Rehoused Delay has a high pitched whine-help?

[Stompin Tom]

I recently rehoused an old Ibanez DL5 (soundtank digital delay) because the flip flop switch stopped working.  I also wanted some more space to do some mods... In anycase, I bypassed the fet switching on the board (so it's on all the time now) and wired a true bypass switch with a 1m pull down resistor.  The problem is that there's now a high pitched whine, even when the unit is bypassed.  This whine changes in pitch when I rotate the delay time pot.  I've also found that if I put a buffered pedal (like a boss chorus) before it in the chain the whine disappears.  Does anyone have any ideas?  Is there anything I can do to filter out this whine?

[Joe Kramer]

Hey Tom!

I rehoused an off-brand analog delay in a hammond box so I could get true bypass, and wound up with a similar problem.  Even though mine is a BBD delay and yours is digital, I think you're getting clocknoise bleeding into the output/input.  There are a few ways to deal with this, and one or all of them may help.  First, keep the input as far away as possible from the clocking part of the circuit.  Also, keep the DELAY control, which carries the clock pulses, away from the input.  It's a good idea to shield the wires going from the DELAY control, as well as the input and output wires.  Make sure the ground connections to your new box are nice and solid (I assume you used a metal box).  If none of these help, you may have to add power supply decoupling caps to the ICs, which will filter high frequency junk out of the power rails.  It may be possible that the FET switch was filtering out some of the very high frequency garbage floating around in the circuit.  In that case, you could add some small capacitance at the input to filter off RF/VHF.

I did all of these to my rehoused delay and now it works fine.  Hope some of this helps you.

Joe

[Stompin Tom]

Thanks Joe.  I'll try moving things around in the box and changing the wires to the delay pot to see if that helps.  It's all pretty tight in there, so I'm hoping that's all it is.  In case it doesn't, how would I add power supply decoupling caps to the ics?  Is there somewhere I can read up on this?  I've built a few pedals and modded a few more, but I'm still don't know much about the more complicated circuits.  A final question, wouldn't adding a small cap to the input change the guitar tone (like in distortion and booster boxes)?  Seems like it would, but maybe not noticeably?  Thanks again.

[Joe Kramer]

In case it doesn't, how would I add power supply decoupling caps to the ics?  Is there somewhere I can read up on this?  I've built a few pedals and modded a few more, but I'm still don't know much about the more complicated circuits.

First off, I can tell you what I understand, but I don't claim to be an expert.  Always verify information by comparing at least two sources.   

A power supply decoupling cap is essentially a huge filter cap.  DC power is like the life-blood of the circuit, and you want that blood to be the purest, flatest DC possible.  Anything (like clock noise) that causes ripples and fluctuations in the DC, essentially making it something more like AC, will be "imprinted" on the signal going through the circuit: that means noise, distortion, etc, and that's bad.  A cap from critical power supply pins to ground will help filter and flatten this out, and incidentally act as a "power reservoir" in times of heavy current draw.  To do this, locate the positive supply pins on the op amps and put a cap from there to ground.  This cap should be placed a close as physically possible to the actual power supply pins, and can be added to the foil side of the board if it's a small enough cap.  Usually a .1 uF ceramic is recommended, but depending on the circuit anything from .1uf ceramic to 10uf or more (tantalum) will work.  Basically, just as in audio, the bigger the cap, the more filtering, and also the more of a "reservoir" it will be.  In general, this kind of power decoupling is considered good design practice whenever you're concerned with the cleanest possible audio signal (which is not necessarily all the
time!  ). 
 

A final question, wouldn't adding a small cap to the input change the guitar tone (like in distortion and booster boxes)?  Seems like it would, but maybe not noticeably?  Thanks again.

The standard electric guitar/amp combination has almost nothing going on above about 4KHz.  You can check this by running your guitar through an EQ and tweaking the 10KHz control.  It might cut some of the noise, but nothing of the guitar tone.  So, you can safely limit the upper frequencies to where the guitar tone isn't affected, but the high frequency garbage like clock noise (say, 10KHz and above) is kept from poisoning the works.  Without seeing a schematic, I can't tell you where exactly to put this cap or how big it should be, but something in the range of 100-470 pf from the input to ground would probably work.  With any luck though, the wire shielding/routing fix will help, and you won't have to resort to more extreme measures.

Joe

[Stompin Tom]

Joe, your explanation makes perfect sense... thanks again.  Now I just have to experiment with the delay to see what'll work.