[sdiy] aging transformer replacement

[jhno]

thanks for the helpful advice; i had not thought to address transformer
buzz in a direct, mechanical way...

>With older transformers of the open construction kind the coils can get
>"loose" overtime.
>A blob of PU construction glue (this stuff expands to fill gaps, a glue
>cousin of PU foam) or
>sometimes just some tactically inserted folded paper, usually fixes the
>problem.

this sounds promising, i'll try a little gorilla glue.

my best theory is that when it heats up, the physical characteristics of
the transformer are altered such that a resonant mode is exposed. i don't
think that unusual current is being drawn.


>See above.  Another,  more likely cause,  is aging electrolytic caps,
>particularly those large ones in the power supplies.  These seem to be quite
>a common problem in a lot of equipment.

yeah, but is it typical that they would cause a transformer to start buzzing?


>> as an aside, i have bought a few 110 to 110 transformers for isolation and
>> power conditioning purposes. these have buzzed louder than anything. very
>> industrial.
>
>Oh really?  I can't imagine why that would be,  unless there was some defect
>in the construction of them.

i think that some industrial transformers, especially the big mothers, just
buzz and/or hum by nature. they are expected to be installed in places
where no one is present to be annoyed.

the devices i have seem somewhat common: 'mini-computer' power
conditioners. at least, common at Quinn's (bay area scrap electronics
mecca). typically just a big transformer and perhaps a couple of capacitors
sized to match.

at the time i theorized that power problems in our warehouse were causing
noise and artifacts in my equipment. my current view, however, is that
whatever the case, an additional isolation stage does not provide any
advantage. after all, every device already has an isolation transformer - !

i am not sure what difference big capacitors could make.

the more exciting (and possibly, yet more bogus) possibility would be to
run on 'balanced power' as seen in the audiophile world. manufacturers of
multi-K$ power conditioners claim noisefloor improvements. i might try it
sometime, if i can convince myself that a device will not mind receiving
120 volts center-tapped (60 per side, ground tied to center).

the audiophile industry has more snake oil and outright con artistry than
any other i have seen...


>I've actually seen a
>transformer blow, apparently to protect the fast-blow fuse connected to it
>from harm.   :)

ah, a design flaw. this circuit should not use a super-fast-blow
transformer. replace with sort-of-fast or slower coil.

jhno