I have something of a conundrum regarding my BII.
I've always had the ambition to equip it with a tube buffer but I ended up with a quick and dirty solution using transformers.
Now, I've lived with this for quite a while and I got to thinking about finishing the project once and for all. However I recently started asking myself why.
The Buffalo II in itself will put out 16mA (pp) of current if needed but using it in voltage mode as I am this will never be needed.
Tinkering around with some old tube schematics I realised that the BII is just as powerful as a small tube would be in a buffer configuration. The tube won't have a lower output impedance than the dac either.
Why bother with building an active buffer if the Zout will be just as high or higher and the Iout is about the same or lower?
I like the tube glow but is there really any reason to deviate from the transformer output that I have today? Apparently it works pretty well with the 100kOhm load that my amps represent.
My amps will clip at 2Vrms and use a balanced connection with true differential architecture.
Is there any reason to go active? My gut feeling is that active should be better but the numbers doesn't seam to support that feeling, am I missing something?
I've always had the ambition to equip it with a tube buffer but I ended up with a quick and dirty solution using transformers.
Now, I've lived with this for quite a while and I got to thinking about finishing the project once and for all. However I recently started asking myself why.
The Buffalo II in itself will put out 16mA (pp) of current if needed but using it in voltage mode as I am this will never be needed.
Tinkering around with some old tube schematics I realised that the BII is just as powerful as a small tube would be in a buffer configuration. The tube won't have a lower output impedance than the dac either.
Why bother with building an active buffer if the Zout will be just as high or higher and the Iout is about the same or lower?
I like the tube glow but is there really any reason to deviate from the transformer output that I have today? Apparently it works pretty well with the 100kOhm load that my amps represent.
My amps will clip at 2Vrms and use a balanced connection with true differential architecture.
Is there any reason to go active? My gut feeling is that active should be better but the numbers doesn't seam to support that feeling, am I missing something?
Hi,
You are right about the output impedance of a straight tube buffer being highish, unless you are using a tube design with feedback you output impedance will be around the same or higher than the DAC itself.
The only compelling reason to use a straight tube buffer would be because you enjoy the color of the sound coming out. Nothing wrong with that. 🙂
I think using a transformer straight out will also yield good results.
If you could find a tube stage that uses super-symmetry (fully symmetrical operation) then that would be a great match - but it would likely still need buffers if really low output impedance is your goal.
Whatever stage you choose you need to filter out unwanted quantization noise. Since your gear is balanced input - this will be less of an issue for you as long as it has really good CMRR.
The output stages I have designed are designed to have good CMRR and include filtering to massively reduce the quantization noise.
Cheers!
Russ
You are right about the output impedance of a straight tube buffer being highish, unless you are using a tube design with feedback you output impedance will be around the same or higher than the DAC itself.
The only compelling reason to use a straight tube buffer would be because you enjoy the color of the sound coming out. Nothing wrong with that. 🙂
I think using a transformer straight out will also yield good results.
If you could find a tube stage that uses super-symmetry (fully symmetrical operation) then that would be a great match - but it would likely still need buffers if really low output impedance is your goal.
Whatever stage you choose you need to filter out unwanted quantization noise. Since your gear is balanced input - this will be less of an issue for you as long as it has really good CMRR.
The output stages I have designed are designed to have good CMRR and include filtering to massively reduce the quantization noise.
Cheers!
Russ
Oh, I realized I have another question regarding the BII.
I've been running it with the standard placid psu but I'm having a hard time finding data as to any recommendations for the power supply ripple.
A buddy of mine is doing some designing on a psu at the moment and asked me to try it out for him. What would the target spec be for psu ripple? I'll target 5.5VDC and the psu should be good for 1A but how good a dc voltage do I need? 100uV, 10uV, 1uV, ...?
I've been running it with the standard placid psu but I'm having a hard time finding data as to any recommendations for the power supply ripple.
A buddy of mine is doing some designing on a psu at the moment and asked me to try it out for him. What would the target spec be for psu ripple? I'll target 5.5VDC and the psu should be good for 1A but how good a dc voltage do I need? 100uV, 10uV, 1uV, ...?
I also have questions about output circuit for BII DAC. Hope it's ok to tag them onto this thread.
After years of using BII with current o/p using IVY yesterday I tried a voltage out circuit with 1 opa627 per channel. I find I really like the sound of voltage o/p - to me it sounds more effortlessly dynamic & detailed, very spacious & very deep bass.
I wonder Russ, if you have done any further experimentation with different types of o/p methods and, if so, if these is any documentation of this anywhere.
thanks
mike
After years of using BII with current o/p using IVY yesterday I tried a voltage out circuit with 1 opa627 per channel. I find I really like the sound of voltage o/p - to me it sounds more effortlessly dynamic & detailed, very spacious & very deep bass.
I wonder Russ, if you have done any further experimentation with different types of o/p methods and, if so, if these is any documentation of this anywhere.
thanks
mike
B-II, R2 and R5 LED resistors value?
Hey guys, could not find a general B-II thread anymore, so I figured here was as good as anywhere, sorry for the OT question.
Just want to know what the original value and case size is for the LED resistors on the B-II board, I am rejuvinating mine for an experimental project, thx!
Hey guys, could not find a general B-II thread anymore, so I figured here was as good as anywhere, sorry for the OT question.
Just want to know what the original value and case size is for the LED resistors on the B-II board, I am rejuvinating mine for an experimental project, thx!
Last edited:
A value from 221R to 1K should work fine (depends on how bright you want to go). I can't remember if the case size is 0805 or 1206 I will check later today.
Last edited:
Thanks
Russ and Brian, for some reason I am not getting my usualy notifications from diya... Anyway, ended up going with Susumu 0805 681R, cleaned everything up, and the old B-II is working with 45.1854 clock on board. Yeah, now to direct wire the Ethernet interface...
Russ and Brian, for some reason I am not getting my usualy notifications from diya... Anyway, ended up going with Susumu 0805 681R, cleaned everything up, and the old B-II is working with 45.1854 clock on board. Yeah, now to direct wire the Ethernet interface...
- Status
- Not open for further replies.