From Oakley Sound via analogue Sent Mon, Jun 18th 2018, 09:37
From Oakley Sound via analogue Sent Mon, Jun 18th 2018, 09:37
> Nice to know there's noise cancellation, even if it's not true balanced. It is true balanced. Balancing is about matching impedances in the hot and cold lines - it does not require differential signalling. Solely balancing output impedances is sometimes called 'impedance balanced' but this is only an indication that it is not differential. It is still balanced and can be treated as such. The problem with just matching impedances is the quality of the impedance setting resistors (1K is surely too high by the way) and the ability of the single op-amp output to perform as a zero ohm voltage source across the whole audio frequency band. I'm not sure a TL072 is quite up to the job as its open loop gain falls pretty fast. Generally the lower the output impedance the better but most op-amp stages will perform badly with long lengths of cable so those output resistors tend to be quite high. Matching of input, and output, impedances between hot and cold can make all the difference when it comes to how much noise rejection can be achieved. The traditional way of creating such a balanced output usually involved a transformer and it made sense to have inverted and non inverted signals present as a differential signal. Differential signals are easily created with a centre tapped transformer. As Brian says this gives a 6dB increase in maximum headroom over non differential systems. Electronically balanced output stages (EBOS), ie. ones done with two simple op-amp stages, are not necessarily better than 'impedance balanced' outputs. Using a non balanced jack (TS) in a balanced TRS output will short the cold output and lead to excessive currents being drawn; the output impedance setting resistor of the cold output having to take the full output signal. AD8200 users take note. Feedback systems can be employed to prevent the cold output being active but this can complicate the electronics and cause possible instabilities under some loads. THAT Corp have a nice little 8-pin driver chip that does the job admirably, although discrete versions can be found in most Soundcraft desks for example. Transformers still out perform electronic input stages with regards to noise rejection. Although THD and bandwidth accuracy are usually better with an all electronic system. Probably the best noise rejection electronic input stage is obtained by Bill Whitlock's In-Genius input stage: http://www.thatcorp.com/1200-series_High_CMRR_Balanced_Line_Receiver_ICs.shtml Tony www.oakleysound.com