Quote:
Originally Posted by spritzer /img/forum/go_quote.gif
The hum is a well documented problem and it's because the diodes and caps in the bias supply are shot. Replacing them will fix it.
|
Good to know. Those caps are of questionable quality anyway. Most of them, at least.
Is there any wisdom passed down through the ages regarding the voltage ratings? 1kv caps are no problem to source - i can get very nice MLCCs at 1kv - but some of them are rated for 2kv, and there is a 220pf 3kv cap in each earcup.
Surely 3kv is way over the actual voltage in the circuit, maybe they just wanted thicker dielectric for better frequency response? At 3kv, the replacements i can get from digikey or mouser look a heck of a lot like the original. At 1kv, I can get silver-mica caps. Surface mount, but large enough that they should be able to straddle the pads on the board.
Quote:
A normal 600v teflon wire is fine for the bias supply. If you want to be extra careful you can always put heatshrink around the wire or encase it in a teflon tube. |
I was actually thinking about using the existing captive power cord. It's just the plug on the end that's gone missing.
I have about 14 pounds of misc teflon (PTFE) insulated silver plated stranded wire, mostly 22awg. A lot of it is shielded twisted pair, with a teflon outer jacket. This stuff is mil-spec (NASA surplus, I'm told), and highly abrasion resistant (read: stiff and waxy, not soft and cuddly).
I don't have the mil spec number for it, but it should probably be able to do the job. I would have to get a tighter strain relief for the E.9. It's bright white, but i have plenty of pleasing black-with-white-stripe braid sheathing.
Unfortunately the shielded twisted pair is a bit stiff for headphone cable use. The ESP9 stock cable is pretty short, and the extension cord that came with my kit is very stiff. I suppose i should investigate some of this 'star quad' stuff. Or i could just bundle 5 wires together inside some braid sheathing, but Koss seems to have used shielded 4-condutor.