1) the added mass to the board and potential wiggly mechanical connection / vibration possibility.
2) The distance between the low z power feeding and the consumer.
3) OK, I wrote "for most" as I think opening up this level of interface creates options and makes it harder in general to create a good predictable solution if one would evaluate the statistical outcome of a lot of implementations. But hey, Ian is probably doing this for himself aiming at the ultimate solution and not a mass-market product
Would a "sense" pin on the board connected to the point of consumption be usable??
All this is of course nit picking on the 8:th decimal :-D
/
2) The distance between the low z power feeding and the consumer.
3) OK, I wrote "for most" as I think opening up this level of interface creates options and makes it harder in general to create a good predictable solution if one would evaluate the statistical outcome of a lot of implementations. But hey, Ian is probably doing this for himself aiming at the ultimate solution and not a mass-market product
Would a "sense" pin on the board connected to the point of consumption be usable??
All this is of course nit picking on the 8:th decimal :-D
/
Last edited:
1. yes, because everyone knows sound waves travel from speakers directly through enclosures, always striking the vertical surface first....
Ian already recommended soldering it if permanent installation is desired.
2. its the decoupling caps supplying the transients... low noise is most important
3. if it were a mass market product he would have stopped long ago and charged a lot more.
Ian has already demonstrated an ability to actually measure as well as model the minutiae many can only ponder or guess at the existence of. he knows there are better solutions for low noise than an LDO, but also knows everyone at this level or at least most of us, have already formed an opinion of what that is, which may or may not coincide with his own preference.
he would never please everyone, most will without certainty plug in a power source that is per-concieved to be 'the best' and clocks in general are an area people are pretty fanatical about, best leave the crazy people to their own devices
Ian already recommended soldering it if permanent installation is desired.
2. its the decoupling caps supplying the transients... low noise is most important
3. if it were a mass market product he would have stopped long ago and charged a lot more.
Ian has already demonstrated an ability to actually measure as well as model the minutiae many can only ponder or guess at the existence of. he knows there are better solutions for low noise than an LDO, but also knows everyone at this level or at least most of us, have already formed an opinion of what that is, which may or may not coincide with his own preference.
he would never please everyone, most will without certainty plug in a power source that is per-concieved to be 'the best' and clocks in general are an area people are pretty fanatical about, best leave the crazy people to their own devices
Last edited:
Ian,
One question:
I got the Dual-Clock-Board from you with the 22./24 clocks on it and TP3/TP4 etched as text on the PCB for this speed. I exchanged the clocks to crystek 957 22/24. Plug and play.
If I now want to get to 44/49: What to do ?
I have read your guide on double speed, but it says: TP3/TP4 again. for the 44 and TP4/TP5 for 22.
SOOOOOOOOOOO.....has there been a different version of the board or a firmware-update ? Or what was the "MASTER-Switch" which changes to right setting for 22.x from TP3/Tp4 (as etched on the PCB and written in your V1 guide) to TP4/TP5 (as written in the double XO-Speed guide) ? I am confused.
One question:
I got the Dual-Clock-Board from you with the 22./24 clocks on it and TP3/TP4 etched as text on the PCB for this speed. I exchanged the clocks to crystek 957 22/24. Plug and play.
If I now want to get to 44/49: What to do ?
I have read your guide on double speed, but it says: TP3/TP4 again. for the 44 and TP4/TP5 for 22.
SOOOOOOOOOOO.....has there been a different version of the board or a firmware-update ? Or what was the "MASTER-Switch" which changes to right setting for 22.x from TP3/Tp4 (as etched on the PCB and written in your V1 guide) to TP4/TP5 (as written in the double XO-Speed guide) ? I am confused.
Last edited:
no....what would happen if I pull the 22 clock now with tp3/4 and plug the 44 in ? any damage possible or should i simply try ?
Blitz,
They gave correct answers.
You just need to make sure the dual xo firmware is the most updated version which is V3.3 by now. Software will take care for anything else.
Nice weekend.
Ian
Found the sticker, it was under the board...and as I have connected the board with everything it was not easy to find...3.3, yes.
So it was plug and play without changing anything...easy.
Only question left: I put the jumper between TP3 and TP4...as well with the 22.x clock. Made nice music...but according to your guide this was wrong. TP4/TP5 would have been right, correct ? So, I should change the TP3/4 to TP4/5 with 22.x and see how this sounds ?
By the way: The Crystek 22.x with TP3/4 sounds completely different than the Tentlabs 45.x. The crystek much like a Direct heated Triode...very rich tonality, very fluent and liquid sound. The Tentlabs sounds like a heavy-feedback transistor amp based on bi-polars in comparison. Clean, controlled, but technical sounding. Unbelievable how much a clock can do.
So it was plug and play without changing anything...easy.
Only question left: I put the jumper between TP3 and TP4...as well with the 22.x clock. Made nice music...but according to your guide this was wrong. TP4/TP5 would have been right, correct ? So, I should change the TP3/4 to TP4/5 with 22.x and see how this sounds ?
By the way: The Crystek 22.x with TP3/4 sounds completely different than the Tentlabs 45.x. The crystek much like a Direct heated Triode...very rich tonality, very fluent and liquid sound. The Tentlabs sounds like a heavy-feedback transistor amp based on bi-polars in comparison. Clean, controlled, but technical sounding. Unbelievable how much a clock can do.
Firmware upgraded after hardware, so please don't mind about the marks on the PCB. Please follow the document with 100% to make sure every setting is correct.
Jumpers are just telling the processor how to swich Fs, without changing anything to clocks, so will not affect sound.
Ian
Last edited:
I changed the jumper to what the manual says...instead of tp3/4 Tp4/5...not sure if this changed the sound, but for sure it changed the frequency the board shows. Before it was correct, now the board shows half of the frequency, so a 96khz recording now is 48khz in the leds of fifo board...so, is this intended or back to tp3/4 ?
For double speed mode, Yes.
Real frequency decided by the XO your are using, not the jumper.
LEDs show different for normal and double frequency.Please read the pdf file for more detials
.Ian
Thanks for your patience...reading helps...
so, if I get it right....for the 22.x both Settings work. Difference is
TP3/4: The Dual Clock module works in normal mode, therefore the speed LeDs are showing the right speed as well.
TP4/5: The module works in Dual speed mode. Not needed for 22.x clocks, but for 45.x, but neverthelss possible.
Right ?
so, if I get it right....for the 22.x both Settings work. Difference is
TP3/4: The Dual Clock module works in normal mode, therefore the speed LeDs are showing the right speed as well.
TP4/5: The module works in Dual speed mode. Not needed for 22.x clocks, but for 45.x, but neverthelss possible.
Right ?
Very close to my subjective impressions. What i find the only subjective downside is the reduced back to front perception, everything else seems just fine.
Did you try at all the kit clocks?
It seems with reasonably high quality I2S input, improvements cannot be taken for granted even if the underlying principles of operation are so compelling. Cabling, ps noise, etc probably become the real bottleneck.
Have you experimented at all with the USB cable since using Ian's kit? One would expect this should now become inaudible, right? And also fine tuning OS and software for minimum latency. What about your Sotm card, have you tried using a standard USB port?
I have currently not yet started with FIFO in the chain to swap stuff and learn. It will be interesting as both can happen: The chain is less sensitive now or more sensitive now...question is how much FIFO "corrects" or if it is more like a magnifier which shows now with a higher level of clarity what is going on.
BEFORE FIFO:
The Sotm USB was very audible. Much more liquid and tonality rich than the normal USB-Port from any motherboard I tried (and I tried a lot). It is worth the money. With normal Motherboards in any case, but even with an Alix-Board.
On USB-Cable, I used normal Kimber-Cable CU. Nevertheless I plan to build my own cable, shorten it to 20cm and make it Ground and Data only and Silver-Cable. You can buy these cable for 500 Euro or build them yourself for 20 Euro.
I will for sure try as well silver-cable for connecting the FIFO with WAVEIO and BIII.
Latency changes where very audible already before, I played around with the CPU and Ram-Latencies and change the switch to the Netgear GS108E which is a big improvement, especially in terms of reliable delivery of data stream for the HD stuff. NO DROP OUTs anymore since this one takes care of the distribution !!! As well a nice sonic improvement for free.
MPDPUP is as well mandatory, much more musical than any Windows stuff, and I tried a lot. (CMP, Jrover, Foobar, Jplay etcetc; Hardcore tweaking of Windows XP, Vista,7). As well the Linux-Tweaks suggested in the MPDPUP-Threat are audible.
Since the FIFO I have now a slight, small hum in my left speaker and I need to fix this first, we need a real quiet system to do these kind of evals.
Outlook:
- I will get from Ed the 44/49 clocks and will compare.
- Battery supply is in Prep as well.
- If I have this together, maybe Ian's S570 is already in the house and we will see.
- Ackodac is ordered to do a comparison to the BIII as well and have the U-Fl cable game to try.
- Moving from Async to Sync with Ackodac and S571
- AckoDac will be powered by Paul Hynes-Regs or
- Salas Shunt-Regs are in the house since this week and
- the D1, the Sen and the CCDA based on 12B4 are in process to be build.
Those three outputstages will require rebuilding the power-amp to fully balance, no phase-splitter anymore.