Synchronous Clocking/Hum Update
OK. So first I moved my USB receiver so that it is closer to the B-II, the I2S lines are now about 7 cm, and the masterclock line (on U.FL micro coax) is about the same. At first I tried only terminating the shield at the source end. At this point the hum was much reduced, but still audible with the ear at about 10" from the woofer cone, in both speakers. Now many might be OK with this level of hum (inaudible, even at on the quietest night, at the listening position, and certainly below the noise floor of even my lowest noise recordings), but knowing that there was virtually none before trying the synchronous clocking, I kept tweaking...
I decided to ground the shield for the MC cable at both ends: I carefully soldered a micro bnc to pad R17, and ran a short ground lead to it, then I used a 5 cm U.FL cable between the boards. Now the hum is entirely gone from the R channel, and the L channel only has hum audible with the ear within 2" of the woofer cone. The hum now present is likely due to pickup in the analog output wiring, as the L channel leads are closer to the transformer side of the chassis. I will probably experiment with shielding in the chassis to reduce the hum on the L channel, but it is likely now entirely inconsequential...
I would recommend anyone experimenting with synchronous clocking to go ahead and ground the clock distribution cable at both ends, and to use some U.FL or W.FL coax cables.
Russ, great news on the teleporter. When you say no problems with longish runs of cat 5, do you just mean no dropouts or loss of lock, or have you looked at MC and Bit Clock waveforms on the 'scope and seem no degradation in the waveform? Seems almost to good to be true! Any potential advantage to cat 6 with the additional shielding?
OK. So first I moved my USB receiver so that it is closer to the B-II, the I2S lines are now about 7 cm, and the masterclock line (on U.FL micro coax) is about the same. At first I tried only terminating the shield at the source end. At this point the hum was much reduced, but still audible with the ear at about 10" from the woofer cone, in both speakers. Now many might be OK with this level of hum (inaudible, even at on the quietest night, at the listening position, and certainly below the noise floor of even my lowest noise recordings), but knowing that there was virtually none before trying the synchronous clocking, I kept tweaking...
I decided to ground the shield for the MC cable at both ends: I carefully soldered a micro bnc to pad R17, and ran a short ground lead to it, then I used a 5 cm U.FL cable between the boards. Now the hum is entirely gone from the R channel, and the L channel only has hum audible with the ear within 2" of the woofer cone. The hum now present is likely due to pickup in the analog output wiring, as the L channel leads are closer to the transformer side of the chassis. I will probably experiment with shielding in the chassis to reduce the hum on the L channel, but it is likely now entirely inconsequential...
I would recommend anyone experimenting with synchronous clocking to go ahead and ground the clock distribution cable at both ends, and to use some U.FL or W.FL coax cables.
Russ, great news on the teleporter. When you say no problems with longish runs of cat 5, do you just mean no dropouts or loss of lock, or have you looked at MC and Bit Clock waveforms on the 'scope and seem no degradation in the waveform? Seems almost to good to be true! Any potential advantage to cat 6 with the additional shielding?
I mean both. There is no degradation at all that I can detect.
LVDS is perfect for the application.
No need for CAT6 but any twisted pair type cable will work pretty well.
Russ, Brian my Buffalo II has problem with locking S\PDIF. I listened the dac half year without any problem. I tried everything. Different USB to S\Pdif converters, regular CD...
Finally I disconnected everything and powered the dac from battery. The power amplifier was also disconnected from main supply.
The laptop was running on battery and the LOCK diode still blinking. Is there anything I could try to solve the problem?
Thanks,
Angel
Finally I disconnected everything and powered the dac from battery. The power amplifier was also disconnected from main supply.
The laptop was running on battery and the LOCK diode still blinking. Is there anything I could try to solve the problem?
Thanks,
Angel
Synchronous Clocking
Doing this is described elsewhere in this thread, go back and take a look for pictures, Bunpei has been doing this for awhile with the SD Transport. First, you need an I2S source with masterclock output, and I would not consider trying this unless you are confident that your source is low jitter, as using a synchronous clock eliminates the ASRC and its "jitter reduction" on the ESS chip.
To provide a synchronous clock, first depower the oscillator on the B-II board, by either removing the clock Trident, or if not using Tridents, removing the associated inductor on the underside of the board (see trident instructions for finding this inductor). Then the Masterclcok signal lead can be soldered to the pad at R17 between the ESS chip and the Oscillator on the board-use your meter to learn which is the right pad for the clock input to the ESS chip. As mentioned previously, I recommend using a coax cable for clock distribution (U.FL works well) and grounding the shield at both ends. Keep the clock distribution cable as short as possible (as well as the other I2S leads).
I am no expert, but this is working well for me. If more expert readers feel a need to comment, or have opinions on "better" ways to do this, please chime in!
BTW, I am curious why my B-II has no resistor at R17, is this a change that was made, or should R17 be present?
Doing this is described elsewhere in this thread, go back and take a look for pictures, Bunpei has been doing this for awhile with the SD Transport. First, you need an I2S source with masterclock output, and I would not consider trying this unless you are confident that your source is low jitter, as using a synchronous clock eliminates the ASRC and its "jitter reduction" on the ESS chip.
To provide a synchronous clock, first depower the oscillator on the B-II board, by either removing the clock Trident, or if not using Tridents, removing the associated inductor on the underside of the board (see trident instructions for finding this inductor). Then the Masterclcok signal lead can be soldered to the pad at R17 between the ESS chip and the Oscillator on the board-use your meter to learn which is the right pad for the clock input to the ESS chip. As mentioned previously, I recommend using a coax cable for clock distribution (U.FL works well) and grounding the shield at both ends. Keep the clock distribution cable as short as possible (as well as the other I2S leads).
I am no expert, but this is working well for me. If more expert readers feel a need to comment, or have opinions on "better" ways to do this, please chime in!
BTW, I am curious why my B-II has no resistor at R17, is this a change that was made, or should R17 be present?
No, but you probably don't have a good reason to have it there in the first place.
Russ,
Your answer puzzle me. I have the BII with 100 MHz crystal. Is R17 there by mistake?
Please clarify,
D.
I think you mean you have the 100Mhz clock. If were a crystal it would be required (along with two caps which you cannot install while the clock is in place).
As Brian said in the first run of that revision there were R17 populated there. I had completely forgotten about that.
It won't harm anything in the least. It is just not necessary. Rest easy. Things are as they should be. Cheers!
Russ
I have no problem with other dac, connected in the same system.
Buffalo is in metal box, grounded.
Any help, please ...
Thanks,
Angel
Hi Angel, it sounds like one possibility is that you might have a damaged comparator (you would need to check the output of the comparator to know for sure).
Post to the support forum and we will see if we can help you fix it.
Cheers!
Russ
Russ/Brian, Will the Buffalo II ever be available again? I would like a dual mono setup, but the price of two buffalo IIIs is too much. Mind you, I could always sell the BII and extra output stage, but I wonder if others may be in my position too.
Is there any real benefit of BIIIs in dual mono vs BIIs?
Is there any real benefit of BIIIs in dual mono vs BIIs?
Hi, I'm finishing my buffalo II project output tube.
My source is a MacMini and connected to the DAC via TOSLINK. I've placed a module toslink and connected to D1 and GND at the entrance of buffalo.
I've plugged in because I could not lock SPDIF when playing at 96KHz. Is it right or do I have to put Single S / PDIF Level Converter Kit?
On the other hand, I want to put a USB-I2S converter based on xmos. I have then 2 signal inputs:
- SPDIF connected to D1 and GND
- I2S inputs connected to the buffalo.
What do I need to manage these two inputs?.
I am confused between sidecar, MUX, 4:1 spdif, etc.
Thanks for the help
AL
My source is a MacMini and connected to the DAC via TOSLINK. I've placed a module toslink and connected to D1 and GND at the entrance of buffalo.
I've plugged in because I could not lock SPDIF when playing at 96KHz. Is it right or do I have to put Single S / PDIF Level Converter Kit?
On the other hand, I want to put a USB-I2S converter based on xmos. I have then 2 signal inputs:
- SPDIF connected to D1 and GND
- I2S inputs connected to the buffalo.
What do I need to manage these two inputs?.
I am confused between sidecar, MUX, 4:1 spdif, etc.
Thanks for the help
AL
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.