Posted: Tue Oct 18, 2011 1:29 pm Post subject:
Diode Filter Subject description: Square to Saw?
I was messing around with diodes while in chat, and came up with something interesting. A schematic was demanded, so here it is.
It does work with just one diode, but the sound seemed to be very unstable (it would cut in and out and be filtered more or less). The extra diodes seemed to stabilize it (on breadboard at least).
diodefiltersamples.gif
Description:
Examples of how the filter changes the sound.
Filesize:
205.3 KB
Viewed:
527 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
diodefilter.gif
Description:
Schematic for the filter.
Filesize:
7.16 KB
Viewed:
548 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Here is a sound sample. I start with the source sound (16' and 8' square wave with mild lowpass filtering) and then turn up the diode filter. Then I play a little solo so you can hear how it sounds being played.
Thanks. I was told that it creates an amplified feedback loop and the diodes act as highpass filters.
I also added a 0.1uF cap to ground right before the first cap near the sound in. The feedback loop was boosting the system noise when no note was on, and the cap stops that from happening.
The cap at the sound in was changing my source, so I moved it to pin 2 on the opamp. It moved the spike to be at the front of the squarewaves instead of the middle. It makes some really cool sounds when combined with lowpass filtering. I'll record a demo and post later.
Joined: Aug 11, 2011 Posts: 411 Location: Raleigh NC
Posted: Tue Oct 25, 2011 10:13 am Post subject:
The only thing adding extra diodes SHOULD do is allow more current through that path. It may be that if there is a parasitic capacitance involved, that adding the extra diodes makes it bigger, allowing some sort of high-frequency bypass to occur.
It would be interesting to change it to one diode and a small capacitance value in parallel and see what that does...
Alternately you could try a large value resistor in parallel.
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
Posted: Thu Oct 27, 2011 6:39 am Post subject:
Under small signal conditions, when diodes are in parallel, only one will actually conduct - meaning that the other diodes are really doing nothing at all.
You can test this yourself with several LEDs. Put them in parallel (all facing the same way) and power with a 9v battery and a resistor (like 270 ohms). Only one will light. Now remove the one that lit up and a different one will light, but only one.
This happens because each diode is ever so slightly different. When a diode turns on, it presents a voltage across it's terminals (about 0.6 volts, the "diode voltage"). Let's say diode A is 0.600 volts and diode B is 0.610 volts. If they are in parallel, diode A will prevent diode B from conducting because it won't allow the voltage above 0.600 volts.
Diode parasitic capacitance is extremely small and not sufficient to cause any perceptible effect in an audio filter. _________________ FPGA, dsPIC and Fatman Synth Stuff
Time flies like a banana. Fruit flies when you're having fun. BTW, Do these genes make my ass look fat?corruptio optimi pessima
Joined: Aug 11, 2011 Posts: 411 Location: Raleigh NC
Posted: Thu Oct 27, 2011 7:06 am Post subject:
Quote:
Under small signal conditions, when diodes are in parallel, only one will actually conduct - meaning that the other diodes are really doing nothing at all.
You can test this yourself with several LEDs. Put them in parallel (all facing the same way) and power with a 9v battery and a resistor (like 270 ohms). Only one will light. Now remove the one that lit up and a different one will light, but only one.
Thanks for the info. It makes me wonder what's going on then.
The circuit is pretty weird to begin with, so maybe the results he was seeing were just coincidence, or chaos in action????
Or perhaps he's on to something. Maybe, just maybe, he's discovered a phenomenon that no one has seen before. Perhaps with more research and a great deal of effort, this phenomenon can be nurtured and grown like an acorn into a mighty oak tree. Perhaps it will one day be the basis of a whole new technology...A technology that can lift man from his earthly worries into a new future... A bright future.... A future of hope and peace.... Where people will care about each other.... Where no child need go hungry! A future where no one will fear the darkness ever again!!!
maybe 1 diode was getting overloaded, and having 2 gives a way for it to pass voltage once the threshold is met so that the first one doesn't overload.
I actually didn't notice any difference from 2-6. I had 6 holes in the breadboard so i filled them...
But just 1 would often be very loud at the start, and it would randomly cut in and out with some noise.
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
Posted: Thu Oct 27, 2011 10:17 am Post subject:
Cynosure wrote:
lol
maybe 1 diode was getting overloaded, and having 2 gives a way for it to pass voltage once the threshold is met so that the first one doesn't overload.
Not sure what an "overloaded" diode is.
Cynosure wrote:
I actually didn't notice any difference from 2-6. I had 6 holes in the breadboard so i filled them...
But just 1 would often be very loud at the start, and it would randomly cut in and out with some noise.
This is not something that a diode does. I would suggest that "cutting in and out" is an intermittent connection or another problem. Same with the noise, there is something more going on than exactly what is drawn in the schematic, some bad connection, some resistive solder connection, something intermittent, something that is hooked up in a different way than what you show.
Here is a question to answer - can you make another one, just like it, that performs exactly the same way? If not, then there is something that deviates from your drawing. _________________ FPGA, dsPIC and Fatman Synth Stuff
Time flies like a banana. Fruit flies when you're having fun. BTW, Do these genes make my ass look fat?corruptio optimi pessima
Yes - my breadboard is old and often doesn't make good connections, so it is possible that was the cause of the wierdness with one diode, and having a second one gave it another path to go through when the first wasn't connecting.
If you want to use this circuit, then try it with one diode first.
I have it soldered up now and it behaves similar, but the mixing pot has lower resistance and I moved it closer in the circuit to the sound source, so a little bit of the source sound always comes through unless I add some lowpass filtering to it. Also, as I mentioned in a previous post, I added a cap to ground at pin 2 of the opamp. It lowered some of the system noise, but also moved the spike to the front of the wave instead of the middle (I have no idea why that would happen).
I will try and record a vid tonight - or this weekend at the latest. Going to try one of those screen capture apps for my scope and record from the sound in on my comp.
The screenshots of the signals don't give voltages, but if i'm reading the schematic right, with an amplification of about 20 on the opamp, it's likely the output from it is a rail-to-rail square wave. That's then being high-pass filtered by the 1u cap/R combination, and then only the positive-going part of the wave will get through the diodes.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum