Drums Volume is almost non-existent
  • Hello friends,

    I’ve just taken my Anushri out of storage. I assembled it long ago and everything was perfect then, but in testing it now, the drums volume is extremely low on its own. I’ve tested this both out of the main out and the drums out and the volume is very low in both cases.
    Any ideas on where I should be looking to discover the issue? I do have a multimeter I can use to test too.

    Thanks for your assistance!

  • Do you have a large signal on either sides of C2? at the junction of R5 and C4?

    A multimeter is not the right tool – an oscilloscope, or connecting the signal to an amp, is a better way of checking the audio path.

  • I also have an oscilloscope available, but I do not think I have a way of connecting directly to an amp. I will find my osc test leads and see what kind of reading I get.

  • Here’s the reading I got. I have to admit, I’m not savvy when it comes to using this Osc. I mainly use it to watch cool waveforms from the modular system :)

    The channel is set to 1v divisions, so is this considered a strong signal?

    <img src=“http://mutable-instruments.net/forum/uploads/FileUpload/1a/6878175d901ad6fd575b5de84425b1.jpg” />

    1008 x 756 - 156K
  • Ugh, sorry, I was testing C4. Let me go test the correct points.

  • Okay, I’m seeing 24v at both sides of C2 and the junction of R5 and C4

    <img src=“http://mutable-instruments.net/forum/uploads/FileUpload/52/ff2d15c93f5e82a8bc85c372327b77.jpg” />

    1008 x 756 - 164K
  • Looks like the ground of your scope is not connected to anything :)

  • I’m a total amateur. Totally forgot to attach my probe clip to ground.

  • Grounded, I get nada… no signal

  • I’ve got the ground clip of my probe connected to a post I have soldered into the open GROUND contact points near the voltage regulators. When the clip is attached and I test those points, I’ve got no signal. Remove the clip and I get those 24v readings. My oscilloscope is plugged into a grounded outlet.

    Am I failing miserably at this, or is there a malfunction somewhere that is not my brain :)

  • It’s normal to get weird readings when the oscilloscope ground clip is not attached to anything.

    But it’s not normal that you get no signal from the point I’ve mentioned – while you can still hear faint drums.

  • Okay, so I checked again and the signal was so low, that I didn’t notice it. I cranked up the volts division and I can see signal now. So it is there, but its’ very very low at all 3 test points.

  • What about the signal at pin 5 of the ATMega?

  • Same signal strength as the other points, very low, Just about 0.8V p-p

    Bad ATMega?

  • What is the shape of the signal you observe on the pin 5 of the ATMega?

  • Looks like an audio waveform in sync with the drums beats


  • Based on this image from the DIY Resources page, point 16 in the image, which is what I understand as pin 5, is where that wavefrom I recorded is from with the sequencer running.

    <img src=“http://mutable-instruments.net/forum/uploads/FileUpload/1f/5345e46d3906309f5dbf4d109af92c.jpg” />

    640 x 274 - 113K
  • That’s the right pin, but it’s not normal that you don’t see a PWM square signal – unless your oscilloscope has a very shitty bandwidth.

    I don’t think it’s a defective MCU – there might be another build issue, but I don’t really know which one.

  • It’s a 20mhz oscilloscope, so I expect I would see a 39khz pulse. I don’t see anything on pin 5 unless I have the sequencer running

    Okay, I see a pulse, but I have to crank the sensitivity so high (10mv) to see it.

    <img src=“http://mutable-instruments.net/forum/uploads/FileUpload/0d/77ef30645158a97dc2a408f3954fa9.jpg” />

    1008 x 756 - 275K
  • Sounds like that pin is connected to ground with a very low resistance (almost shorted). You could try to measure the resistance from this pin to ground. If there is a short, you can open up the board file in eagle, highligh the corresponding track with the “view” tool and see to which solder joints it goes. Then you have a rough idea where you could find the soldering error.

  • Measures 3.74 k ohms

  • Did you try to firmly push the cpu into its socketß Often the sockets fail.

  • Yah, it’s in there tight. I even removed it and reinserted it.

  • Any other ideas? :-/

  • just to confirm, do you wanna measure in AC mode??? then every PWM looks like spikes, cause well you only see the AC part, that means the signal changes, the DC parts, eg the bottom and plateaus of the PWM are basically DC appear as 0V, exactly like in your pictures…

    PS: I can probe aorund in my anushri and send signals how they should be, I think that might help, BUT this costs me time so dont expect this to happen fast :)

  • If I switch to DC mode, all that happens is the the whole signal measured goes from 0 up to about 4.5, There is no square wave/pulse/troughs, just the straight line that jumps up to 4.5 when set to 1v/division.

    I have no problem waiting. This anushri isn’t going anywhere. I appreciate the help :)

  • Are you confident about the soldering? Maybe you could upload some hi resolution photos of your board. Before we go any further, it could be worth checking the soldering first.

    Also it would be interesting to see how exactly you made the scope measurements. Where did your probe ground go? Maybe the point where you attached it to the board was not acutally sitting at GND due to a cold sodler joint. Just guessing.

  • > then every PWM looks like spikes, cause well you only see the AC part

    That’s not correct, the “DC component” you mention is modulated at audio frequency, so it’s not DC at all.

    So flipping the AC/DC switch should only alter the true DC component – shifting the waveform up or down by 2.5V.

  • haha nice one, indeed you intruige me to test it, but im too lazy right now, its true that AC just gets rid of any DC component, so your 5V dc indeed should just be plus minus around 0V.

    But :) depens a little on the timebasis, cause if the pulses are very slow it will look likes spikes again, dont know if anyone can follow my thought, maybe i hoook up the osci to a signal generator later and then tell with what frequency the PWM starts looking like spikes, my best guess would be around 100 Hz right now, only a guess.
    38kHz should certainly be fast enough, I assume, so pinch is right.

    nope: i was wrong, just tested it on an agilent with integrated wave gen, up to 5 MHz still viewable squarewave, in AC mode. So yes does not really matter if you try catching your PWM in AC or DC mode, as long as you are aware of what you doing. my bad

  • I’m pretty confident about the soldering. I’ve done a lot of these projects in the past decade+. But you never know.

    Note that in the first pic, you can see where there is a little hunk of solder stuck between to pads near the top right. I noticed it after taking the picture and i removed it. This did not change any of the behavior that I’m currently experiencing though. The last pic shows where I have the probe clip attached to a soldered lead onto a gnd pad.

    Hi-Res solder photos:

  • Can’t think of anything sorry

  • What kind of oscilloscope probe do you use? There are probes around with internal attenuation of 1:10, 1:100, or sometimes even 1:1000…

  • Next to the atmega, it looks like there’s a little piece of solder between two pins.

  • Yah, I mentioned that earlier. It’s gone and it didn’t change the behavior.

    My probe is x1 and x10. It’s on the x1 setting.

  • Anyone interested in taking a look at it in person?

  • Where are you located?

  • Tampa, FL, but I’d ship it if someone was willing to give it a once over.

  • I’ll take that silence as a no :)

    Mex, any chance you’ve been able to take readings for me to compare with?

  • still following, though right now not that much of time.
    Can you post again, what readings exactly you want me to do?

    Also, you can put the drums out on a seperate of those mini chinche plugouts, have you already tried that?

  • Yes, I have tried that. The output volume is the same.
    I don’t know what to test from here, this is outside my realm of expertise. If pichenettes is stumped, then I’m not sure where to go from here.

  • You better have it checked by someone, look over at muff´s in the DIY section maybe someone is willing to do it in your general area.

  • That’s a good idea, Varthdader

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion