Archived from groups: rec.audio.tech (
More info?)
Pooh Bear wrote:
> Karl Uppiano wrote:
>
>
>>"Pooh Bear" <rabbitsfriendsandrelations@hotmail.com> wrote in message
>>news:41ABDBB4.1DDC77D1@hotmail.com...
>>
>>>Brian Allen wrote:
>>>
>>>
>>>>Graham Holloway wrote:
>>>>
>>>>>Does anyone here have any specific thoughts on the means of
>>>>>implementing
>>>>>balanced lines, ie. transformers, discrete circuits using IC op-amps,
>>>>>dedicated driver and receiver IC's, etc.
>>>>>
>>>>>I need to buffer other audio circuits and I have yet to find anything
>>>>>that I
>>>>>feel I can confidently use as I have never done this before.
>>>>>
>>>>>
>>>>>
>>>>>Graham Holloway
>>>>>
>>>>>
>>>>
>>>>Here are two resources I found useful. If you are looking for simple
>>>>implementation I have recently been toying with the burr brown DRV134
>>>>and INA 137 and have found them excellent for my purposes.
>>>>
>>>>http
/www.jensen-transformers.com/an/an003.pdf
>>>>
>>>>http
/www.dself.dsl.pipex.com/ampins/balanced/balanced.htm
>>>
>>>I see that Doug Self hasn't learnt how to design for equal inverting and
>>>non-inverting input impedances on his differential input circuits.
>>>
>>>Graham
>>>
>>
>>Either figure 12 or figure 13 look OK to me...
>
>
> I've never seen them actually implemented in any audio gear. Fine in theory
>
> Figs 9 and 10 are classic examples of poor design though. Widely seen in many
> pieces of kit though. Fig 10 looks like it's from a Soundcraft Spirit Folio.
>
> It iseasily fixable though. In Fig 9 change R2 and R4 to 3k3. This will give
> equal loading on both legs with a balanced signal.
>
Doesn'this then match the differential input impedance, but then
unbalance the common mode input impedance? This is what I measure when
I tested it, in fact it matched table 3 in D. Selfs article. Wouldn't
the common mode input impedance matching be more important because the
whole reason for balanced circuitry is to use the common mode rejection
for noise reduction. With audio signals having plenty of signal
typically we are not worried to much about exactly matching the
differential input impedance. Or am I mistaken and the differential
input impedance matching is more important?
Brian
> The scaling of values can be used to work this 'trick' for any gain. You end
> up with some oddball values - easily fixed by a series or parallel arrangement
> of resistors.
>
> Neve used to - probably still do - have custom resistor packs made for this
> topolgy.
>
> I've improved the Fig 10 style arrangement to at least offer symmetrical
> impedances in the '+4' setting.
>
>
> Graham
>
>