modifiying a siren circuit based on the 2N6027

Thread Starter

lowrise4

Joined Sep 6, 2010
34
I asked this question about 2 years ago. I did get a suggested solution, but it didn't work as I had hoped, so I'm re-visiting this project...to hopefully get it to work like I had originally envisioned.

It is based on project 11 in the 'Make: Electronics' book. It's a siren using two 2N6027 Programmable Unijunction transistors...like in the attached pic. I drew the circuit orientation left to right (like my breadboard) rather than top to bottom which seems to be the conventional orientation. My siren circuit works perfectly when I do it as shown. The bottom (audio) oscillator makes an audible high-frequency tone, which is caused to waver up and down in pitch by the low-speed oscillator upstream of it and connected to the audio oscillator's gate.

What I've been wanting to do (unsuccessfully) is add a third (very low speed) oscillator upstream of the low-speed oscillator...to make the entire siren sound itself slowly rise and fall in pitch. So I made a third very low speed oscillator which by itself works fine. But when I connect its cathode (output) to the gate of the low-speed oscillator (as shown by the grey line in my pic), it doesn't work. I even tried a whole range of resistance values for the conenction. I also tried changing the values of the resistors which set the voltage at the gate of each PUT. Still didn't work. Any help is appreciated!
 

Attachments

Thread Starter

lowrise4

Joined Sep 6, 2010
34
Hi,
I first verified that the third stage works by itself - by seeing that its output can cause an LED to flash...which it does. I used a smoothing capacitor on the third stage...like on the second stage.

When I conenct the output (cathode) of the third stage to the gate of the second stage, the best/most modulation of the siren sound I can get (depending on which resistors and capacitors I try) is that it will periodically suddenly inject a higher-pitch sound - and often interrupting the siren sound.

I haven't yet come close to being able to smoothly modulate the entire siren sound slowly up and down in pitch...like the second stage does to the first.
 

Brownout

Joined Jan 10, 2012
2,390
I assume from re-reading your first post, that you connected the last stage through a 10k ohm resistor (not shown on your drawing) If that's true, I can't think of a single reason it's not working. I'm sure you double/triple checked you connections by now. If you have a DMM, I sugges that you measure all the DC bias voltages on the last connected stage. And, if you happen to be using an older, anolog dial pointer type, measure the output of that statge while in operation. If the signal is changing slowly, you would be able to "see" the effects on the meter. Please let us know the results of these tests. If I had the resources, I'd try to build the circuit myself. I think this is an interesting project.

BTW, even if the last stage works, I don't think it's going to produce the effect you've described. Just as the 2nd stage effects the frequency of the 1st stage, the last stage will effect only the frequency of the 2nd one. So, the effect should be that the "wobbles" get faster and slower.
 

Thread Starter

lowrise4

Joined Sep 6, 2010
34
BTW, even if the last stage works, I don't think it's going to produce the effect you've described. Just as the 2nd stage effects the frequency of the 1st stage, the last stage will effect only the frequency of the 2nd one. So, the effect should be that the "wobbles" get faster and slower.
Doh!:eek: Why didn't I realize that after all this time? I indeed noticed that I didn't get the kind of pitch modulation I'm looking for by connecting the third stage to the second. It makes sense: the second stage is operating at subaudio frequencies, not audio. So I must instead make the 2nd and 3rd stages both feed into the 1st stage...a bit like an audio mixer.

So by trial & error I finally found two 'programming' resistance values for the voltage divider at the gate of the third stage PUT which worked somewhat: 15k (positive side) and 33k (ground side). And yes...a 10k resistor in between the third and first stage. And I used two 100uF caps on the third stage to make a relatively slow rise & fall of pitch. It doesn't work perfect, but so far this is the closest I can get to the sound effect I want.

A modification I had to do: the rise (produced by the 3rd stage) was much longer than the fall, so I tried adding a parallel diode & resistor before the first cap of the third stage. This did even out the rise and fall times somewhat. But still, the third stage 'messes up' the siren sound a bit...it changes the waver frequency of the second stage...slows down as the pitch rises, and then speeds up as it drops. Maybe the third stage is feeding some current back up into the second?...but a diode between the first and second didn't help.

So it still doesn't give a clean, rising & falling siren sound, but I'm getting closer to what I wanted. Especially considering that I'm using something as simple as the 2N6027. That is my whole point of this project...to see how much (even if crude) sound synthesis I can get out these PUTs. I've found other modifications that make truly bizarre sounds. But for the moment I'm trying to get a nice modified siren sound. Any further help or suggestions is appreciated.
 

Attachments

Brownout

Joined Jan 10, 2012
2,390
Glad to hear of your progress. Have you tried simulating your circuit? You can download LTSpice for free, and try different values of components much more quickly than you can breadboard them. Then when you do make changes on your breadboard, the values will be much close to what you want.

Keep up the good work!
 
Top