[solved] Bass Drum BUZZZ

Discuss Nava construction and related issues
  • does someone use isolated wires? After I did the bassdrum mod to get rid of the buzz I noticed that the bd circuit is pretty sensitive picking up hum and other interference from the wires, specially the power supply ones.. just try and move a wire closer to the area where q6 and q12 are :shock:
    further I still have some buzz, but it's much better than before, maybe it is just from the wires now, so I'd like to try isolated ones, any suggestion?

    I noticed another small issue, which might also be firmware related.. the kick drum level sometimes varies in one pattern, I couldn't replicate it so far. It's like I programm a pattern with 4 kicks and for example the first kick is just a little bit higher in volume than the other 3 kicks, but it's noticeable and also visible in the meter in the DAW. It doesn't depend if accents are set.. This seems to appear randomly, and just the kick is affected. I'll record a short demo in the evening, it's very subtile, but it's definately there :roll:
  • @AonFluX, i tried this too with a longer wire direct to the solder side, but at this point i already burned the M5218 :\

    @decoder, i had this sometimes too, one step bassdrum being louder then the others.


    i got some more M5218L from ebay, from 2 different sources, i replaced the AL with one of those, on the monitors it sounds fine, but on the headphones i hear some high frequency sound on the kick... never used it with the M5218AL with headphones so not sure it was there too... going to try the M5218L from the second source, or maybe this is also some interference from wires...
  • ok, good that I am not the only one with the different lvl thing :> imo that's a firmware problem too? not thaaat bad but needs fixing please :mrgreen:
  • decoder303 wrote:does someone use isolated wires? After I did the bassdrum mod to get rid of the buzz I noticed that the bd circuit is pretty sensitive picking up hum and other interference from the wires, specially the power supply ones.. just try and move a wire closer to the area where q6 and q12 are :shock:
    further I still have some buzz, but it's much better than before, maybe it is just from the wires now, so I'd like to try isolated ones, any suggestion?
    hey, i just build the BD with suggested fix, using a piece of utp 6cat wire (copper wire with isolation) and made a short vid testing the induced noise with my finger (which is exactly about what you say)

    https://streamable.com/u1kt

    i can also confirm that the "BUZZZ" still can be noticed at the BD full volume + full volume/amplification on my analog mixer (this video was recorded at this level, but only the induced noise + mixer noise is there, which is much louder compared to the "BUZZZ") maybe it can be fixed even better by placing q6/q12 as closer as possible to the opamp ic, but i think it's not worth the effort doing that on this board version and already happy with the current fix and this low "BUZZZ" level.

    this is how i done that :

    http://i.imgur.com/EmMosV6.jpg
  • yeah, it picks up any interference :> I actually meant shielded cables not isolated :D

    here is a quick demo of the bass drum level difference:

    https://www.sendspace.com/file/dunknb

    and here some screens, without any accent, with accent and with accent + total accent.. the bd on step 5 has obviously higher level, it could also be step one, sometimes it's step 9 and 13..

    Image

    Image

    Image
  • decoder303 wrote:yeah, it picks up any interference :> I actually meant shielded cables not isolated :D
    hehe, the funny thing is that i was thinking exact in that way... but the other problem is where to connect the shield - obviously, not to the GND, the whole fix is about to put the trace away from the ground plane. i was searching for an alternate solution indeed (without cutting traces), but decided just to do what was proposed by the designer (or else i would stuck with the BD section for a much longer time) and by looking at the pcb traces i noticed that the ground plane is too separated between some close points, so it can be parasitic inductance/capacitance due to the long ground current path, which can lead to some positive feedback on high frequencies which in turn can lead to the noise generation. effectively it can be solved by different placing of components (and so the traces to them) and/or changing the whole ground plane.
  • redstorm wrote:
    hehe, the funny thing is that i was thinking exact in that way... but the other problem is where to connect the shield - obviously, not to the GND, the whole fix is about to put the trace away from the ground plane.
    Would it help to use a shielded cable, but without connecting the shielding anywhere?
  • For shielding against electromagnetic fielda the shield has to be connected at one side, normally literature tells one side only because if you do it on both sides there is a huge risk for groundloops, but on one pcb there shouldnt be a problem qith gnd loops since the whole board is on the same gnd level

    Shielding is a bit of voodoo as is interference.

    Im also finishing my nava, yocto already finished but not tested causw waiting for a proper case. So i will do the bd fix in some days or weeks and will test around with that too.then i can tell more details of how i gonna to fix it.

    Cheers
  • about the lvl problem, I couldn't replicate the behaviour why some steps have a higher level sometimes. I only know that, if 1 step in a pattern has higher volume it will stay this way until the whole eeprom gets initialized. Might be a (not perfect but working) work around for people who also have this issue. Seems that somehow some information is "stuck" in the memory and stays there until the init procedure is done?

    the buzz isn't bugging me that much anymore since it's way better than before. Of course it would be nice to really get rid of it since at some processing stages of the kick the buzz still is quite noticeable.