balanced/single-ended testing jig

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

blandman74

Well-known member
Joined
Aug 4, 2005
Messages
67
Location
Washington, DC
Hello all, this week I fixed someone's Sta-Level and realized to my chagrin that I do not have a good way of testing the inputs and outputs of balanced audio gear.  My bench looked ridiculous, and the rat's nest of test leads picked up a lot of distracting noise.  I am getting more into repairs and I am pretty sure more balanced gear is going to show up for work.

I wonder if people have thoughts about a simple unbalanced-to-balanced driver for testing purposes.  It would be coupled with a balanced to unbalanced receiver so I can test the outputs.  The driver would deliver a signal/sweep from my sig generator; the receiver would go whereever - 'scope, amplifier, network/impedance analyzer, etc.  This would just be a one-off build for testing performance of audio gear for either repair, or even for design purposes when I have the desire to build myself (or, usually, someone else) a piece of audio gear. 

I have a few questions:

(a) I was thinking that I would just use a THAT 1646 (driver IC) and a THAT 1200 (receiver IC) for each purpose.  Do people have experience with these?  I don't expect to be using long cables, so right now I am thinking about leaving off any EMI protection on the outputs of the 1646. 

My reason for going with this approach, instead of a transformer, is that I want low distortion at a reasonable price.  My gear is humble but someday might be upgraded; so, for example, my sig generator has -55dBc distortion up to 100kHz.  I hope to get something better than that someday.  THAT's datasheets claim far better noise + THD figures than most transformers, I think.  (I *think*.)  Should I just go with some Edcor 10k/10k transformers?

(b)  I also plan on adding some dip switches to the driver's output that allow me to deliberately unbalance the output, by up to 600 ohms.  I wonder what people think of that - useful?  I was thinking 150 ohms, 600 ohms, on both sides (using 1% resistors) so I can also dial-in the (balanced) output impedance of the driver.

(c) Similarly, with the receiver I would have matched resistors, on dip switches, to set the input impedances, at 600 ohms and 10k, and perhaps 100k.

(d) I was going to make this without a chassis ground so that the jig would sit at the ground established by the DUT.  Or, maybe, I should install a ground lift switch?

 
Not very scientific but I checked a few balanced outputs by sending sine signal and viewing positive and negative signal output with a two channels scope. It helps to see any mismatch problem. With pseudo balanced output, I just check positive signal like an unbalanced output.
 
blandman74 said:
(a) I was thinking that I would just use a THAT 1646 (driver IC) and a THAT 1200 (receiver IC) for each purpose.  Do people have experience with these?  I don't expect to be using long cables, so right now I am thinking about leaving off any EMI protection on the outputs of the 1646. 
As has already been mentioned, balancing teh generator is not necessary; in fact sending teh unbalanced signal to either input leg is more informative. But the balanced line receiver is a near-absolute necessity. The 1240 is adequate. I've used a lot of them without any issue.

(b)  I also plan on adding some dip switches to the driver's output that allow me to deliberately unbalance the output, by up to 600 ohms.  I wonder what people think of that - useful?  I was thinking 150 ohms, 600 ohms, on both sides (using 1% resistors) so I can also dial-in the (balanced) output impedance of the driver.
Unless you want to test many units that require impedance matching, it's not very useful in day-to-day. For noise measurements, you want to use dummy plugs.

(c) Similarly, with the receiver I would have matched resistors, on dip switches, to set the input impedances, at 600 ohms and 10k, and perhaps 100k.
Again, that may not be extremely useful for modern equipment.

(d) I was going to make this without a chassis ground so that the jig would sit at the ground established by the DUT. 
You wouldn't be allowed to sell it for safety compliance, but if it's for your own use, you may do that, but you have to know it's at your own risk.

Or, maybe, I should install a ground lift switch?
You may, but you must retain some earth continuity; typically a parallel RC (100R/0.1uF) will do.
 
I have a test kit that you may be interested in.

https://groupdiy.com/index.php?topic=60805.msg770297#msg770297

You can also use it for non 500 series equipment. Power it externally using +-16V, and connect the input and output of the equipment under test to the TRS jacks.
 
This is an obnoxious answer, but if you can find an older Audio Precision test set, the mundane stuff like dealing with balanced and unbalanced gear, and having a quality generator with some good measure of calibration can help out a lot. For example, knowing the gain of a circuit accurately can help, as can knowing the frequency response. If you get a nicer one with a distortion analyzer that can show you an FFT of the residual, it can show you an enormous amount of information about a circuit - what interference makes it into the output, power supply noises, basically a lot of "small" stuff that can point directly to a problem that can be fixed.

Older units are out there, available used, and sometimes you can find one for not all that much money, especially compared to the cost of a competent generator, meter and distortion analyzer.

Again, sorry for the "let them eat cake" response, but a good AP test set is really a nice thing to use, and you can sometimes find them for very little money, especially compared to their new price.
 
Thanks for the responses.  I went ahead with the 1646 balanced driver only.  And yes, someday I will get an ATS or similar, it would indeed be nice to have that stuff taken care of.

I respectfully disagree that it is unnecessary to send a balanced signal to test gear.  I would think the usefulness of this would be obvious, although the point about *also* sending single-ended test signals being useful is well-taken. 
 
The correct answer depends on what you need to test.

It may be useful to have a signal source that exceeds typical audio bandwidth and signal levels (+26dBu seems reasonable and cost effective but transformer output to +30 dBu are not unheard of). 

Balanced-differential inputs can test both balanced and single ended.  You may want to include CMRR test capability (both inputs and outputs)  to confirm proper operation of DUT (device under test).

If testing hot output gear, you need a hot bench.

JR
 
blandman74 said:
I respectfully disagree that it is unnecessary to send a balanced signal to test gear.  I would think the usefulness of this would be obvious, although the point about *also* sending single-ended test signals being useful is well-taken.
Care to explain this "usefulness" ?
 

Latest posts

Back
Top