You know, I just did not think ever to measure the THD on my card. It is working so fine and it give me a such hi-fi sound that I never thought to do it.
But It could be interesting to know the THD level on my board (with Crystek oscillator). To be an fair/right comparison we may use the same setup to measure this. If you can give your setup or describe your method, and the name of the software application you used to measure this, then I will try to reproduce it on my board, and we can see the results afterwards.
The measurements were made using RMAA. I used the headphone output. I improved THD a lot by using LM4562 and improving Xonar -12V line. Each capacitor model gave different filtering resulting in different THD. The current setup has much clearer and much more fresh mids. The 3kHz is very important when producing music as I do this occasionally. It's the third formant of professional singer.
Do you mean that for the measurements -- jitter, whatever -- you've been posting, you are indeed using the SAME computer (as in the RMAA procedures I noted above)?
If you're concerned about using the SAME computer you could, e.g., use another computer (e.g. friend's laptop) for running ARTA or RMAA.
Using another computer would introduce large jitter error from the recording card. The Xonar ST uses CS2000 to multiple the CMI8788 clock twice and feeds it to the ADC. It's better to use card with very low jitter and then compare two playing cards. If the difference can be seen, then the recording card is good enough.
I can't go belows some Asus measurements when using best PSU. However I can gain a lot when dealing with the oscillator PSU. No change in measurements like noise or THD, but the sound changes a lot. Now I find the 7805 with output bypassed with tantalum 27uF and 100nF ceramic to be pretty good. The ceramic value is critical for 24MHz range: external tests show more than 30dB noise reduction. I may find the CLC better: the 50dB noise reduction for 24MHz is possible when using ferrite bead.
You could try the Flea reg., a tuned and tweaked ckt for canned osc's. To work with 12v Molex power, skip the 7812 and all before it.
I use the whole thing in a few other (older) projects -- with good results.
consternation...or is that constipation?
This remark can be a source of consternation...or is that constipation? Uh ... er ... anyway ...the burden of proof on on thee.
Confused! or Corvused
IAC ... if you want to tap a 12v source, build a DEDICATED/regulated pwr source for the TCXO. If you don't want to try the 5v Flea reg. I noted earlier, use Tent's simple ckt here:
Also, as I noted prev., go back to the 12v Molex input terminal BEFORE any Xonar regulator. Use ferrite beads liberally -- they are cheap.
Confused...er... Corvused ...Are you using 12v for the TXCO (which is designed for 5v)? (Remember: you STILL have to reg. it down to 5v)
IAC ... if you want to tap a 12v source, build a DEDICATED/regulated pwr source for the TCXO. If you don't want to try the 5v Flea reg. I noted earlier, use Tent's simple ckt here:
Also, as I noted prev., go back to the 12v Molex input terminal BEFORE any Xonar regulator. Use ferrite beads liberally -- they are cheap.
Last edited:
Hollowman, I used the 7805 for the TCXO. The input voltage for this regulator comes from the +12V/GND line on the board, just from the 7812 output. I do not make use of ferrite beads. It's hard to tell the difference and the board tuning is harder. I keep the required elements to the minimum.
I started to make intense use of tantalum caps on the xonar card. Ordered much more smd tantalums and some more smd ceramics. Meantime I improved the connection to the headphones. This card can give the sound I've never heard on headphones.
I started to make intense use of tantalum caps on the xonar card. Ordered much more smd tantalums and some more smd ceramics. Meantime I improved the connection to the headphones. This card can give the sound I've never heard on headphones.
The RFS may need some break in, it wasn't used for over a week. The FC is ok. The same voltage, capacitance is different. The FC has 80mohm ESR at 100kHz, the RFS is uknown but quite comparable with DF. Such a small difference but the clearer impulse is much better sounds more airy, more clean, has better resolution and much more depth.
RFS?
What's 'RFS'? Cap brand, I assume. Panasonic FM are known for the break-in period. But when they do, sonics really snap in.
Coris, please set up the recording source using xonar panel. Sometimes the change is required for proper switch to work. Looks like the measurement taken with the wave out mixer. Also the peak is important. I my case I can't reach more than -8dB due to card.
Hollowman, the RFS is the Elna Silmic II. The card measurements at this level is hard. After a few minutes somethings gets little noisy and the measurements are worse.
Hollowman, the RFS is the Elna Silmic II. The card measurements at this level is hard. After a few minutes somethings gets little noisy and the measurements are worse.
I do not have the Xonar panel. I don't want it and I don't need it. I have installed only the driver. I have used Line IN and the L/R channels output connected in loop through an potentiometer (1/10). Actually I do use the STX only to play files. The IN stage in my card is normally unused.
Do you have some suggestions about the recording level for those measurements?
After I will finish the last mod (battery powered oscillator) I will remake the measurements.
Do you have some suggestions about the recording level for those measurements?
After I will finish the last mod (battery powered oscillator) I will remake the measurements.
AFAIK you don't have to install the xonar panel. You may run it and it will only give a smal latency during the Windows session. Restart and it's clear. The measurements you've taken are from the "wave" input, not the real. I get the silimar when I choose the "wave" input. Seems your driver doesn't want to switch to line input. Weird.
OK. I have disabled "Wave" and enabled/default "Line in" before measurements. Loop in between output - Line in. I could see that Line in were active (level meter worked). I will verify again.
I have finished now the battery powered oscillator. A big increase in quality. The fidelity is just exceptional, but the most impressive now is the sound stage. It gets more depth and both the instruments and voices are very well separated in space. The speakers are just transparent. The sound is in the room and fill it. Very impressive!
I have finished now the battery powered oscillator. A big increase in quality. The fidelity is just exceptional, but the most impressive now is the sound stage. It gets more depth and both the instruments and voices are very well separated in space. The speakers are just transparent. The sound is in the room and fill it. Very impressive!
Last edited: