Hi,
I would like to share some experiences and would like to get your view / hypothesis to improve as well.
I just finished two projects:
- A BIII with Xmos input, PLacid HD Supply, Unbalancer from Broskie build with Texas Comonents, Mundorf PSU CAps, VCap CU and JJ 802S and Ecc99. All CAlbled with Mnundorf 0,5mm Gold-Silver Wire. A nice 20H Choke in the PSU of the unbalancer.
- A modifified EE Minimax (ESS9018 based) with mtech USB-Input board, modified with BUrson OPamps, Elna Silmics, NIce WBT-Nextgen Outs etc. I use the OPamp-Output as the TUbe Out is a poor design.
After 500 hours of break-in now, this is what I hear:
- The BIII with tube-out sounds much more like solid state than the eemax: It is far more crisp, extreme transparent, but as well somewhot "technical" sounding, tight, the tones have not really enough relaxed freedom to vibrate
- The eemax is softer in a very deleicate, nice way. It is much more musical than the BIII. It has a certain "graininess", it is not as dead quite as the BIII, it has not this technical sound, is more emotional. Tones are vibrating better and are more realistic.
The BIII sounds more like a Amp with heavy feedback, while the EE sounds like a nice DHT.
Well...that was certainly not what I expected.
The design goal was: Best of both worlds. Obviously the ESS9018 can sound very special, as demonstrated in the EE minimax. Very transparent, but as well very musical. I wanted to preserve this, but add the transparence and naturalness of a good tube-stage. I wanted to get rid of the layer of graininess and noise and wanted to lift the still existing curtain.
Some theories / hypothesis I am coming up to follow-on ( and here I would appreciate A LOT your help not to run in circles and to discuss which is the right hypothesis):
1. The Mundorf Silver/Gold Cable with only 0.5mm is too small in diameter and sound nice in trebles, but cant transport the signal
2. The V-Cap CUs put a technical signature to the wole thing. The Unbalancer has two Caps in the Signal Path, while the EE Minimax in Solid ( in my modded version) has none.
3. The JJ tubes have this kind of sound
4. I hear here the nature of passive I/V-conversion vs active i/v conversion: The eemax has active, while the unbalancer needs a 50 ohm resistor
5. The mundorf power cap are polyprops, we hear the PSU here
6. The BIII board sound more sterile (doubt that)
Which of these hypothesis would you advise me to test first ?
Update:
The other difference is that the m2tech evo in the minimax plus works only with kernel streaming while the Xmos accepts both and was running under Asio. With Asio you have a more crystalline, airy, but tigtly controlled sound while Xmos with Kernel Streaming is as well more Grounded and now for the first time more realistic. It seems that Asio was needed before to compensate for a muffy output stage, while with the super-transparent Broskie-Stage the Asio-Character is simply to much.
I still want more "vibrations", but this was already a good step in the right direction.
I would like to share some experiences and would like to get your view / hypothesis to improve as well.
I just finished two projects:
- A BIII with Xmos input, PLacid HD Supply, Unbalancer from Broskie build with Texas Comonents, Mundorf PSU CAps, VCap CU and JJ 802S and Ecc99. All CAlbled with Mnundorf 0,5mm Gold-Silver Wire. A nice 20H Choke in the PSU of the unbalancer.
- A modifified EE Minimax (ESS9018 based) with mtech USB-Input board, modified with BUrson OPamps, Elna Silmics, NIce WBT-Nextgen Outs etc. I use the OPamp-Output as the TUbe Out is a poor design.
After 500 hours of break-in now, this is what I hear:
- The BIII with tube-out sounds much more like solid state than the eemax: It is far more crisp, extreme transparent, but as well somewhot "technical" sounding, tight, the tones have not really enough relaxed freedom to vibrate
- The eemax is softer in a very deleicate, nice way. It is much more musical than the BIII. It has a certain "graininess", it is not as dead quite as the BIII, it has not this technical sound, is more emotional. Tones are vibrating better and are more realistic.
The BIII sounds more like a Amp with heavy feedback, while the EE sounds like a nice DHT.
Well...that was certainly not what I expected.
The design goal was: Best of both worlds. Obviously the ESS9018 can sound very special, as demonstrated in the EE minimax. Very transparent, but as well very musical. I wanted to preserve this, but add the transparence and naturalness of a good tube-stage. I wanted to get rid of the layer of graininess and noise and wanted to lift the still existing curtain.
Some theories / hypothesis I am coming up to follow-on ( and here I would appreciate A LOT your help not to run in circles and to discuss which is the right hypothesis):
1. The Mundorf Silver/Gold Cable with only 0.5mm is too small in diameter and sound nice in trebles, but cant transport the signal
2. The V-Cap CUs put a technical signature to the wole thing. The Unbalancer has two Caps in the Signal Path, while the EE Minimax in Solid ( in my modded version) has none.
3. The JJ tubes have this kind of sound
4. I hear here the nature of passive I/V-conversion vs active i/v conversion: The eemax has active, while the unbalancer needs a 50 ohm resistor
5. The mundorf power cap are polyprops, we hear the PSU here
6. The BIII board sound more sterile (doubt that)
Which of these hypothesis would you advise me to test first ?
Update:
The other difference is that the m2tech evo in the minimax plus works only with kernel streaming while the Xmos accepts both and was running under Asio. With Asio you have a more crystalline, airy, but tigtly controlled sound while Xmos with Kernel Streaming is as well more Grounded and now for the first time more realistic. It seems that Asio was needed before to compensate for a muffy output stage, while with the super-transparent Broskie-Stage the Asio-Character is simply to much.
I still want more "vibrations", but this was already a good step in the right direction.
Last edited:
Hi Blitz
Interesting observations. I have recently build a BuffaloIII with Paul hynes regs and the unbalancer. The unbalancer is setup with the same JJ tubes like you. coupling caps auricaps and RTI film. I was also surprised with the sound - to me it gives a too hifi like sound with too much emphasis on air and treble extension - the 3d image is amazing I must say. But Sometimes I feel it sounds a bit artificial. Definately also not what I expected.
I tried to change the auricaps to russian PIO k40y-9 to get a more relaxed full bodied sound. This helped a little. Next I will try change JJ ecc802s to see what it brings.
Did you manage to tweak the unbalancer to there more warm natural side?
I am really interested in hearing other people's impressions with the buffalo dac and unbalancer combo - especially also if they have tried other I/V stages with buffalo and compared it to the unbalancer
Best regards
Søren
Interesting observations. I have recently build a BuffaloIII with Paul hynes regs and the unbalancer. The unbalancer is setup with the same JJ tubes like you. coupling caps auricaps and RTI film. I was also surprised with the sound - to me it gives a too hifi like sound with too much emphasis on air and treble extension - the 3d image is amazing I must say. But Sometimes I feel it sounds a bit artificial. Definately also not what I expected.
I tried to change the auricaps to russian PIO k40y-9 to get a more relaxed full bodied sound. This helped a little. Next I will try change JJ ecc802s to see what it brings.
Did you manage to tweak the unbalancer to there more warm natural side?
I am really interested in hearing other people's impressions with the buffalo dac and unbalancer combo - especially also if they have tried other I/V stages with buffalo and compared it to the unbalancer
Best regards
Søren
Update:
upgraded my BUffalo III to Dual Mono and therefore could lower the Input resistor to 20 Ohm. I included now as well Ian's Fifo with Dual-clocks and Crystek 957.
Already a big step forward, each time.
But the biggest step which really brought 3D and Msic and all the technical sounding stuff away: Changed from V-Cap CU to Duelund CAST. WOW. WOW.WOW.
In many places I have put in Z-Foils, but for the I/V-Resistor the only really clear and highly resolving resistor has been the wirewound Neoohm 0.1% precision resistor. Much more attack and detail, "LIve"-Feeling. I will test some more resistors for the I/V, but the Neoohm is currently leading the pack. Really impressive what the Unbalancer is capable of if you select carefully the parts.
upgraded my BUffalo III to Dual Mono and therefore could lower the Input resistor to 20 Ohm. I included now as well Ian's Fifo with Dual-clocks and Crystek 957.
Already a big step forward, each time.
But the biggest step which really brought 3D and Msic and all the technical sounding stuff away: Changed from V-Cap CU to Duelund CAST. WOW. WOW.WOW.
In many places I have put in Z-Foils, but for the I/V-Resistor the only really clear and highly resolving resistor has been the wirewound Neoohm 0.1% precision resistor. Much more attack and detail, "LIve"-Feeling. I will test some more resistors for the I/V, but the Neoohm is currently leading the pack. Really impressive what the Unbalancer is capable of if you select carefully the parts.
Hi Blitz
Which 12bh7 did you end up using - did you use current production or NOS?
Best regards
Rolle
I did not a big swapping contest there yet. Currently I use a TAD. Which is surprisingly good. But I need to revisit this as the whole character of the Unbalancer changed dramatically due to the other parts changes.
To give you an exapmple: I thought that I need to intregrate a stronger L-Pad before the tweeter when using the V-cap, as is was to treble-biased. Now with the Duelunds, you nearly would say: NOt necessary anymore...plus getting MORE transparency. Sound a bit weird, but they are much faster and better than V-CAp. It is not 13 bs 14. It is 13 vs 18. (humblehifi-Score-points.) The Duelund CAST arre a must with the Unbalancer. You will ne very surprised. BTW they are on sale at hificompents.uk
To give you an exapmple: I thought that I need to intregrate a stronger L-Pad before the tweeter when using the V-cap, as is was to treble-biased. Now with the Duelunds, you nearly would say: NOt necessary anymore...plus getting MORE transparency. Sound a bit weird, but they are much faster and better than V-CAp. It is not 13 bs 14. It is 13 vs 18. (humblehifi-Score-points.) The Duelund CAST arre a must with the Unbalancer. You will ne very surprised. BTW they are on sale at hificompents.uk
Last edited:
I'm not surprised about the cap results, Duelund has never let me down.
A friend of mine got really nice results using "Shinkoh" resistors for I/V
(BuffaloIII / Feastrex transformer output). That may be worth a try.
What cap are you using for C1 ?
Buy the way I'm still waiting for the new batch of unballancer pcb's,
seems like I have been waiting for ever
.Thanks just checked the site (amplifier valve kits, HIFI pre-amplifiers, speaker kits,AMP Parts, upgrade components - right?) Would love to try those duelund cast pio, but even on sale the price is
for caps. Will try the TAD 12bh7 though - thanks for the tip.I joined OPC's new take on the pass d1 group buy - waiting for components to arrive. Interesting to see how it compares to the unbalancer. Will probably wait with the expensive modding until I managed to compare the two I/V's
I' am also using a Buffalo III with an Broskie Unbalancer circuit. For the unbalancer I used Z-Foil TX 2575 resistors, ony the I/V ones are Neohm wirewound. As coupling caps I'm using V-CAP CuTf between the stages and a Mundorf Silver/Oil bypassed by V-Cap CuTf at the output of the Unbalancer. I never got the chance to try Duelund CastPio, but already thought about it. I tried a lot of different tubes, even Siemens CCa and Telefunken E88cc. For me by far the best combination are Voskhod 6n23p from the 70s or early 80s as input tubes and Tungsram E80CC as cathode followers. I didn't find any alternative that came close. For me a big improvement was to use a battery(nimh) for biasing the input tubes.
Yes, for the 6n23p I needed more than the given 10mA. I never tried a rial ccs circuit for biasing, but even though I'm using clean power with salas sshv2, using a battery was still a big improvement. I thought it would be critical to use a battery in a differential amp, but its absolutely not, no hum or anything at all.
Doesn't matter what tube I used, with the battery bias it was always a much cleaner and smoother sound.
Blitz,
You are right on WRT Duelund caps. I can also recommend the VSF, they are
cheaper and still in same league of performance.
WRT tubes, I think NOS Mullard or Amperex are the go, not sure who used to
make BH7's.
WRT resistors, old tantalum + AB carbon comp are also worth trying.
This whole equation with Sabre is so damn complex, it will go right back to
the DAC board itself, layout, bypass, power supplies etc.
Sounds like you have it running pretty well though.
Goodstuff
Yes, a lot. It is unbelievable transparent and musical wit the ability to lift your whole High-End-Equipment. It gives you a basis to devlop speakers, anplifiers etc to the next level. But the integration of the FiFo was as well important. Today the EE has no chance.
Think about it: I can hear now the "right" or "wrong" how a resistor sounds. I hear not only "different", but what is right or wrong, which is a new quality for my chain. An unbelieable 3D-Effect which sounds very natural, "simply there". At the same time very good, fluid tonality.
Think about it: I can hear now the "right" or "wrong" how a resistor sounds. I hear not only "different", but what is right or wrong, which is a new quality for my chain. An unbelieable 3D-Effect which sounds very natural, "simply there". At the same time very good, fluid tonality.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.