Test & Measurement interface for Soundcard

An other interface is born. A few simple mods/changes were done: DC input choke to remove DC/DC converter RF, light bulb protection at the input instead of the 2W resistors (see this thread for details) and added 1/8 stereo jacks in parallel of the 1/4 jack, for better compatibility with my own sound card. Thanks again Mr Millet for this great design!
 

Attachments

  • SB_F-B.jpg
    SB_F-B.jpg
    97.4 KB · Views: 1,040
  • SB_IN.jpg
    SB_IN.jpg
    151.7 KB · Views: 1,021
I suggest that instead of removing D10/12, D11 / D12 be replaced by a string of 1N4148 diodes the number of is suitably selected to put the protection limit closer to the full scale voltage of the soundcard YOU are using.

Note: I got better THD (flatter increase) vs frequency by using THAT parts for the line recievers/drivers) (see posts a few pages back)

Hi all - I'm purchasing parts to start my build..including parts for a few tweaks, and I can't wait! :)

My sound card is an M-Audio M-Track Eight, which is an external USB model.
M-Audio

The balanced line inputs are rated at a maximum of +16dBu.

Am I correct in understanding that you are suggesting changing D10/D11, D12/D13 to be series string like the following;

+16dBU = 13.82v(p-p) = +- 6.91v
Assuming 1N4148 vF = 0.62v = 10x in series

With this series string of 10, would I replace D10 or D11 with 9?

Or would the limits imposed by the input diodes and gain of proceeding stages make this number in series unnecessary?

Many thanks!
 
I think it is wise to retain some form of voltage limiter.

The voltage limiter is a shunt regulator.
It needs a current limiter before it. That is usually just a resistor. It could be a PTC resistor to help avoid some added noise.

The combination of resistor + shunt should be designed to survive your worst case accidental input.
How big a voltage might you accidentally apply to your sound card?
 
Last edited:
Hi Andrew, many thanks.

I'm unsure to be honest (I'm still learning!). I initially assumed the prior input stages clamp at ~1.2v(p-p) hence the maximum output to the stage we are discussing would be low, however on reflection I *think* it is clamping relative to the rails. Is this correct?
 
Last edited:
Good question. @pmillet: What is the worst case / failure-mode voltage that could make its way to the sound card if D10 and D12 are removed?

Certainly no more than the power supply, which is normally +/-5V. In reality probably no more than +/-4V, since the output buffers usually won't swing all the way to the rails.

I would think that most professional sounds cards would be fine with that...

The important clipper is D5/6/7/8...

Pete
 
Last edited:
Certainly no more than the power supply, which is normally +/-5V. In reality probably no more than +/-4V, since the output buffers usually won't swing all the way to the rails.

I would think that most professional sounds cards would be fine with that...

The important clipper is D5/6/7/8...

Pete

Thanks! Can R28 (200R series) be shorted then, or just leave it? Definitely get better performance out of the circuit this way and I agree, it should be a problem with at least my setup.
 
Another interface lives. The project was very well documented, esy to build and very professional.

Be warned I am a complete newbie at this. I'm looking for feedback on my first couple of measurements... Do the results look valid?

The first is a Petes interface hooked to a Focusrite scarlet 2i2. The 1k tone is generated by the computer using audio tester. Voltage out was 4.4vac

The second is a 1k tone generated by my phone playing through an old hafler amp. Same test set up same voltage.

I hope to get more comfortable with using the computer to do measurements so I can properly set up my newest 4D32 tube amps.
Thanks,
Evan
 

Attachments

  • test1jpg.jpg
    test1jpg.jpg
    266.5 KB · Views: 460
  • test dh200 4.4v 1k jpg.jpg
    test dh200 4.4v 1k jpg.jpg
    229.1 KB · Views: 453
Trobbins, I wish I could answer intelligently. I really don’t know. I’m having a hard time getting my head around this whole thing. It sems simple on the surface ... inject a signal ... look at the output... but as I actually try to get meaningful information out of the computer I am running into a lot of things I don’t understand.

Not related to your question. When I use the gen mon feature on Pete’s interface I can output plenty of voltage from the computer. This is how I generated the first plot. But when I try to use the Soundcard to drive the amp I’m testing I can only get about 300mv

I need to ponder and play with this quite a bit more
 
Last edited:
Member
Joined 2016
Paid Member
Hi all,

I'm in the market for an audio analyser that can do more than my poor FFT from my Siglent scope. Not really skilled yet in the area, but I need the usual things like FFT, THD measurements, etc.

I have gone through the whole thread with interest and intend to build it from Pete's PCB and panels. However, I'm working with a Linux laptop (Ubuntu) so instead of building a desktop PC an cluttering my lab with it, I would rather use an external USB sound card and ideally some Linux software. I don't mind paying for the software, but Virtualbox emulation is a pain and if I can go native I'll be happier.

Anyone has experienced USB sound cards and linux software ?

Regards,
Charles
 
Member
Joined 2007
Paid Member
We recently finished the Interface for Soundcard...I say WE because my nine year old son populated at least 75% of the PCB (with me hovering over his shoulder of course). It's hard to be patient and not grab the iron out of his hands. Anyway, the device works as advertised, was pretty simple to build, and all parts were quite easy to acquire (mostly copy and paste from the BOM to DigiKey and Mouser. I had a hard time debugging it at first as one of the pins on an IC was out of it's socket. This a great tool and I appreciate the hard work Peter put into it...Thank You. I could attach pics (to prove it really happened), but I think you guys already know what it looks like. Thanks Again!!!
 
Soundcard interface setup with Juli@ soundcard and ARTA

Hi,

I recently completed my Soundcard Interface. It went together perfectly and it was a pleasure assembling it! I earthed the selector switch shaft and lined the box with mu-metal both as suggested earlier in this thread.

It appears to work fine testing with a signal generator and DMM. I don't have an oscilloscope at the moment so I can't be absolutely sure.

I'm trying to use the Soundcard Interface with a Juli@ sound card and evaluating ARTA for simple amplifier frequency response and THD/spectra measurements.

Not understanding how to use ARTA could well be my problem here.

I've attached the loopback spectrum.
loopback spectrum.png

Looks good and can be improved significantly by reducing the level.

If I connect-up the soundcard with the right channel as reference and select "Gen Mon" to link the sound card output to the soundcard input I should be able to measure the contribution of the soundcard interface as many others have done on this thread.

I had to reduce the input level to -12dB avoid clipping in the Time Record window. I've attached the result and as you can see it's relatively poor with THD of 0.1% and THD+N of 0.11%. At least the background noise is down to 140dB and there's no peak at mains frequency.
-12dB spectrum.png

Any suggestions as to what I'm doing wrong here?

Thanks.
 
Looks good and can be improved significantly by reducing the level.

Looks very typical for Juli https://www.diyaudio.com/forums/equ...ensation-measurement-setup-4.html#post5584954


If I connect-up the soundcard with the right channel as reference and select "Gen Mon" to link the sound card output to the soundcard input I should be able to measure the contribution of the soundcard interface as many others have done on this thread.

Look at the card diagram. The monitoring feature feeds incoming data from inputs to output mixer, not the other way round.

scheme-big.png
 
Success!

Now that I better understand how to use ARTA I'm getting similar results to others.

I just need to see if I can reduce the 15dB mains bump. Interface is currently powered from a computer USB port. I thought earthing the switch shaft and mu metal shielding top and bottom would fix that.

GenMon 239mV 200mV scale.png