All I have on hand now is direct from DAC to amp or with a Schiit Sys in between. This little pre sounds better than both alternatives.
Details. I hear little details in the music that are not present without it. It really is remarkable.
My BA3 pre is almost done but I've been sidetracked by all my other little projects. Next up is an integrated amp withy this board and a modified ACA Mini that runs with a bipolar supply. I'll hopefully get the BA3 done to use with an F4 and will do a comparison. But who knows when?
I love these Pass designs. What a blend of simplicity, elegance and sound quality. Thank you, thank, you thank you Papa!
Details. I hear little details in the music that are not present without it. It really is remarkable.
My BA3 pre is almost done but I've been sidetracked by all my other little projects. Next up is an integrated amp withy this board and a modified ACA Mini that runs with a bipolar supply. I'll hopefully get the BA3 done to use with an F4 and will do a comparison. But who knows when?
I love these Pass designs. What a blend of simplicity, elegance and sound quality. Thank you, thank, you thank you Papa!
Hi All-- I've built a working version of B1 rev 2 (2019) and now want to optimize it to make it as transparent as possible.My amp (Goldmund Job 225) has very high gain (35db) and is very wide bandwidth and seems to show up ANY changes to anything in the audio chain (Interconnects, power cables, quality of the wire used) . The buffer will be driven by a DAC with 200 ohm output impedance.
In Post 8 marc brown asked: Can competent DIYers remove to improve the sound? Which resistors?
and Zen Mod replied with: R105 in above schematic make it just few ohms
QUESTION 1) I'm assuming that Zen Mod is literally meaning 2 or 3 or 4ish ohms on the output resistor? I intend to use some fancy (expensive) Vishay Naked type transistors and they are pricey so I hope to get it close to right the first time.
QUESTiON 2) Is there any way to tweak the INPUT resistor values to improve sound?
(I'm removing the 25K volume pot that is paralleled with 1Meg ohm resistor. I plan to replace both with a single equivalent resistance value of 24K.. and do volume control in the source). Would decreasing both input resistors (the one shunted to ground and the 1K series) with lower resistors possibly sound better? How about higher impedance on the input? (To those that say just "listen": I don't have a lot of components right now to play around with and I plan to order some high quality resistors).
Thanks!
In Post 8 marc brown asked: Can competent DIYers remove to improve the sound? Which resistors?
and Zen Mod replied with: R105 in above schematic make it just few ohms
QUESTION 1) I'm assuming that Zen Mod is literally meaning 2 or 3 or 4ish ohms on the output resistor? I intend to use some fancy (expensive) Vishay Naked type transistors and they are pricey so I hope to get it close to right the first time.
QUESTiON 2) Is there any way to tweak the INPUT resistor values to improve sound?
(I'm removing the 25K volume pot that is paralleled with 1Meg ohm resistor. I plan to replace both with a single equivalent resistance value of 24K.. and do volume control in the source). Would decreasing both input resistors (the one shunted to ground and the 1K series) with lower resistors possibly sound better? How about higher impedance on the input? (To those that say just "listen": I don't have a lot of components right now to play around with and I plan to order some high quality resistors).
Thanks!
You can remove or reduce the value of the output resistor. If going for a low value just go for a value, 5-10R, exact is uncritical.
You can safely reduce the value of the 1K resistor by half and perhaps get better sound/flatter frequency response. It is there to hinder parasittic oscillation with the source. Adjusting the value of the 1M resistor will not alter sound as far as I can see, it is there to safeguars in case the wiper goes mad if I understand it correctly.
You can safely reduce the value of the 1K resistor by half and perhaps get better sound/flatter frequency response. It is there to hinder parasittic oscillation with the source. Adjusting the value of the 1M resistor will not alter sound as far as I can see, it is there to safeguars in case the wiper goes mad if I understand it correctly.
Thank you andynor and R-K Ronningstad (and indirectly Erno)--
Yes it is about reducing the DACs output impedance and reducing the level... if my DAC has simply a 200ohm output impedance I am assuming that would imply the R at the input shunting current to ground would also be 200 ohm to drop the voltage 6dB. Is that correct?
I am assuming that the current into the JFET gate would be almost nil (as far as calculations). Wouldn't it be helpful to have some R at the input to the gate (to avoid oscillations as andynor implied)? Somehow it seems a bit low to me to have the input resistor shunting to ground at 200 ohms at the input to the buffer? (I have never done buffer/audio design. It seems people like ratios of 1 to 10 for source output impedance to preamp inputs-- but does that not apply here because of the high impedance of the JFET or something?)
Would it make more sense to have a "T" shaped set of 3 resistors? Where R1 at the input is in series with the source, where R2 would be a shunt resistor, and R3 would be the resistance from the shared node to the gate? Would that somehow be better for loading or power dissipation or impedance matching? (I wouldn't know what values to make R1,R2,R3)
I'm asking these questions because I don't really have a feel for input impedance matching or design. Any suggestions are most welcome!
(Oh...and I'm using +12V and -12V as my supply rails... and have matched FETS from the store).
Thanks!
Yes it is about reducing the DACs output impedance and reducing the level... if my DAC has simply a 200ohm output impedance I am assuming that would imply the R at the input shunting current to ground would also be 200 ohm to drop the voltage 6dB. Is that correct?
I am assuming that the current into the JFET gate would be almost nil (as far as calculations). Wouldn't it be helpful to have some R at the input to the gate (to avoid oscillations as andynor implied)? Somehow it seems a bit low to me to have the input resistor shunting to ground at 200 ohms at the input to the buffer? (I have never done buffer/audio design. It seems people like ratios of 1 to 10 for source output impedance to preamp inputs-- but does that not apply here because of the high impedance of the JFET or something?)
Would it make more sense to have a "T" shaped set of 3 resistors? Where R1 at the input is in series with the source, where R2 would be a shunt resistor, and R3 would be the resistance from the shared node to the gate? Would that somehow be better for loading or power dissipation or impedance matching? (I wouldn't know what values to make R1,R2,R3)
I'm asking these questions because I don't really have a feel for input impedance matching or design. Any suggestions are most welcome!
(Oh...and I'm using +12V and -12V as my supply rails... and have matched FETS from the store).
Thanks!
Thank you R-K Rønningstad-- I understand what you are doing... you are making the voltage divider at the input. I'm wondering if it helps to drop the resistor values in half or less (instead of using 25k maybe use 10K and instead of 1K much less than that)? Would that help the sound possibly? All of my interconnects are a helix design which tends to really minimize reflections or oscillations. Is this a better way to drop the voltage 6db rather than using a single resistor as shown in the diagram you provided above (16B)?-- thanks so much! I know everything is a trade off... I'm just trying to get the best mix I can to improve sound and minimize parasitic capacitance and inductance. (It seems anything I put between the DAC and the power supply can be heard and some of the resolution is affected. I'm after as transparent as possible).
Thanks andynor... but wouldn't the input impedance of the buffer, with the two 25K resistors and 1K resistor be close to 50K input impedance -- so dropping each in half would still provide 25K total impedance? But I do feel like the source has a wimpy driver (not discrete devices) so even if it's equivalent to 50K... it sounds like that may be still a good thing? (thank you again!)
Why two 25k’s? The 25k in the original circuit is the pot.
I run mine with a ladder pot, and so the B1 has only the 1k and 1m.
A drawing program with the iPad. Terrible. Paper, pencil and camera would have been better.
HI andynor-- you asked why the two 25K resistors? I believe I can answer that. R-K Rønningstad responded to my inquiry and what might be a topology and values of resistors that he thought might be good in order to be easy for my source (DAC) to drive AND in the process reduce the gain by 6db. I am not using a pot because I want to make the simplest most transparent circuit I can with a minimum of added capacitance and inductance and wiring. (I'm able to control the volume at the source but I have to really crank it's output down if I drive my amp directly). My amp is super sensitive to any input, it maxes at 0.75V and has 35db gain and very wide bandwidth. So I'm trying to get rid of all pots and replace them with the best audio transistors I can find. So the configuration with the two 25K resistors allows dropping the input voltage in half before it reaches the gate of the JFET. (I'm also going to remove the 1Meg resistor because in the diagram, 25K in parallel with 1Meg leaves an equivalent resistance of about 24K -- so I"m going to just leave it at 25K). It isn't what most people want and it's very tuned to my own particular source and amp. I may actually try and reduce the gain even more than 6db. I'm hoping the buffer will allow a better impedance match from my source to amp. (Right now there is some music that sounds kind of lifeless when I directly drive the amp with the DAC, and that is compared to when I was able to drive my amp with a better source).
I see, I see. I should have seen that, just jumped to it - sorry.
Anyways, your goal of a clean signal path will be disrupted by the added (rather highish R) series resistor in the signal path. It may prove a better solution than digital volume control, but if I were you I would test this. You can mount the series 25k resistor with pin sockets, that way you can replace it with jumpers and do A/B testing with digital VC.
Don’t worry about input impedance, standard 25k-ish is good enough with a 125 ratio to your source is good enough for starters. You could mount the second 25k resistor also on pin sockets, to A/B input impedances between ish 25k and 50k.
Anyways, your goal of a clean signal path will be disrupted by the added (rather highish R) series resistor in the signal path. It may prove a better solution than digital volume control, but if I were you I would test this. You can mount the series 25k resistor with pin sockets, that way you can replace it with jumpers and do A/B testing with digital VC.
Don’t worry about input impedance, standard 25k-ish is good enough with a 125 ratio to your source is good enough for starters. You could mount the second 25k resistor also on pin sockets, to A/B input impedances between ish 25k and 50k.