From: Philip A. Marshall
Subject: Re: Looking for cheap isolation transformers
X-Newsreader: Forte Agent 1.92/32.572
Date: Sat, 19 Oct 2002 01:09:01 GMT
NNTP-Posting-Date: Fri, 18 Oct 2002 19:09:01 MDT
On Fri, 18 Oct 2002 01:08:57 -0700, Watson 'Atto Parsec' Name
>In article <firstname.lastname@example.org>,
>email@example.com talked about...
>> On Wed, 16 Oct 2002 19:02:59 GMT, Philip A. Marshall
>> >I'm looking for somewhere to get some cheapish 1:1 transformers to use
>> >as isolators in audio-frequency circuits. If anyone has any ideas,
>> >lemme know.
>> OK. I found some stuff at Mouser, and promptly figured out that I
>> didn't really know what I was doing :)
>> All I want to do is experiment with adding transformers to the inputs
>> and/or outputs of some of the guitar effects I'm building in order to
>> avoid ground loops with the power adapters and the signal cables. So,
>> I just want a 1:1 transformer, which i would assume that I need the
>> input and output impedences to be the same (that's the only thing
>> that's spec'd on most of these transformers, besides i/p and o/p
>> resistances). So, as long as the impedences are the same, and it's
>> listed as an audio transformer, I should be ok? (I'll of course use
>> the appropriate capacitors to protect it from DC voltages yadda yadda
>> My other question then would be how adding one of these would load
>> down my circuit? Would it just be like adding a shunt resistor with a
>> resistance equal to the spec'd resistance?
>> Also, to find the self-inductance of one of the sides of the
>> transformer, can I just use the impedence as the magnitude of Z, the
>> resistance as the real part, and then figure out the imaginary part?
>> That would get me down to 2*pi*frequency*inductance, but they don't
>> say what frequency the tests were done at. I would assume from the
>> other values given that it was 1KHz (ie, impedence variation is given
>> at 1KHz, and the 0db point is taken at 1KHz for the frequency
>> (here's a typical spec sheet so you can see the info I'm getting:
>> http://www.mouser.com/catalog/specsheets/114207.pdf )
>> ANY help at all here will be greatly appreciated. I'm better than
>> halfway through an EE degree, so I have most of the tools for
>> analyzing these types of circuits, I just lack some practical
>> knowledge of some of these things.
>Yes, you should not add a transformer to the input because the input has to
>be high impedance, 100k or so, and at low levels, transformers pick up hum
>from magnetic fields.
If I'm using a 1:1 transformer, it shouldn't change the input
>You should not have to add a transformer to the
>output because you can use a low impedance output such as an emitter
My point is not to change the output impedence, but to break the
direct ground connection from input to output, sortof like the same
reason isolation transformers are used to prevent people being shocked
in the case of a fault.
>But if you do add a transformer to the output, use a transformer
>that is for wide band audio, not a transformer made for narrow band
>telecomm use, such as for a modem. The transformer from mouser.com that
>you specified above is for narrow band telecomm use, not hi fi audio.
Er, yes. Thanks for pointing that out. I just grabbed the first spec
sheet I found, but I see from the frequency response that it would be
lousy, despite the fact that it's billed as an "audio" transformer.
>When you do find a decent wide band transformer, you may get a case of
>sticker shock; so then you will be aware of one more reason why designers
>try to avoid using transformers.
True, this may well be the case. I know the good hammond ones run
anywhere from at least $15 to $50 and above. I'll get some cheap
ones, see how they sound, and then decide. It's just an experiment.