[quote author="CJ"]Lots of hum on the non transformer piezo input side due to unbalanced guitar type signal.
I wonder if I changed that first stage to differential input on the piezo channel would help?
.....Some clown camer in here and started talkin bout batteries being noisy, remember that?
I didn't wanna say anything, but batteries have capacitance that shunts any noise that it generates.
Battery power noisy? Come on!.....
Is there any way to put a slight plus bias on those coupling lytics?
I read that if lytics are reversed biased, they can generate noises and pops for hours after the event, until the effects go away.
Food for thought.
[/quote]
(1) If the piezo is truly floating and the cable is really 100% effective for electric fields you shouldn't have hum. A diff input would be pointless unless you are trying to reject conducted errors, or had no choice but to run unshielded. Make sure the pickup itself is adequately shielded too.
(2) Batteries are not completely quiet, self-capacitance notwithstanding---they couldn't be, given the electrochemical processes at their heart*. The noise is mostly at low frequencies. I posted a link in the middle of a regulator discussion thread to something showing actual battery noise measurements a while back. What of course they are superb for is isolation. Given that your amps have good PS rejection they should be wholly adequate.
(3) 'Lytics get into trouble anywhere you want to look depending on how susceptible the circuit is. Most leakage current under
normal bias is circa tens of nanoamps for values in the low single to two-digit microfarad range. The CYA spec of manufacturers is much much worse, so much so that in many apps they would be unusable. And there is noise in that leakage current too. I recommend bipolar if you're not sure what the bias is going to be.
When the application is uncritical and the reversal of bias not sustained, I've put 50V 'lytics in circuits that occasionally see 2-3V reverse. Once, when I missed a silkscreen reversal in a bypass for a uC and associated stuff on a little board in a satellite powered speaker, the part which may have been a 10V or 16V one was subjected to continuous reverse
![Oops! :oops: :oops:](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
5V
Nervously testing some units, I found that the current never got horribly large initially, and then over time the 'lytic reformed and worked fine! I forget how many we had shipped at that point (maybe of order 10k), but the relief was palpable.
*Standard cells are pretty damn good, but you can't draw any current from them. Even then NIST had banks of them to average for the standard voltage reference. I think now they have tied the volt to frequency using jospehson junction paraphernalia.