Test tones into balanced inputs
« on: February 15, 2013, 06:28:48 PM »
When setting up or testing audio equipment, it is often required to present a test tone at the input of the device under test.

I am confused as to how this works with a balanced input on the DUT, if the test tone generator is unbalanced.

An example may be something like a 1kHz tone at +4dBu.

To connect an unbalanced tone generator to a balanced input on some device, do I need to convert the tone generator output into a differential output and thereby increasing the test tone by +6db when received by the audio device.  Or should I just create a balanced connection for the test tone, so that a +4dBu signal is carried only on the "hot" side of the connection?

I hope I am expressing myself clearly enough for someone to help enlighten me.

Thanks everyone.  Paul H


Re: Test tones into balanced inputs
« Reply #1 on: February 16, 2013, 06:23:18 AM »
As usual the answer is it depends - in this case on the topology of the input you are driving. I am pleased you used the word differential because most electronically 'balanced' inputs and outputs are differential. It is because of this that all the 'should I raise an unbalanced output by 6dB' questions arise. If the input you are driving is truly balanced then the answer is no. If it is not then the answer is probably.

Sorry not to be able to give a more definitive answer but only if the input you are driving is a transformer can you say a definite no.



'The only people not making mistakes are the people doing nothing'


Re: Test tones into balanced inputs
« Reply #2 on: February 19, 2013, 08:34:52 PM »
Thanks Ian! 

I was thinking that I could build a simple opamp buffer for the test signal, adding a balanced impedance network to the output, as per Bill Whitlock's Jensen white paper (section 2.4):

Would that be a reasonable way of presenting a test tone to equipment with a balanced input (of whatever topology)?


Re: Test tones into balanced inputs
« Reply #3 on: February 20, 2013, 11:23:14 AM »
have you considered a new tone generator? I have minirator mr2 that I use for in the field stuff. It has xlr and rca outputs on it.  For on the bench I just use a  usb daw interface that a small 2 input 2 output. Then using software can make it do whatever I need. I can generate anyone tone or other signal I want and can even use it as an analyzer with the right software.
You tell me whar a man gits his corn pone, en I'll tell you what his 'pinions is.


Re: Test tones into balanced inputs
« Reply #4 on: February 21, 2013, 07:33:30 AM »
There are lots of useful little boxes which do this in the studio. I use a Redbox unbalanced to balanced convertor which I have used for home studio work. To be honest I am not sure it made all that much difference when testing my DIY projects, which echoes some of the above observations!


Re: Test tones into balanced inputs
« Reply #5 on: February 23, 2013, 04:46:06 PM »
My Motu balanced I/O when connected with balanced loopback cable (output to input) shows a value in my RTA(s) which is 6dB less than I expect.

I have a CRO measuring each phase to ground and also the difference between those signals.

The signals on the hold and cold are equal in magnitude and 180deg out of phase with respect to ground. Across the hot and cold is double that magnitude.

When I convert the peak-peak values to RMS and dBu I compare that to what the RTA shows the dBu amplitude of the signal peak received and it is 6dB down.

I've confirmed that using various iso transformers and other test generators, multimeters etc.

In my case, the RTA sees the incoming signal as an 'unbalanced' in terms of magnitude.

I note that when I'm calibrating my RTA to absolute voltage and 600R loads and continue on my way.

If I want to display balanced signals on the CRO I use a couple of different iso transformers to unbalance or I measure with the cro on 2 traces, each phase to ground, flip one trace and add it the other.

On my system, this has a strong bearing on the reported dBu levels of noise floor. It's easy to have a system set to report stellar results. Without a Pro system to compare with and set you straight, you need to be vigilant that what you think is RTA 0dBu actually is when compared to the 'outside world'.

So, I cross check with as many bits of gear I have, to make sure I haven't got my figures wrong.

I would say I can (reasonably) accurately measure down around to 0.4mV ac on the CRO without using measurement amplifiers and maybe 0.1mV ac on the multimeters.   Thats somewhere around -75dBu to -86dBu. At that low level, it is 100ths of a uV differences to go lower, which I can't measure except thru my RTA.

So thats why my calibrations then indicate a noise floor on loopback of around -83.5dBu on my studio (Motu 2408MkIII) and -84.5dBu in my test bench (Motu 828MkII).

The difference with dbFS is that my Motus crap out analog-wise at around +10dBu (6.93Vpp into 10Kohm inputs I think), at which point distortion is quite high from there onto the digital clip. No idea what +dBu that is.

So, the figures for S/N and noise floor and so on reported by most dbFS audio test applications are generally 15dB or so better than what my RTA measures at, calibrated to a reasonable level of confidence to around -83dBu.

SO there you have it.


« Last Edit: February 24, 2013, 07:41:01 PM by alexc »
I ping therefore I am


Re: Test tones into balanced inputs
« Reply #6 on: February 24, 2013, 05:10:23 AM »
Evidently I am going to have to experiment!


Re: Test tones into balanced inputs
« Reply #7 on: February 24, 2013, 03:12:07 PM »
I use my old HTC myTouch 3G for quick signal generation. My scope says it puts out a very clean sine wave even when cranked and even a decent square. It's not ideal, but it works great for field testing.
- Rodney

" better call Kenny Loggins, 'cause you're in the danger zone."


Re: Test tones into balanced inputs
« Reply #8 on: February 24, 2013, 07:38:21 PM »
One thing I'm considering is using some of Rochey's Eden preamps as measurement amplifiers.

That's some super figures that could make a nice pair of instrumentation gain cells for boosting signals to measurable range for me, without adding significant amounts of their own distortion or noise components.

I ping therefore I am


Re: Test tones into balanced inputs
« Reply #9 on: February 25, 2013, 02:04:06 AM »
I recently bought a Bitscope mixed signal oscilloscope, with which I am still getting my head around.

It can generate test tones and waveforms, but I need to rig up a way of applying signals to gear.

My main issue was how to apply the unbalanced test tone to balanced gear.


Re: Test tones into balanced inputs
« Reply #10 on: February 25, 2013, 05:13:05 AM »
ISO boxes !! Or, you could get a good quality instrumentation amp to debalance. Or cheat and just check one side.

Sometimes they need to be different types, but you can start by making up a pair of 600R:600R and/or 10K:10K.

I used a commercial 4way balancing box and replaced the transformers inside to 600:600.
I use it when feeding an unbalanced sig gen signal to a balanced input or testing a balanced output with a cro, quick style.

Mostly does the job, albeit with a dB or so off loss.
Freq resp should be good (where low Z is appropriate). Some phase shift to be expected.

Generally though, I'm feeding balanced inputs and receiving balanced outputs with my audio interface, which is balanced.

For well behaved stuff, the signal on the + and - is equal and phase shifted 180deg. You can take a short cuy.
Not always though - some are not symmetrical around ground.
I ping therefore I am


Related Topics

  Subject / Started by Replies Last post
3 Replies
Last post February 21, 2006, 04:39:14 AM
by mick
0 Replies
Last post April 10, 2009, 09:18:51 PM
by dmlandrum
11 Replies
Last post July 29, 2010, 04:06:01 AM
by MartyMart
0 Replies
Last post July 19, 2011, 12:35:59 AM
by JohnCE