High pitch oscillation in a digital LFO trem....

[G. Hoffman]

I'm building a trem using an MV-53 LFO and, essentially, the audio path from a Tremulus Lune.  I've got it passing sound, but I've run into two problems.  There is a 10k (or there about - actually closer to 9k) oscillation in there somewhere, and it is bleeding through the power supply to the analog audio path.  The LFO and the audio path are on separate boards, and the only points of contact are the 9V, the ground, and of course the LFO's output, which has a separate ground connection, and is of course optically isolated from the audio path.  When I run the LFO and audio off of separate power, the oscillation goes away, so I'm assuming that is where the problem is coming from.  BUT, when I tried to isolate them with a resistor, the audio path stopped passing audio.  So, and suggestions on getting rid of the oscillation?  Altering the ground scheme didn't help, only powering it from separate batteries.

The other problem is that the depth control (identical to the one seen HERE) cuts out at the middle 1/3rd to 1/2 of its range, or rather, when the depth control is in that middle range the audio cuts out.  I've looked at the output of the LFO on my oscilloscope, and the problem is NOT in the output of the LFO - the LFO out goes up and down exactly the way it should, but for some reason by the time it gets to the other side of the opto (VTL5C2) it is cuts out the audio completely.  I've tried different values of load resistors on the opto, and there doesn't seem to be any change.  I've played with the various trimmers, and it doesn't help.  I've also replaced the pot, to no avail.

So, any suggestions on the next trouble shooting step?


Gabriel

[cloudscapes]

10khz sounds like the PWM driving the LFO output. add a low pass filter right after the LFO out. as physically close to the LFO ic as possible. 0.1uf should do it.

[MoltenVoltage]

Hi Gabriel,

Working on our new line of chips (in the beta stage), we have made a lot of progress in quieting down digital/analog circuits.

Here is a link to a revised datasheet for MV-53 posted tonight:
http://www.moltenvoltage.com/downloads/TW_LFO_Sequencer_-_MV-53_-_Datasheet_-_v2.pdf

Note that the other schematics listed on the main page, http://www.moltenvoltage.com/mv-53.html have not yet been changed.

On datasheet page 2, you will see there is now a separate ground for the pot inputs, AVDD, and pin 27, and and the analog power to the chip is more filtered.  There is also heavier filtering on the power supply at both voltage regulators.

There is also a separate analog ground for the op amps (which will also be the ground shared by your analog circuit).

The ferrite on AVDD is not absolutely necessary but helps quiet down the analog power to the chip.  For thru-hole, we have been using the Laird Ferrites.  998 ohms is the highest value we could find at Mouser.  Part # 875-28C0236-0EW-10.  If you are using SMT, try 2k5, 50mA - Murata - Mouser # 81-BLM18BD252SN1D

I can't explain why the depth control would be cutting out the audio when in the middle of its range.  Are you sure you have the Depth pot going to 9 volts, rather than 3.3?

Please keep us posted.

Thanks!