Test & Measurement interface for Soundcard - Page 82

dch53 · 13th January 2019, 02:24 PM

It must have been picking up the mains from the environment.

I have thin sheets of mu metal stuck to the inside of the top, bottom and sides of the case. I linked the side pieces to the top and bottom sheets with blobs of solder and earthed them which I hadn't done before.

The mains hump has disappeared altogether.

This is without the higher performance components. I included these when I ordered all the parts so I might as well fit them.
Loopback spectrum -30dB 230mV earthed mu metal shield.png

phofman · 13th January 2019, 02:54 PM

Well done. I did not know your enclosure was not fully shielded.

For shielding smaller non-conducting enclosures or compartments a solderable copper tape has been of great help to me 50mm x 3m EMI Copper Foil Shielding Tape Conductive Self Adhesive Barrier Guitar | eBay

MagicBus · 14th January 2019, 07:52 PM

Here is how I built mine. I used a generic cheap metal box modified to the rear to fit the pcb. I built it with the THAT chips and the light bulbs but sans the LCD. I put connectors for my DMM for this job. I attach loop measurements with a behringer UCA 202 usb soundcard. The performance is primarily compromised by the soundcard. I'm making some thoughts to transplant it into the interface box, bypass its analog in/output and connect the chip directly to to the interface making it a stand alone device.

dch53 · 15th January 2019, 01:36 AM

Thanks Magic. I included the THAT parts and lamps in my component buy so I can go down that route.

I was confused by your first trace with the fundamental at -1.5dBFS. I realised there must be something wrong with my calibration because I can never get anything like that. Even -3dB produces serious clipping according to Time Record.

Paragraph 1.5.2 Calibration of Soundcard Input Channels has never made sense to me. Step 3 says "Press the button 'Generate sine (400Hz)' and monitor the input level at bottom peak-meters..." When I do that, the peak-metres don't register anything.

I've just left input calibration unchanged from the value resulting from the output calibration. If I set the Generator to -30dB (240mV) and use the 200mV scale on the soundcard interface I get -28.3dBFS.

I'll have to learn how to use STEPS too.

MagicBus · 15th January 2019, 07:11 AM

I started from the same level as you about a year ago and still learning so, I can only describe what I've found hoping it will help. It seems you have digital clipping, you will know how ugly that is when you feed real music to your speakers under this condition. Anyway, you can go to your soundcard control panel and see if you can adjust line in levels - you should. Set out level to max and adjust input until you get the best results for calibration and measurements. To give you an example, in my soundcard, the input level slider has a range from 0 to 100 and I had to take it down to 3. Once you get the calibration right, STEPS is easy to use.

jan.didden · 15th January 2019, 07:46 AM

Quote:

Originally Posted by phofman

Interesting that the 50Hz peak was not eliminated by the balanced input. Any reason for that?

Not to beat my own drum, but a SilentSwitcher from the diyaudio store, fed by a Powerbank, will completely cure that.

Jan

phofman · 15th January 2019, 08:53 AM

Jan: Does that mean the peak is caused by 50Hz noise in power supply? Or 50Hz noise in the ground loop caused by the way the device is powered?

Or perhaps induced into the tool from the outside?

IMO the easiest way would be trying to power the tool from the soundcard itself. Power lines are easily available in Juli (pinout e.g. hifi.slovanet.sk :: Zobrazi? t?mu - Anal?gov? ?as? ESI Juli@ + I2S )

MagicBus · 9th February 2019, 12:06 PM

Quote:

Originally Posted by MagicBus

Here is how I built mine. I used a generic cheap metal box modified to the rear to fit the pcb. I built it with the THAT chips and the light bulbs but sans the LCD. I put connectors for my DMM for this job. I attach loop measurements with a behringer UCA 202 usb soundcard. The performance is primarily compromised by the soundcard. I'm making some thoughts to transplant it into the interface box, bypass its analog in/output and connect the chip directly to to the interface making it a stand alone device.

In case anyone would be interested, I finally did that. Performance is slightly improved and it's now very handy to use. One usb cable plug n' play, no drivers required.

dch53 · 10th February 2019, 02:06 AM

Incorporating the soundcard into the box is a great idea. What voltage levels are you putting into the interface and what setting are you using? I suspect you're using higher levels than I am.

Attached are my best results with the standard ICs and shielded box. Best results are on the 2V scale.

You'll noticed that the distortion results stop at 75.4Hz. This is a problem I'm having with STEPS: the sweep terminates after a random number of steps. I have to click "Stop" to regain control of the app. I posted a question on the ARTA thread but I still have the problem.

MagicBus · 10th February 2019, 10:13 AM

As already said I'm not a specialist myself but willing to help this is what I've done: PC interface set to 2V range, soundcard output level to max (100/100), input level quite low (2/100). But this setup may not apply to other soundcards. I found that working during the calibration procedure, where I suspect you may have done something wrong. Better try to calibrate without the PC interface in the test rig and follow the instructions to the letter. Remember to take AC measurements with your DMM and of course you will need to do calibration separately for STEPS. Also at steps check single channel level at measurement setup.

13th January 2019, 02:24 PM	#811
dch53 diyAudio Member Join Date: Jun 2015 Location: Perth, Western Australia	Better shielding did the trick! It must have been picking up the mains from the environment. I have thin sheets of mu metal stuck to the inside of the top, bottom and sides of the case. I linked the side pieces to the top and bottom sheets with blobs of solder and earthed them which I hadn't done before. The mains hump has disappeared altogether. This is without the higher performance components. I included these when I ordered all the parts so I might as well fit them. Loopback spectrum -30dB 230mV earthed mu metal shield.png

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Recommendations on pc audio interface for loudspeaker measurement	diypole	Multi-Way	9	15th July 2009 01:45 PM
Request for advice / high performance soundcard for measurement	Gordy	Solid State	8	14th February 2008 05:14 PM
MCM Speaker Measurement Interface / USB Woofer Tester	sdclc126	Multi-Way	0	17th January 2007 02:00 AM
Tenma Speaker Measurement Interface	johngalt47	Multi-Way	1	6th January 2007 03:17 AM
Soundcard interface.	Mr Evil	Everything Else	5	13th July 2005 12:38 AM

New To Site?	Need Help?
Register to Participate Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

13th January 2019, 02:54 PM	#812
phofman diyAudio Member Join Date: Apr 2005 Location: Pilsen	Well done. I did not know your enclosure was not fully shielded. For shielding smaller non-conducting enclosures or compartments a solderable copper tape has been of great help to me 50mm x 3m EMI Copper Foil Shielding Tape Conductive Self Adhesive Barrier Guitar \| eBay

15th January 2019, 01:36 AM	#814
dch53 diyAudio Member Join Date: Jun 2015 Location: Perth, Western Australia	Thanks Magic. I included the THAT parts and lamps in my component buy so I can go down that route. I was confused by your first trace with the fundamental at -1.5dBFS. I realised there must be something wrong with my calibration because I can never get anything like that. Even -3dB produces serious clipping according to Time Record. Paragraph 1.5.2 Calibration of Soundcard Input Channels has never made sense to me. Step 3 says "Press the button 'Generate sine (400Hz)' and monitor the input level at bottom peak-meters..." When I do that, the peak-metres don't register anything. I've just left input calibration unchanged from the value resulting from the output calibration. If I set the Generator to -30dB (240mV) and use the 200mV scale on the soundcard interface I get -28.3dBFS. I'll have to learn how to use STEPS too.

15th January 2019, 07:11 AM	#815
MagicBus diyAudio Member Join Date: Aug 2016 Location: Athens, Greece	I started from the same level as you about a year ago and still learning so, I can only describe what I've found hoping it will help. It seems you have digital clipping, you will know how ugly that is when you feed real music to your speakers under this condition. Anyway, you can go to your soundcard control panel and see if you can adjust line in levels - you should. Set out level to max and adjust input until you get the best results for calibration and measurements. To give you an example, in my soundcard, the input level slider has a range from 0 to 100 and I had to take it down to 3. Once you get the calibration right, STEPS is easy to use.

15th January 2019, 08:53 AM	#817
phofman diyAudio Member Join Date: Apr 2005 Location: Pilsen	Jan: Does that mean the peak is caused by 50Hz noise in power supply? Or 50Hz noise in the ground loop caused by the way the device is powered? Or perhaps induced into the tool from the outside? IMO the easiest way would be trying to power the tool from the soundcard itself. Power lines are easily available in Juli (pinout e.g. hifi.slovanet.sk :: Zobrazi? t?mu - Anal?gov? ?as? ESI Juli@ + I2S )

10th February 2019, 10:13 AM	#820
MagicBus diyAudio Member Join Date: Aug 2016 Location: Athens, Greece	As already said I'm not a specialist myself but willing to help this is what I've done: PC interface set to 2V range, soundcard output level to max (100/100), input level quite low (2/100). But this setup may not apply to other soundcards. I found that working during the calibration procedure, where I suspect you may have done something wrong. Better try to calibrate without the PC interface in the test rig and follow the instructions to the letter. Remember to take AC measurements with your DMM and of course you will need to do calibration separately for STEPS. Also at steps check single channel level at measurement setup.