i have 2 WTPA and both have the same playback problem

hi
i bought a WTPA kit when it came out and finished it .
everything work but there's a playback problem
here's an audio example http://www.sendspace.com/file/xmslzc (first the sample then the WTPA playback)
i rechecked everything, replaced some component, tried different power supply etc ......
i started to think that the ram chip was broken and ordered a new kit but this time with presoldered ram .
i finished it and BAM !!! exactly the same problem.
what could possibly cause this problem twice in a row ?
thanks in advance for your help .
Loïc Moreau

Comments

  • bump
    nobody has/had a similar problem ?
  • I'm afraid my problem is different. But your file sounds like some sort of effect is on. Did you try turning down the effects?
  • thanks for helping
    yes i tried so many things to make that stupid cyclic noise desseapear that i'm starting to think it must be a really obvious mistake .
    i tried switching all effects on and off, i also tried the bail! fonction but i still get that annoying noise
  • Hey Loic --
    I'm afraid that what you're hearing is digital bleed into your audio path. There are a couple things you can do to lessen the problem, but it's probably not something you did wrong -- it's a limitation of the design, at least the way it is now. Although yours does sound particularly noisy....

    Make sure your level going into the ADC is as high as possible before clipping. This will help your SNR. You can use your gain control to help with this too, since most of that noise shows up after the gain stage. This is the most helpful thing you can do. Also make sure that the body of your potentiometers is soldered to the ground plane (the big holes in the board where the pots mount).

    W/r/t board changes:

    First, you can cut the two traces on the top of the board immediately above VR3. These are used to feed the overdub circuit which is not used, and which makes a connection from the "clean analog" side of the board to the dirty one. These connections are unnecessary.
    Next you can cut the trace going to pin 34 of the AVR. Same reason as above.
    Next you can cut the trace from pin 7 of the TLV2464 to pin 33 of the AVR, and replace it with a shielded wire with a little slack. The shield of the wire should attach to analog ground (near the quad op amp). This should make an audible difference in the noise floor.
    Finally, you can beef up the values on C17 and C15, although I am not sure how much this will help.
    Also, If you don't use a guitar input ever, you can reduce R12 from 1M to 100k or so. This will help keep noise pickup down at the input.

    This noise pickup is one of my biggest priorities with the next version of WTPA, because it bugs the hell out of me, too. The AVR's ADC is actually pretty noisy, so I'll never be able to get rid of the noise entirely (in WTPA, anyway) but I bet I can do an order of magnitude better than it is now. Cross fingers.

    Hope this helps,
    TB
  • hi todd
    thanks for helping
    i tried everything you said but it doesn't work better .
    it's so frustrating i have 2 wtpa and both are useless.
    i notice that the cyclic noise volume is proportional to the volume of the recorded material .
    is that normal ?
  • Weird. If you did all that stuff (board changes too?) then maybe something IS wrong with your WTPAs.
    There should be a little noise, especially if your recording is quiet, but nothing that would make the device unusable.
    Is your signal source a laptop, by any chance? Once in awhile I've gotten noise from a laptop sound card that I've occasionally (incorrectly) attributed to WTPA.
    Other than that, is your wall wart a switching supply and if so can you try changing it?

    Try running the thing off a 9v battery, ground the input, don't attach any cables, and record silence with all the gains all the way down.
    Then, once you have the silence recorded, connect the output to something like an amp or a mixer and listen to it.
    At that point you should have a little noise, but not much. Sometimes it's hard to tell how loud the noise is in this situation because there isn't a signal present.

    Don't adjust the output levels, and plug an ipod or mixer or something into the input and the input gain at like 12 o'clock. The recorded output should be REALLY loud compared to the noise floor.

    Let me know, and sorry,
    TB
  • hi todd
    thanks a lot for the help.
    i feel so stupid i tried the wtpa with a battery and the noise is gone .
    i tried with a lot of different wall wart including the "ultra clean power supply" i made for my stompboxes http://www.generalguitargadgets.com/diagrams/ultra_clean_ps_sc.gif but they all make noise.
    now i need to find or make an "ultra mega super clean power supply" cause i hate batteries
  • Don't feel stupid. You got your problem fixed and that's what matters!

    Still....

    While the above schematic is not, say, medical clean (the 317 is nothing special, though the designer did go to a lot of trouble to bypass the adjust pin, which helps) I bet it is still WAAAY cleaner than most average wall warts. And since you had what sounded like digital noise on the WTPA, it seems weird that it would be coming FROM the PSU itself (since the above schematic is a linear supply).
    Seems more likely that there's some weird cross-talk involving your power connection and another signal connection, or some ground loop, or (less likely but still possible) digital noise on your power line.
    I'd check that stuff out before warming up the iron to build some crazy toroidal supply (not that it wouldn't be fun).

    Using batteries is not only a good way to get "clean" power, it's also a great way to "float" the circuit to see if your troubles are ground / reference related.

    But yeah, other than that I hate batteries, too.

    Also -- kudos for sticking with this problem for this long and getting it figured out!
    Best,
    TB
  • One way to make a really silent power supply is to actually use a rechargeable NI-MH 9V battery to power the WTPA, and then use the mains power pack in parallel to charge the 9V battery AND power the WTPA at the same time. The 9V battery acts as a super huge smoothing capacitor while the power pack is running the WTPA, and is also handy as portable battery power when not connected to mains. I discovered this when I had a guitar pedal that was particularly noisy when powered from mains but silent on battery.

    Cheers, graham
  • That's a good idea. However, the lazy engineer who made the pedal should get smacked up. I don't want to get smacked up :-)
  • edited August 2009
    thanks gmeredith i'm gonna try that !
    i killed a few batteries allready
  • You may have to put a resistor on the input to the battery if it charges too fast and the battery gets hot after only an hour or so of charging (the 9V ones don't have very high capacity). Just keep your eye (or hand) on the battery temp the first couple of runs.

    Cheers, graham
  • hi
    here's an update on the situation .
    i have use the wtpa with dying batteries with succes but it never lasted long so i got pissed and stopped using it.
    the other day i listened to sample i made with it and decided to solve the issue .
    a friend gave me a lab power supply so i tried powering the wtpa with it .
    my wtpa works with 7v no more no less .
    if use a little bit more that cyclic noise come back, if i use a little bit less the wtpa doesn't power up correctly.
    so i told myself there was something fishy in the power supply section and replaced the 470uf caps, 1n4004's and 7508.
    i still had the same problem so i finally made a vero the size of a stamp to get exactly 7v with a lm317 .
    now it's working but i don't have any idea what cause that problem and why both my wtpa's have it .
    hope that may help someone with the same problem.
  • edited May 2010
    Hi,
    I had similar cyclic noise as you described, I tried your fix using the LM317 and it solved my noise issue.
    Maybe I' ll experiment with ferrite beads and the 7805 regulator to see if it also removes the noise.

    Thanks!


    Edited for spelling
Sign In or Register to comment.