Author |
Message |
dentaku
Joined: Apr 24, 2013 Posts: 6 Location: Canada
|
Posted: Mon Sep 30, 2013 3:48 pm Post subject:
saw/triangle of timing cap out of a speaker on my breadboard |
|
|
Being very beginner at making sounds with hardware instead of software I've noticed that you can get triangle and saw-ish looking waves straight from the timing capacitor of most square wave oscillators.
As you all know, you can plug a small speaker into the square wave output of a simple oscillator (I've done this with a 555, 556, an old UJT I found in a bin and even an LM339 comparator) and you get sound.
BUT that nice TRIANGLE or curved SAW that shows up on the oscilloscope or even on the computer through the line in doesn't do anything when you plug in an 8 or 4 ohm speaker into the breadboard.
QUESTION?: What's the easiest way to hear the triangle or saw-ish wave output of the capacitor charging and discharging on a speaker plugged directly into my breadboard? |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Tue Oct 01, 2013 4:01 pm Post subject:
|
|
|
There isn't nearly enough power at the timing cap to drive a speaker. You can see it on a scope because the scope probe is high impedance and doesn't load down the circuit much. The 4 or 8 ohms of a speaker is almost like a short compared to the timing circuit and will just squash what little signal there its.
At the very least you'll need to buffer it, something like a TL07x opamp should do the trick.
While some circuits can produce sound through a little speaker, it's generally not a good idea to connect a speaker to something not designed to drive one. You're probably stressing whatever is trying to drive it and could potentially damage that part of the circuit. If you want to probe around with a little speaker, I would build a small speaker driver circuit. The line in of your PC should generally be safe as well, it is a high impedance input. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
dentaku
Joined: Apr 24, 2013 Posts: 6 Location: Canada
|
Posted: Tue Oct 01, 2013 5:32 pm Post subject:
|
|
|
Today I decided to go looking for information on impedance, buffers and impedance matching etc. so now that I read your post I guess I wasn't heading completely in the wrong direction.
Thanks for the info.
Could I use something like an LM324 or LM358 "voltage follower" to make a unity gain buffer amplifier like fig. 3 of this Wikipedia page?
http://en.wikipedia.org/wiki/Buffer_amplifier
I don't have much to choose from but I DO have a few salvaged (an a few never used) opamps I can try.
It seems to me that this could also be done with a FET or even a simple BJT although it might be finicky for a beginner like me to get working correctly.
I'm just experimenting with stuff that makes sound I guess, as a way to learn about electronics after a lifetime of working with software. |
|
Back to top
|
|
|
gdavis
Joined: Feb 27, 2013 Posts: 359 Location: San Diego
Audio files: 1
|
Posted: Wed Oct 02, 2013 3:23 pm Post subject:
|
|
|
That's the idea.
You can design a transistor buffer but that's trickier to configure. Opamps are easier to work with and perform better than a single transistor. They basically do all he hard work for you.
Impedance matching applies when you're transferring power, i.e. driving a speaker. You want as much power as possible from the amplifier to transfer cleanly to the speaker. This is achieved by matching the output impedance of the amplifier with the input impedance of the load (speaker).
The signal path before the amplifier is different. This is typically small signal, you want to minimize the power transfer. The ideal is to have a zero output impedance and infinite input impedance. Of course realistically this isn't possible but we generally try to have a small output impedance driving a large input impedance.
Which is why connecting a speaker directly to these circuits generally isn't a good idea. Small signal circuits aren't designed to provide the power to drive a speaker. That's what the speaker driver/amplifier is for, to bridge this gap. _________________ My synth build blog: http://gndsynth.blogspot.com/ |
|
Back to top
|
|
|
dentaku
Joined: Apr 24, 2013 Posts: 6 Location: Canada
|
Posted: Thu Oct 03, 2013 11:24 am Post subject:
|
|
|
Last night I was messing with a simple UJT oscillator that obviously puts out a pulse that you can hear out the speaker but the saw from the cap doesn't.
I was curious what would happen if I made it blink and LED like I did when I first started messing with electronics.
I ended up making a darlington pair with two 2n2222 NPNs which at slow speeds does exactly what I expected... the led fades in then quickly shuts off as I was expecting a saw wave would (slow attack, fast decay).
Then I realized that if it could light an LED it's totally powerful enough to drive a speaker and it did. It was even a bit too loud but less annoying sounding than a narrow pulse.
I guess the next step is to do this with an LM324 or something like that.
At least I'm learning something |
|
Back to top
|
|
|
Mongo1
Joined: Aug 11, 2011 Posts: 411 Location: Raleigh NC
|
Posted: Sat Oct 05, 2013 7:59 am Post subject:
|
|
|
Since you're learning about all this stuff, it might be helpful to really understand the sawtooth - what it really is.
Many oscillators work in a similar fashion. There is a Capacitor and Resistor. Current flows through the resistor, and charges the cap. The oscillator has circuitry that monitors the voltage on the cap with a comparator, and when it reaches a certain threshold, a switch is activated that quickly drains the voltage back out of the cap. Then the cycle repeats.
So - you only need to control the rate at which the capacitor charges to control the frequency of the oscillator. You can do that manually by varying a potentiometer, or electronically in several different ways.
The sawtooth is created by that charging cycle on the cap. It can only happen because when electrons are deposited on the cap, they have no way to leave (until we close that switch).
When you add a resistor from the cap to ground, you effectively give the electrons somewhere to go. If it's a high value (1 Meg), it doesn't matter too much, because the charging process can probably deposit electrons faster than they can be drained. The lower the resistor value, the more it will effect the process. If you keep dropping that resistor value lower, the draining process will eventually overwhelm the charging process, preventing the sawtooth from forming. When you add a speaker, you're effectively adding 8 ohms in that spot - it's almost a dead short. There's no way the sawtooth can form under those conditions.
Adding the buffer works because the input to the buffer is a very high impedence. It doesn't drain the cap (not much anyway). Once the buffer is in place, you can do just about anything to the buffer output, without affecting the sawtooth formation.
As you play around with this stuff, and examine different oscillator circuitry, it's helpful to really analyze which components are involved in the core of the oscillator. When you build the circuit, those components are vital to the performance of the oscillator.
Gary |
|
Back to top
|
|
|
dentaku
Joined: Apr 24, 2013 Posts: 6 Location: Canada
|
Posted: Sat Oct 05, 2013 11:38 am Post subject:
|
|
|
That was very well explained, Gary.
It also helps because that's pretty much what I figured was happening so at least I'm not heading in the wrong direction.
I'm going to see if I can use a buffered UJT oscillator as an LFO to modulate the frequency of another oscillator (maybe from a 555 or an LM339).
I've been trying to make oscillators in different ways so I can learn new stuff.
One night I even built a homemade vactrol out of an LDR salvaged from a nightlight and an LED to control a 555 oscillator using the audio output of a computer. Kind of a crude MIDI to CV setup. |
|
Back to top
|
|
|
|