I am building a pitch-only theremin and I designed this, as it was the simplest circuit I could understand lol. I was hoping to get feedback as to if this looks like it will work. The circuit consists of two 555 timer circuits wired in the astable mode. the idea is that the right ("Fixed") one oscillates at about 447kHz and the left one oscillates at between 444kHz and 447kHz, depending on the position of the hand (far away, the two oscillators have equal frequencies so no sound is produced; the 5k pot is to tweak the freqeuncies so that the base frequencies match). I assumed the maximum capacitance of the antenna and hand is around 1 pF. I'm not sure if the 555 can run that fast, or if it was okay to use such a small cap in the frequency determining part of each 555 circuit, but from what I've researched the values do not seem too unreasonable. the 100 uF caps before the XNOR gate are to get rid of DC. the XNOR is supposed to sum the two square waves, with a high output only when both signals are the same voltage (both low or both high). then the final low pass filter to get rid of the 400kHz+ overtone and leave just the beat frequency of the two signals.

So my questions are:

-does this seem like it would work when soldered on a pcb (not breadboard)?
-should I run the output into computer speakers (that have a built in amplifier) or will this power its own speaker?

I apologize if this post is overlong or unintelligible, but I'd appreciate any help I can get!

 

Looks like an interesting circuit. If you can get it to work I might publish it (if that is OK with you). I'm wondering if you can get the variable 555 to actually work like you are planning. Have you tried it yet? I've always thought of 555 as very stable (within the limits of an RC oscillator).

What is the XOR gate (what chip number)?

Your isolation from what I presume is a CMOS gate could use some work. Something like an emitter follower perhaps.

Another scheme using a diode bridge for the XOR gate just occured to me. I don't know if it will work, but it will be interesting to try.

I'll draw it up in a bit for show and tell.

 

Looks like an interesting circuit. If you can get it to work I might publish it (if that is OK with you). I'm wondering if you can get the variable 555 to actually work like you are planning. Have you tried it yet? I've always thought of 555 as very stable (within the limits of an RC oscillator).

What is the XOR gate (what chip number)?

Your isolation from what I presume is a CMOS gate could use some work. Something like an emitter follower perhaps.

thanks, I haven't yet tried it...I'm new to soldering pcb's and I need to pick up a few parts first. I was planning on using an XNOR gate, not an XOR, but it probably wouldn't matter as long as the duty cycle is 50% for the 555. I was just going to use the 4077 series CMOS chip, but I am realizing now that it has super low output lol, so I guess I'd need a transistor, as you said above...I think :S

 

Work through the logic of a simple diode bridge rectifier. It can be a XOR circuit.

I would not assume you can get it to vary frequency that easily, it is an interesting experiment I've never tried.

 

OK, here is a circuit I actually built almost a year ago. It has a lot of similarities to your concept, which is why I'm showing it.

The biggest problem I had was the two oscillators kept wanting to sync up. This is because the power supply didn't have enough filtering. I suspect we're going to have to add a small RC filters to pins 4 & 8 to isolate them. Where the LED/resistor is located put a audio transformer, and I think you will have something very close to what you want.

The diode bridge is a hetrodyne circuit. It creates the sum and the difference from the two inputs. In both cases (LED and audio) we are interested in the difference frequency. I find it interesting that it is also a form of phase detector, as is a XOR gate.

 

The biggest problem I had was the two oscillators kept wanting to sync up. This is because the power supply didn't have enough filtering. I suspect we're going to have to add a small RC filters to pins 4 & 8 to isolate them. Where the LED/resistor is located put a audio transformer, and I think you will have something very close to what you want.

for the filter should I use a low pass filter? if so what frequencies should I be blocking? and for the diode bridge, should I be using schottky diodes? i know their switching times are faster than other diodes. also, do you think the capacitances used for the timer circuits are within the range at which the circuits can operate?

 

I don't think you need to worry about the diodes overall, a simple bridge circuit will work. You can also use 1N914 or 1N4454 types, they are more than fast enough.

Figure a 10Ω resistor between pin 8 and Vcc, with a 220µF cap filtering pin 8 to ground. It may be overkill, but I suspect it is needed.

The more I think about it the less I think the variable frequency 555 side will work. However, since I don't really know it is worth the experiment. I also think you are going to have a few problems reaching 440Khz. While in theory it is within reach of the 555 in practice it may not work so well. We shall see, this kind of experiment is all part of the fun!

 

will the sum frequency be filtered out by the diode bridge? and what's the 150Ω resistor for? just a load so the diode/speaker doesn't draw too much power? sorry for all the questions! I appreciate the help with this.

 

The resistor is to limit current to the LED. LEDs always require something to limit current, always. Without it the 555s or LED or bridge would burn out.

Use the values you calculated for the 555s, they looked good to me. The above circuit is only a reference.

The sum and difference will be created by the diode bridge, as it heterodynes the two frequencies together.

 

The sum and difference will be created by the diode bridge, as it heterodynes the two frequencies together.

I understand that the diode bridge returns the sum and difference of the input signals. So then won't both the difference, the frequency I want driving the speaker, and the sum, the one I don't, be fed to the speaker? in other words, shouldn't I add a cap to ground so that the high frequency is attenuated? also, between the output of the 555's and the input of the bridge should there be a cap in place to block the dc component of the output signal?

 

Why bother? If it is 440Khz then the sum will be 880Khz. You may be younger than me, but I doubt your hearing goes that high.

 

haha you have a point. okay no filter after the bridge, but what about the caps between the 555 outputs and the diode bridge? And to construct the circuit I was going to use a pcb from Radioshack that has the same layout as a breadboard

I know breadboards are horrible with small capacitances because the individual connections have significant capacitances, but do you think the pcb version without clips will still cause this problem?

 

I didn't see any mention of thermally-induced drift. I think that will be a big problem, IF you can keep them from syncing up due to injection locking.

 

I would use a conventional protoboard to make the first one. Ron is right, there may still be issues with the design. You are at the build it and see phase.