QA400 test interface

I have just finished building a simple passive test interface for use with my Quantasylum QA400 spectrum analyser. Schematic is here:

https://drive.google.com/file/d/0B_n67A1hN3qtd2U0Yk9XdGNPLTg/view?usp=sharing

It has an XLR input from the generator which is isolated by a Carnhill VTB2281 600:600 transformer. There are two reasons for this. The first is that the QA400 has an unbalanced output so the transformer converts this into a balanced source. The other reason is is protects the generator from phantom power which might destroy electronically balanced sources. The output of the transformer is fed to the output XLR via a switch. This switch selects either the transformer or a short circuit or a 150 ohm resistor which are used for noise tests.

On the return side from the equipment being tested is an input XLR across which is a 600 ohm load consisting of a pair of 270 ohm resistor in series with 33 and 27 ohm resistors. The total is exactly 600 ohms. A switch selects  the meter output to either a direct connection to the input or across the 33+27 ohm resistors which makes a 20dB attenuator for measuring large signals. There is no isolation transformer because all of my designs have a transformer at the output so the signal is balanced and floating. This means it does not matter that the unbalanced input of the QA400 connects one side of the balanced signal to ground. If you are testing an electronically balanced output with the QA400 then you might need to include an isolating transformer.

This interface is universally applicable. I use it with my Lindos test set and also with my Focusrite 2i2 as an interface to RME or REW. Pics to follow.

Cheers

Ian

I have updated the design of the passive test box. As well as the QA400, I have been using my Focusrite scarlet 2i2 USB audio interface coupled with the REW progam for audio testing. The new version of the test set overcomes a problem with the generators of both the QA400 and the 2i2. The problem is they cannot output high resolution very small signals, say -40dBu and -60dBu for testing mic pres. To do this you need to set the generator to 0dBu (to get max resolution) and feed it through a pad. As switched pad has therefore been added to the desing.

Also I noticed a very small amount of 50Hz hum *-100dBu) was picked up by the Carnhill transformer in some orientations. So I have replaced it with a Sowter 5069 which has a mu metal case (the hum is now absent).

Schematic of the new version:  https://drive.google.com/file/d/0B_n67A1hN3qtRmhlU3ZqUmMzdkk/view?usp=sharing

Pic of the first version: https://drive.google.com/file/d/0B_n67A1hN3qtTWVRVDJBZktHRlU/view?usp=sharing

and of the new version: https://drive.google.com/file/d/0B_n67A1hN3qtM3NFcUJ1a2dSYzg/view?usp=sharing

In some ways the REW is a better tool than the QA400. You can look as distortion spectrs in real time and it does regular sweep frequency responses, neither of which the QA 400 offers.

Cheers

Ian

Ian,

Don't want to hijack this thread, but just had a quick question regarding your Scarlett 2i2.  I just purchased one for on-site audio repair/testing.  First thing I found when doing a loopback test was that when the "direct monitoring" button was "ON" I get interaction between left and right channels.  I'm using RMAA.  In other words when direct monitoring, raising the left channel level also slightly raises the right channel and vice versa.  When direct monitoring is off, there's no problem.  Do you see this on your unit?

Rgds,
Jeff

ruffrecords said:

In some ways the REW is a better tool than the QA400. You can look as distortion spectrs in real time and it does regular sweep frequency responses, neither of which the QA 400 offers.

Click to expand...

The absence of a decent visualization interface and the difficulty in creating specific tests for the QA400 led to it being exiled on a distant shelf. For me it's an unfinished product, like a PC without an OS.

deveng said:

Ian,

Don't want to hijack this thread, but just had a quick question regarding your Scarlett 2i2.  I just purchased one for on-site audio repair/testing.  First thing I found when doing a loopback test was that when the "direct monitoring" button was "ON" I get interaction between left and right channels.  I'm using RMAA.  In other words when direct monitoring, raising the left channel level also slightly raises the right channel and vice versa.  When direct monitoring is off, there's no problem.  Do you see this on your unit?

Rgds,
Jeff

Click to expand...

I remember I had to very very specific about the the way the controls are set and also to use the XLR input rather the the TRS because they seem to act differently. I am pretty sure I have direct monitoring off because I think iy sums the input and output signals but I am not sure. Unfortunately there does not seem to be a block diagram for the 2i2 so it is hard to know just what goes on inside. I do not know if mine has the exact interaction you mention. I will check it out and let you know.

Cheers

Ian

abbey road d enfer said:

ruffrecords said:

In some ways the REW is a better tool than the QA400. You can look as distortion spectrs in real time and it does regular sweep frequency responses, neither of which the QA 400 offers.

Click to expand...

The absence of a decent visualization interface and the difficulty in creating specific tests for the QA400 led to it being exiled on a distant shelf. For me it's an unfinished product, like a PC without an OS.

Click to expand...

It's only advantage is that its answers are in dBu. Since I built the interface unit I have discovered REW. It is a much more competent piece of software and works well with my Scarlett 2i2, it is cross platform and free. Its only disadvantage is its answers ae in dBFS.

Cheers

Ian

ruffrecords said:

Its only disadvantage is its answers ae in dBFS.

Click to expand...

It's indeed the case of all generic-soundcard-based solutions; in the absence of a recognized standard for soundcard calibration, it is what it is... Fortunately, I think it's easy enough to calibrate using a DMM and being only half-clever.

ruffrecords said:

I have just finished building a simple passive test interface for use with my Quantasylum QA400 spectrum analyser. Schematic is here:

https://drive.google.com/file/d/0B_n67A1hN3qtd2U0Yk9XdGNPLTg/view?usp=sharing

It has an XLR input from the generator which is isolated by a Carnhill VTB2281 600:600 transformer.

Click to expand...

Begging the question of the "Loop Back" residual distortion of the transformer versus  a number of things, particularly frequency and signal level.

arnyk said:

ruffrecords said:

I have just finished building a simple passive test interface for use with my Quantasylum QA400 spectrum analyser. Schematic is here:

https://drive.google.com/file/d/0B_n67A1hN3qtd2U0Yk9XdGNPLTg/view?usp=sharing

It has an XLR input from the generator which is isolated by a Carnhill VTB2281 600:600 transformer.

Click to expand...

Begging the question of the "Loop Back" residual distortion of the transformer versus  a number of things, particularly frequency and signal level.

Click to expand...

Good points. The  output from the Scarlett is 0dBu at which level the transformer distortion is negligible (compared to the distortion in the tubes designs I am measuring). For lower levels (for mic pre inputs) I use an attenuator post the transformer. REW allows you to calibrate the loop back frequency response and apply it to measurements. I did this but it only adds about 0.1dB correction at 20Hz and 20KHz.

I have not checked loop back distortion at frequency extremes ( but with a 44K sample rate there is not much point above 10KHz). Low end does need checking. Thanks for reminding me. I am still trying to coax REW to set the Scarlett to 96K sample rate.

Cheers

Ian