Isolate the audio input

Johnny2Bad

Member
2010-05-04 7:51 pm
I agree that an appropriate input transformer is the way to go ... not exactly cheap for the good ones but for example I was looking at a Jensen that was flat (less than +0/-0.2 dB deviation) from 5 to 40K and perfect for line level. $70 a channel, though. You may well be able to get away with less, depends entirely on the application.

Optocouplers are used in digital (toslink) inputs ... they don't have a stellar reputation even in that application although there are also those that argue they're fine and the Golden Eared are full of it, but they do eliminate any possibility of ground loop hum entering the input, which is certainly possible with S/PDIF coax or AES/EBU.

Is that the goal? I would go with transformers for analog inputs if that's the case. Considering that you probably can't listen to any non-live performance, let alone any except a completely un-amplified live performance, anywhere on the planet, without having some transformers in the signal path, I can't see what could be more professional; it's what the pros use every day.
 
Last edited:
Is this ClassD of fredos. I don't have amplifier of fredos but using the audio input isolator appears to be very good. It also has some examples of isolator circuit using optocoupler.
 

Attachments

  • optoORtranformer.jpg
    optoORtranformer.jpg
    88 KB · Views: 192
  • IsoladorAudioOpto.jpg
    IsoladorAudioOpto.jpg
    113.6 KB · Views: 190
Using any kind of input isolation will actually add noise since both line trafos and optocouplers are non-linear. The only advantage is that you eliminate the risk of a ground loop and incorrect connections.

It's most often used in pro audio because pro audio engineers knows that sound engineers are complete morons that will sometimes connect an XLR output from one amp to the XLR input of another amp. But the easy way to avoid damage is to just add a clamp on the input.
 
It's most often used in pro audio because pro audio engineers knows that sound engineers are complete morons that will sometimes connect an XLR output from one amp to the XLR input of another amp. But the easy way to avoid damage is to just add a clamp on the input.

It's used in pro audio to match impedances (in particular with microphones), in DI boxes (to allow otherwise incompatible gear to interact, eg iPod for sound checks, direct boxes for plugging guitars directly into consoles, etc,) and to allow the inevitable long cable runs without significant HF loss and eliminate ground loops. They also block DC, and RF interference. Surely these are not all undesirable features in an audio chain.

It is probably impossible to buy a recording that does not involve listening to transformers used to create it. If you have even one favourite recording, I am going to suggest they chose the audio transformers wisely.

There is very little to dislike about a well-designed and properly implemented audio transformer, sonically. There is much to dislike about poorly designed and improperly implemented transformers sonically. Mostly they are used less today because they tend to be heavy and expensive, not because they inherently sound bad. Far from it, but, like everything, the chosen unit must be appropriate for the task.

Almost every time I heard a "bad" transformer you could trace the essence of the problem to trying to get away with choosing a cheaper example when the design calls for something that actually works under all expected conditions. They are certainly not the ideal answer to every audio problem, but on the other hand I find my life much easier when I don't dismiss well known working solutions out of hand.
 
Last edited: