Sound card?

:

driver.

Same PC, three sound cards. Maybe I'll bring a fourth in tomorrow, but it's an internal (PCI) card and inside another system.

:

driver.

It might be worthwhile to try the best one on another PC, preferably one of very different pedigree. Often it's the PC that causes most of the noise.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

:

driver.

I put transformers (no attempt at matching them) on the Line-in (L&R) and that fixed the 5MHz coming from the internal sound card. It didn't do anything for the SoundBlaster External though. As expected, the internal sound card isn't all that great though.

I ordered the sound card Steve suggested. We'll see how that works. The boss told me to buy a better sound card if we need it but spending more money is obviously no guarantee of better performance.

On a sunny day (Wed, 28 Jan 2009 17:12:29 -0600) it happened krw wrote in :

I had a strong about 10MHz signal here some time ago, mainly in the evening, it came from the CB antenne cable's coax shield, working as a normal antenna. In Amsterdam one time the AM broadcast stations were so strong they interfered with ripple measurements on a power supply I designed (you could see the audio AM modulation on the scope), picked up by a huge high power wire wound load with just the right number of turns to tune it I guess. So perhaps your 5MHz signal is a shortwave radio station, or some local RF source. (Tune a radio to 5MHz and locate it?) That would explain why the transformer filters it out, it does not go that high.

100pF would have done the same trick too perhaps. My 10MHz signal had fading too...

Something internal to the PC makes more sense than picking it up out of the ether. My scope's frequency counter never got a good lock on it so I don't know its exact frequency. It could even be the sound card's codec leaking. Dunno about a cap filtering it out. It was everywhere.

That's certainly a possibility, though using a crappy sound card as a crappy scope is bad enough. Bastardizing a high quality sound card seems to be a waste. ;-)

Balanced in/out would be quite useful for us though, since that's a lot of what we do. A reasonably high input impedance is needed though. I'll take a look at what's out there.

I've found a mechanical noise source at a client that had bothered them to no end, where a Stanford Research analyzer had thrown in the towel. "How on earth did you find that?" ... "With my laptop".

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

shots:

to have,

I write my own..

.

oice control.

myself.

I have a Griffin imic that I can't get to run on Suse linux. Any experience with that device?

Just an FYI for linux sound users. I run several Diamond XS71 soundcards in the linux box. The C-media chipset has good support.

For test equipment purpose, I bought a DSA. I've given up arguing with people that say soundcards are as good as test instruments. It's not worth the effort. You end up testing the soundcard/PC and not the DUT.

For really high quality playback, my Logitech Slimdevices Squeezebox Classic is great. It sounds as quiet as a CD deck. You can generally get them on line for $50 off if you can find the secret password on the net. Great software too. It does FLAC. You can store the audio in FLAC on your server, then stream in MP3 if need be The only bug I found is the switcher makes AM noise. However, it is external, so you can run the box on a linear.

Do you run them under Linux?

A caveat, with the 24$ USB box, go through all the windows sound settings menus, all 3 or 4 of them, and turn everything off so you get a straight through wave with no added "processing". That can be harder then it sounds. Also do not use "Media Player" as the source, if you can help it. Try "Goldwave" or similar. Goldwave also has a "evaluate expression" function for passing a math derived waveform out to the card.

Steve

I figured it is cheap enough just to buy one and hack with it. I need more sound on my linux box, and Goldwave is a windows program.

On a sunny day (Wed, 28 Jan 2009 17:09:56 -0800 (PST)) it happened " snipped-for-privacy@sushi.com" wrote in :

Never heard of it, so I looked it up. Cannot find the word 'linux' on their ste, seems more Apple oriented. So, the normal procedure is then: grml: ~ # cd /usr/src/linux grml: /usr/src/linux # cd drivers grml: /usr/src/linux/drivers # cd usb grml: /usr/src/linux/drivers/usb # grep -iR Griffin * input/powermate.c: * A driver for the Griffin Technology, Inc. "PowerMate" USB controller dial. input/powermate.c: * certainly faster than the human hand could turn it. Griffin say that it loses a input/powermate.c: * Griffin were very happy to provide documentation and free hardware for development. input/powermate.c:#define POWERMATE_VENDOR 0x077d /* Griffin Technology, Inc. */ input/powermate.c:#define POWERMATE_PRODUCT_NEW 0x0410 /* Griffin PowerMate */ input/powermate.c:#define POWERMATE_PRODUCT_OLD 0x04AA /* Griffin soundKnob */ input/powermate.c:static char pm_name_powermate[] = "Griffin PowerMate"; input/powermate.c:static char pm_name_soundknob[] = "Griffin SoundKnob"; input/powermate.c:MODULE_DESCRIPTION( "Griffin Technology, Inc PowerMate driver" ); input/Kconfig: tristate "Griffin PowerMate and Contour Jog support" input/Kconfig: Say Y here if you want to use Griffin PowerMate or Contour Jog devices. input/hid-core.c:#define USB_VENDOR_ID_GRIFFIN 0x077d input/hid-core.c: { USB_VENDOR_ID_GRIFFIN, USB_DEVICE_ID_POWERMATE, HID_QUIRK_IGNORE }, input/hid-core.c: { USB_VENDOR_ID_GRIFFIN, USB_DEVICE_ID_SOUNDKNOB, HID_QUIRK_IGNORE },

No luck for the 'imic', but they may help you if you want to write a driver it seems.

grml: /usr/src/linux/drivers/usb # grep -iR imic * misc/usbtest.c:/* ISO tests ... mimics common usage serial/mct_u232.c: * - Mimic Windows driver by sending 2 USB 'device request' messages serial/mct_u232.c: /* Mimic the MCT-supplied Windows driver (version

1.21P.0104), which grml: /usr/src/linux/drivers/usb #

Nope

I dunno which Cmedia chip you use, but I think I have one soundcard laying about that uses one. and that one is noisy with all the sliders at zero.

DSA? What is that?

Sure, you need real test equipment.

I get -72dB from the Ensoniq soundcard as noise floor. As the PC fan makes a lot more noise, plus all the other stuff here, that is fine with me. Audio people going for -120dB will need a special dead room. So, 24 bit audio is wasted on me, in fact I use the compander a lot. If you want to play your music in the car, try procressing it like this, this script I have in /usr/local/sbin/ucompand_m: You call it with a mp3 file as argument, and it makes a compressed wave file, with maximum amplitude at exactly 100%.

The companding in this script mimics a NE570 chip.

echo "Making wave file." mpg123 -w /tmp/q1.wav $1

echo "Getting max amplitude." sox /tmp/q1.wav -e stat 2> /tmp/stat.txt

echo "Extracting volume from stats" cat /tmp/stat.txt | awk '/Volume adjustment:/ { if (1 < 2) print $3 }' >

/tmp/vol

vol=at /tmp/vol echo "volume=$vol"

if [ "$vol" == "1.000" ] then echo "Companding wave file." sox /tmp/q1.wav -t wav $1-companded-unity.wav compand 0.003,.013.5

-80,-40,-60,-30,-40,-20,0,0 1

echo "Deleting temp file." rm /tmp/q1.wav else echo "Creating unity wave file." sox /tmp/q1.wav -v $vol q1-unity.wav

echo "Deleting temp file." rm /tmp/q1.wav

echo "Companding wave file." sox /tmp/q1-unity.wav -t wav $1-companded-unity.wav compand

0.003,.013.5 -80,-40,-60,-30,-40,-20,0,0 1

echo "Deleting temp file." rm /tmp/q1-unity.wav fi

echo "Ready"

exit

----------------------- You could re-encode it to mp3 again by adding the line: lame $1-companded-unity.wav

:

A cheaper solution is probably a USB soundcard and a ground isolating USB hub.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn\'t fit, use a bigger hammer!"
--------------------------------------------------------------

:

Well, we wanted a bit more than that. Timed recording, auto-spool onto LAN, no Windows crashes, and so on. Works like a champ now. If you need to record something outside the timer settings such as a memorial service or choir practice all you have to do now is press start and at the end press stop.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

" USB controller dial.

Griffin say that it loses a

free hardware for development.

=A0 =A0/* Griffin Technology, Inc. */

=A0 =A0 =A0/* Griffin PowerMate */

=A0 =A0 =A0/* Griffin soundKnob */

; ;

driver" );

support"

owerMate or Contour Jog devices.

=A0 =A00x077d

OWERMATE, HID_QUIRK_IGNORE },

OUNDKNOB, HID_QUIRK_IGNORE },

er it seems.

request' messages

r (version 1.21P.0104), which

g about that uses one.

I'm using the cards for digital demodulation from radios. The noise floor is not an issue as much as just getting the thing to work in Linux. It uses the CM8738 driver.

Dynamic Signal Analyzer, not a very descriptive name. HP3562

ile,

0.003,.013.5 -80,-40,-60,-30,-40,-20,0,0 1

mpand 0.003,.013.5 -80,-40,-60,-30,-40,-20,0,0 1

On a sunny day (Thu, 29 Jan 2009 12:05:26 -0800 (PST)) it happened " snipped-for-privacy@sushi.com" wrote in :

Thank you, learned an other abbreviation, I did try googling, but it found things like Digital Signature Algorithm... :-)

I do not understand how your problems are occurring, it is rather contrary to the basic design of these sound cards.

Just the same try looking in the semi-pro low-pro audio region like RME Hammerfall. The kind of equipment used to make a compromise home studio.

Interesting. Here in the states both 5.0 MHz and 10.0 MHz are reserved for certain special government broadcasts, both with distinctive modulation. No other stations would be authorized. Now if the interference were at 4.77 MHz i would expect a PC internal source.

It's the WWV broadcasts. I hope they keep them running, very handy to calibrate stuff.

That would represent the vintage Model A of PCs :-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.