Working on some new circuit designs that aren’t working out
I’m designing a little voltage controlled synthesizer board, but unfortunately there’s something wrong with my circuit and it literally hurts to listen to. I mean, it sounds pleasant enough, but the harmonics on the edge of the range of hearing are super, super loud. These nearly inaudible harmonics cause extreme ear fatigue, so much that my ears hurt for hours after listening to this for less than 30 seconds.
You may ask why I keep listening to it, well, I’m trying to fix the problem! And I can’t tell if the problem is there until I rewire and try again. Unfortunately by the time my ears hurt it’s already too late, there’s no “hurts a little bit” with this thing which is unfortunate (he says, as he types with a pounding headache).
Filtering the output causes even more oddball harmonic glitches (and loses many of the effects that make this circuit (otherwise) fun to play around with.
This is a project I’ve been working off and on with for months, and I think I’m just going to have to give up with my current schematic and design. Which is a shame because anything else I implement will take a lot more parts and/or a lot more design time.
For what it’s worth, this is the schematic for the core of the VCO: http://www.ecelab.com/circuit-vco-555.htm
The fun bits that I was adding all had to do with control voltage and modifying the control voltage before it hit the 555. In order to tame those harmonics I tried a 100 ohm/1mfd RC filter, I tried lowering the output voltage, I tried a number of other things a few months ago which I don’t even remember but still got ear fatigue after a fairly short amount of time. Ah well.
February 11th, 2010 at 5:22 pm
John, I don’t know what all you’ve tried for filtering, so forgive me if I’m retreading old ground.
The 555 is going to produce a (roughly) square wave output, which shouldn’t be unpleasant to listen to — it’s one of the foundations of subtractive synthesis, after all. So your description of the harmonic pain is surprising — yet I know exactly what you mean and from the lucidity of your description I can’t possibly doubt that you’re hearing exactly what you say you are.
I’m wondering about bad interactions with the 555 output going straight into the inductive load of the speaker. Have you tried running the 555 into a small audio amp (LM386?) and see whether that makes any difference?
Likewise, if you really do want to filter the output, I wonder about testing with an active op-amp filter rather than a simple RC.
Just a few thoughts.
February 12th, 2010 at 1:21 am
I’ve tried rc filters and single and double pole op-amp active filters, at a number of different cutoff frequencies.
That’s a good idea with the 386, but I’m doubtful and here is why. When I hooked it up to a scope, I noticed the waveform looked a little odd. Imagine the signal pegged at 5v, with extremely brief spikes down to 0v. The way the circuit varies the tone, it changes the length the waveform stays at 5v, but NOT the length of time the waveform stays at 0v.
So I can filter this pretty hard and get it down to a sawtooth, or near enough. But the amplitude is so low it’s hard to hear. Bringing the volume up with an external amplifier brings back the ear hurting.
April 17th, 2010 at 10:41 am
Could you put a picture or diagram of your breadboard up please? I’ve been trying to build this circuit from ecelabs with no luck.
April 29th, 2010 at 8:49 pm
Sadly no, I disassembled the circuit off my breadboard. I do remember that it was tricky to lay out. Make sure you have the chip the right way up and are connecting each pin according to the pin numbers on the schematic. The schematic doesn’t visually match with what the best breadboard layout is, and I found myself almost hooking it up wrong a couple of times.
April 29th, 2010 at 8:57 pm
Also, it’s not on the schematic, but make sure you have some beefy bypass capacitors near the power supply for the chip. On the timer edge, the 555 in this configuration pulls a ton of current. If you don’t have that capacitor in place the chip won’t be able to get enough power from the supply for its peak transient current needs. At certain tuning ranges the VCO actually stops emitting sound and pulls a little over 120ma (!) so watch out for that as well.
I used a 220 uf electrolytic capacitor in parallel with a 0.1 uf ceramic for the bypass caps.
See here for more info about bypass (also called decoupling) capacitors: http://www.seattlerobotics.org/Encoder/jun97/basics.html Also Wikipedia: http://en.wikipedia.org/wiki/Decoupling_capacitor