slow. Very slow.
You got me right
As I use no digital filters in my DAC, all filtering needs to be done in an analog manner with inductors (coils and trannys) and capacitors. My previous filter setup for my Crystal CS8412 digital reciever SPDIF input used a 0.22uF coupling capacitor and an 1:1 input transformer. I have now changed this and adopted the principle of balanced lines to my SPDIF signal lines (very experimental), by adding another identical 1:1 transformer, and wiring them 1+1:2. A new XLR connector replaces the old 2-pole SPDIF connector. My hope is that balancing the SPDIF line will lower the noise enough to allow me to remove the 0.22uF capacior from the circuit. Time will show.
Needless to say, a balanced SPDIF line needs an unbalanced to balanced converter in the transmitting end of the cable. This adapter also uses a simple 2:1+1 transformator conversion from 2-pole to 3-pole.
That's all for now. I am kept busy with studies at the moment (every moment), so progress is slow. Very slow.
pedroskova said:
I take it you're not a vinyl kinda guy.
You got me right
Hello folks,
I have been thinking about using four 6S45P tubes (single triodes) instead of my two 6SN7 (dual triodes). This will apparently give a gain of about 100, and I wonder; This should be sufficient to drive headphones directly from the DAC output stage, no?
I've got ca. +/-0.2V swing across the input transformer secondary, the triodes will amplify the signal to about +/- 20V across the output transformer. I will probably run the 6S45P tubes by a 3K5 load, so the output transformer primary will be 3K5+3K5 with B+ on centertap, but what output voltage swing is desired on the OPT secondary when it's purpose is to drive headphones?
Can anyone give me some advice on how to calculate the OPTs for this application?
I have no specific target headphones in mind...
I have been thinking about using four 6S45P tubes (single triodes) instead of my two 6SN7 (dual triodes). This will apparently give a gain of about 100, and I wonder; This should be sufficient to drive headphones directly from the DAC output stage, no?
I've got ca. +/-0.2V swing across the input transformer secondary, the triodes will amplify the signal to about +/- 20V across the output transformer. I will probably run the 6S45P tubes by a 3K5 load, so the output transformer primary will be 3K5+3K5 with B+ on centertap, but what output voltage swing is desired on the OPT secondary when it's purpose is to drive headphones?
Can anyone give me some advice on how to calculate the OPTs for this application?
I have no specific target headphones in mind...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.