Yes, DCB1 without coupling caps - this must be good for sound.
So as core it make sence for balance to take 2x DCB1. The question only wich PSU is better: original from DCB1 or UltraBib. Original PSU from DCB1 depends from quality of filtering caps, Russian tephlone ft3 are the best. But UltraBib seems much more independet in quality parts.
Very interesting question.
So as core it make sence for balance to take 2x DCB1. The question only wich PSU is better: original from DCB1 or UltraBib. Original PSU from DCB1 depends from quality of filtering caps, Russian tephlone ft3 are the best. But UltraBib seems much more independet in quality parts.
Very interesting question.
Hello Salas!
Many of my audio components PSU build in paradigm of your shunt regulators. So wanted to have some typical approach in case of DCB1. UltraBib is successor Bib1.1 and looks more modern than original PSU in DCB1. Besides, last one depends on the quality of capacitors, a bit difficult in turning and so on.
So for balanced version the first aim: take 2x cores from DCB1 (2x 4 = 8pcs of 2SK170BL and other sands near them), 1pcs UltraBib + 10V, 1pcs UltraBib - 10V.
Could you clarify some moments:
Many of my audio components PSU build in paradigm of your shunt regulators. So wanted to have some typical approach in case of DCB1. UltraBib is successor Bib1.1 and looks more modern than original PSU in DCB1. Besides, last one depends on the quality of capacitors, a bit difficult in turning and so on.
So for balanced version the first aim: take 2x cores from DCB1 (2x 4 = 8pcs of 2SK170BL and other sands near them), 1pcs UltraBib + 10V, 1pcs UltraBib - 10V.
Could you clarify some moments:
- better to have raw bridge PSU with large amount of capacitors: 4700uF 25V x 4pcs?
- also configure UltraBib on large current: 1A or even 1.5A? May be 600mA will be very good for scene size. More than 600mA may be give abnormal huge scene, what you think?
- does it really needed delay section or it could be omitted?
- does it make sense to add in UltraBib teflon capacitor 0.1-0.22uF between: C1 and R1, C1 and J1 (drain)?
- what else components must be changed for good working DCB1 core?
- it must be 4pcs of UltraBib (+/- 10V) or 2pcs (+ and - 10V) will be enough for balance?
- Yes, better with much capacitance in raw PSU.
- No, in Ubib much spare current doesn't decrease rail modulation by the load like in the classic DCB1 regs. Up to 200mA is good.
- No it doesn't make sense at all. Only accepted bypass is across C2 noise filter Vref capacitor.
-No changes in input & output parallel or series resistors values. Only parts quality subjective choices.
-Four Ubib polarity sections equals dual mono which means more box space and expenses.
Demands a single four secondaries custom Tx or two duals too. Up to you.
Thank you for detail explanations.
As you explained here https://www.diyaudio.com/community/threads/salas-hotrodded-blue-dcb1-build.171715/post-3673538 , it is better to have 2x dual PSU (4 power rails). So best choice 4x UltraBib (2x +10V and 2x -10V), 2x toroid 100VA with 2x 15VAC in each. In this case each UltraBib set with 200mA or 100mA?
In this case complitely insane example of NAIM552 clone:
it reques 8 rails in buffer and 4 in gain section.
As you explained here https://www.diyaudio.com/community/threads/salas-hotrodded-blue-dcb1-build.171715/post-3673538 , it is better to have 2x dual PSU (4 power rails). So best choice 4x UltraBib (2x +10V and 2x -10V), 2x toroid 100VA with 2x 15VAC in each. In this case each UltraBib set with 200mA or 100mA?
In this case complitely insane example of NAIM552 clone:
it reques 8 rails in buffer and 4 in gain section.
Output pairs 2sk170 , copper bar drilled four holes, matched transistors in, fill with epoxy, no drift, dc offset, is rock steady 2.5 mv.
A trick I wanted to share.
Pic 2 was before I noticed drift. Yes yes it's the old design, but still rocks,
A trick I wanted to share.
Pic 2 was before I noticed drift. Yes yes it's the old design, but still rocks,
Attachments
Last edited:
Dumb question time... I made one of these many years ago but lost interest in audio for a few years. Just getting it back out and I didn't label the in/out sockets. I'm assuming the board is fed by the input sockets and then it outputs to the volume control and then the output sockets. Is that right? I spent ages searching for clues online ;-( so thought I'd just ask.
It's plugged in and it's working beautifully but I have one more probably naive question....
I stopped using it a long time ago because I was using an AV receiver to power my six channel LXmini speakers (with subs). Now I have a straight power amp so the DCB1 is attenuating the volume. It sounds a lot better from memory, maybe due to increased power from the amp and gain structure.
The speakers use two miniDSP 2x4 units that take 5-24v power (150mA). So... is it possible to use the 10v supply from the middle of the DCB1 board to power these? Any drawbacks? And if so can I just parallel two sockets from the board?
I stopped using it a long time ago because I was using an AV receiver to power my six channel LXmini speakers (with subs). Now I have a straight power amp so the DCB1 is attenuating the volume. It sounds a lot better from memory, maybe due to increased power from the amp and gain structure.
The speakers use two miniDSP 2x4 units that take 5-24v power (150mA). So... is it possible to use the 10v supply from the middle of the DCB1 board to power these? Any drawbacks? And if so can I just parallel two sockets from the board?
Nice to know that it works and sounds nice. Not only the gain structure is more adequate. This combination probably beats that AVR for performance a lot too. Here's a THD sweep from my DCB1 prototype at 1VRMS output.
Regarding tapping off from the DCB1's shunt regulated PSU to power peripherals its generally doable. Especially digital boxes seem to benefit a lot from a linear supply because no more switching noise bleeds in the ground. We had many positive reports with using L-Adaper for such. Drawbacks aren't obvious to predict but someone must optimally have the lab means to confirm no compromise in the preamp's own spec. Due to possible interference on the rail. But in your case 150mA x2 load is too much for the current limit setting if its a normal dissipation build and not a super hot rod PSU bias job with huge sinks as some folks used to make.
Regarding tapping off from the DCB1's shunt regulated PSU to power peripherals its generally doable. Especially digital boxes seem to benefit a lot from a linear supply because no more switching noise bleeds in the ground. We had many positive reports with using L-Adaper for such. Drawbacks aren't obvious to predict but someone must optimally have the lab means to confirm no compromise in the preamp's own spec. Due to possible interference on the rail. But in your case 150mA x2 load is too much for the current limit setting if its a normal dissipation build and not a super hot rod PSU bias job with huge sinks as some folks used to make.
Thanks Salas for the information and quick reply. I used a 10 ohm 5W resistor so I guess that's a 600mA hotrod?
If there's a risk of it compromising performance I'll leave it be, the idea was mostly for convenience but I also read somewhere that the miniDSP units are susceptible to power supply noise. No doubt there's a better solution, maybe just a better quality wallwart/smps. There are upgrade options for the 2x4 units themselves, or even Nelson Pass' passive crossovers so I won't go chasing rainbows!
If there's a risk of it compromising performance I'll leave it be, the idea was mostly for convenience but I also read somewhere that the miniDSP units are susceptible to power supply noise. No doubt there's a better solution, maybe just a better quality wallwart/smps. There are upgrade options for the 2x4 units themselves, or even Nelson Pass' passive crossovers so I won't go chasing rainbows!