Jump to content

MartinC700

  • Posts

    45
  • Joined

  • Last visited

  • Country

    country-ZZ

Retained

  • Member Title
    Freshman Member
  1. Hi Forrest, Never could find the short,it was between the clock output and the board ground. I ended up just bpassing the board by mouting the clock upside down and wiring directly to it. I did note that the board came with one of the capacitors mounted backwards so you could check all those. Not that for the Evo clock you should wire directly from the oscillator clock, don't use a resistor in series as the Evo uses a parallel input resistor with a 75 ohm load. Then use 75 Ohm wiring and connectors. Let me know how you get on. Post a picture if you still can't find the fault. Regards
  2. Thanks xiddox, I your instructions were great ! I did have one issue still, and had to make another change to the details in the post, specifically the line $ chmod u+s m2tech With help from someone familiar with unix, we changed it to $ chmod o+x m2tech and it now works. Thanks again for your help.
  3. Hi Xiddox, Sorry my mistake. I am a Mac novice, I opened the Apple script editor, pasted in the text from the Australian site then tried to save. Whatever format I tried to save in it gave the same error, something like "expected end of line" and would not allow me to save. Is there a plain text editor on the Mac and an extension I shoudl save with instead of using script editor...again, sorry if these are dumb questions, not a Mac used except for using the Macbook as a music source!
  4. Did you try the script? I am a novice on the Mac, but tried to compile the script and it just threw up and error.
  5. A point that many fail to comprehend is that looking at component cost bears little relationship to the cost of a DAC to manufacture. An example I would make is to look at Twisted Pear Audio DIY Dac's. Russ and Brian probably just break even on their products which are designed by them and I believe sold out of their homes. A good Sabre based DAC with PSU's, analog stage, some digital input switching etc without a case will cost you around $600. Add your assembly labor and a case and you are looking at around $1k. Now factor in a companies normal overhead rates and profit and you would be looking at say $3k, which is right where a comparable DAC from say Weiss would be. Now consider offshore manufacture, simpler design, cheaper parts 1000x the quantities (still very small compared to other consumer markets) and you are still at the $300 level. So, boutique DAC's are expensive, it's a function of quantity mainly!
  6. Works with the EVO too! Edit: Well, it worked fine until I rebooted the Mac, now the Young driver doesn't show up. Oh well, back to the Hiface driver. Anyone have the same problem after rebooting when using the Young driver with the HiFace?
  7. Problem solved - looks like much of the noise was due to the fact I had the clock board floating with respect to ground. Grounding the clock board and then checking the external clock as it enters the EVO showed a much better trace. There is some undershoot/overshoot but reading up on thr issue, that is normal for a clock line that is parallel terminated as the EVO appears to be. Parallel termination has more ringing than series termination, but retains the signal edge better, so my guess is that thats what/why M2Tech use it. Listening now there is a definite step up from the internal clock. It was a tough learning experience but worth it!
  8. Hi Guys, The supply to the clock is very clean, it takes the 5v output from the Valab clock board (with the clock removed), then uses a TPA Trident shunt to reduce to 3.3v. I see no noise putting the scope on the power supply at all. I also have a 0.1uF film cap right at the + supply for the clock. I will try looking at the clock in the EVO also and see how that looks and report back, that shoudl tell me if it's real or a measurement issue. Interestingly the oscillator on the VALabs board originally put out a sine wave - probably not the best for a timing signal.
  9. Quick update on the external clock. Firstly, looking at the output of the clock on my scope (LeCroy 100MHz), the clock trace looks terrible, with what appears to be much overshoot and ringing. I initally though that it was merely an artefact of impedence mismatch of the probe, however, it appears inherent to the Crystek clock as it is independent of whether the oscillator is driving a 75ohm load (via 75 ohm connectors and cable) or unconnected. That said, the EVO locks on to the external clock and it's external clock LED lights. Onto the sonics. The EVO is feeding a Twisted Pear Buffalo II DAC via SPDIF. I can switch the external clock on and off, so its pretty easy to compare the EVO running with internal and external clocks. Well, to be honest I really don't hear any significant difference. Perhaps the external clock presentation is slightly different, however in a blind test I am sure I could not differentiate the two. Since the clock signal appears pretty bad on the scope, I am rather surprised that the extenal clock did not sound worse, so perhaps with some more tweaking, the external clock could be advantageous. The other possibility is that the Buffalo DAC is insensitive to the small difference that the clocks introduce, after all the EVO (supplied by an external shunt regulator) is pretty darned good already. My plan is to mess around with the clock a little more to see if I can clean up the signal, trouble is my knowledge of digital signals is very limited. Regards
  10. Hi Forrest, Many thanks for the offer, I probably am OK as I now have a replavement XO and will likely use a small slave board with the new XO on. The XO to ground connection is a bit of a mystery, I cut the output trace in a couple of spots and it still shorts to ground (even with the XO removed). All you need to do is put an ohmeter between the XO out terminal and the board ground plane to check. It is interesting that in connecting up the orignally supplied VA labs clock to a 5v supply, the output is a pure sine wave, where as the Crystek and other precision clock oscillators normally output a square wave, presumably because the fast rise time produces a more accurate timing trigger. Should have the new clock up and running this weekend, will post when it's done. Regards Martin
  11. My schedule is a bit delayed as is seems that the board had a couple of faults. The first was easy to fix, one of the electrolytic caps was reversed. Unfortunately the second was more of an issue, the clock output was shorted to ground, hence it fried my 957 clock. A clock one is on the way and should be running next week. Regards
  12. Valab board arrived, looks pretty well thought out, from what I can see the TL431 used outputs to a Darlington pass transistor set up. The supply is then split and RF filtered, one line to the clock and one to the counter. My board has a component error, one of the power supply output filter caps is the wrong polarity, not very good QA from TerraDac/Valab. Should be easy to mod though. Plan is to just remove power from the CMOS counter/divider, and like your setup, add the Trident to the existing 5v supply to provide 3.3v to the replacement Crystek crystal. Should have it up and running this weekend. Regards
  13. The acoustica article was where I though the information was but looks like it was in the manual for the PFM Flea clock. This manual doesn't appear on the index to the Acoustica site, but if you Google 'PFM flea manual' it will show up. The clock board I ordered is the same as yours, and my plan was merely to swap out the supplied XO with the new Crystek when it comes available, either using like you a Trident to supply the 3.3v to the clock or alternatively a PFM Flea based regulator. Looks like the Crystek clock maxes out at 25mA current draw and the Trident shunt could be set to draw a total of say 50mA. Not sure what the overall current draw of the Trident would be but my guess is that is close to the max rating for the onboard regs on the clock board so the input resistor on the circuit may have to be altered to match. When the XO board arrives I will do a bit of reverse engineering to see how they implement their TL431 'double' regulation scheme. Regards Martin
  14. Hi Forrest, I also have the 957 on order, so was planning on not using the counter. As far a the effect of the counter on jitter, my information was not first hand but was from the Acoustica site, which is usually pretty trustworthy. I plan to try it direct then via the 74HC163 and see if I can tell a difference. I believe that I may have ordered the same basic board as you from VALAB. Do you know if the board has resistors on the output of the XO or after the CMOS gate? Regards Martin
  15. Just ordered the parts to try a similar experiment, thanks for your report! I notice that you appear to have retained the 74HC163 CMOS counter in your circuit. From what I understand, unless you need to frequency divide, you could just take the XO output to a 75 ohm resistor then to the EVO, there by reducing noise / jitter further. Regards
×
×
  • Create New...