RCA invented ordinary 4xxx series Cmos logic. They show a linear amplifier made from a CD4069 inverter (an MM74C04 is the same). The current is very low especially with a 3V supply so a 10M feedback resistor is used. The graph here from RCA shows a gain of nearly 500 times with a frequency response to only 800Hz when it has a 1k input resistor.
But the author of the Bat Detector used 74HC04 high speed high current Cmos inverters and did not need a gain of 500 times and he wanted a wider frequency response so he used 33k feedback resistors. (My opinion).
Hi AudioGuru, I just sent him an Email.
Now waiting for his response.
Well with the 9 Volt Supply, I should not only be getting the Wider Bandwidth, But also the Higher Gain.
It definately Is Not creating any Excessive Noise, So Extra Gain is not a problem.
Chemelec, that's a kind offer and I appreciate it, but I'm enjoying designing my own. So thank you for the offer but I'm quite happy as I am.
You asked about my mic and I've had another look at the Knowles website. Found an application note on ultrasonic application of their microphones which gives some really useful information.
**broken link removed**
This gives the frequency response for some of their MEMS and electret condenser mics out to 100kHz.
I'm going to order the FG23629 electret mic. It will be easier to actually mount to something, although it's still very small. Plus the MEMS frequency response fluctuates a bit whereas the electret mic just ramps down in a way that can easily be compensated for with amplification.
I can't help feeling this mixer is getting to be way too sophisticated for such a simple task!...
I built a heterodyne bat detector about 10 years ago, and just used an 'off the shelf ' bog standard 4052 CMOS multiplexer as a switching mixer. Because the LO for a switching mixer is a square wave, only the 'signal' input - and its inverse - need to be linear. Much simpler !
I finally got round to trying your method for myself. Thanks for the schematic you attached. I built up a similar circuit and I agree with you that it's a cheap and simple. However the output still contains the spurious signals I referred to previously! The signals are the result of harmonics of both the LO and the signal mixing together, producing audible diff frequencies. In this instance, since the LO is just driving a switch on/off I'm not sure how the harmonics are getting mixed up in the output.
Having just looked at the maths in the link you provided it does show that the square wave harmonics get through. So it has to be concluded that there's nothing that can be done about these spurious signals. Without a sine LO that remains sinusoidal throughout the variable frequency range, harmonics will be present. In fact I tried using a signal generator to produce the sine LO and the harmonics could still be heard, albeit very quiet.
So accepting this, I'll stick with an op amp relaxation oscillator for simplicity. I'll also stick with the NE612 mixer since I already have the ICs and designs for it.