How do I distort thee, let me count the ways PART II

Continued from here (http://www.thereminworld.com/forum.asp?cmd=p&T=3906&F=1)

One thing to be aware of in the Schematic of 'Glasgow Digital Theremin' (http://www.physics.gla.ac.uk/~kskeldon/PubSci/exhibits/E9/cir2.gif)
is that the 4093 gates shown, are, in fact, Shmidt NAND gates and not Schmidt inverters - this means that there are pins which are not shown.. ALL unused pins should be connected to +Ve, as otherwise they float and can cause all sorts of problems!

One could probably use a single 40106 hex Scmidt inverter IC, and need only one IC instead of the 3 required in this schematic... I expect it was done this way to isolate sections of the circuit - but I doubt if this 'isolation' is required - if one decouples the IC well (low ESR capacitor across its supply pins, CLOSE to the IC).. I have made similar oscillators using a single 40106 with no problems.

It is probably worth checking other pins in the schematic.. for example, the output pin of the 4098 is not shown.. its Q output (for the selected monostable) is pin 10, its !Q output is pin 9.. Either of these could be used - but if Q is selected, the integrated voltage (A) will rise as frequency increases, and if !Q is selected,the integrated voltage (A) will fall as frequency increases.. It is also probably a good idea to tie unused inputs on the unused monostable to something - pins 3,4+5 all to +Ve.

Fred Mundell
Fundamental Designs Ltd.
Electronics Consultant.
<- See Profile Image for Email.
Designer of Theremins and other alternative electronic music controllers and instruments.

Good suggestions, Fred! CMOS gates become "linear amplifiers" of sorts if you bias their inputs at half-rail. This used to be an old hacker's trick many years ago ... need an amp? have an unused CMOS gate available? Bias it into the linear range and voila ... a "free" amp.

It wasn't a great idea then or now, but people did it. Back to the point ... if you don't want to blast a CMOS gate into oblivion, make sure you terminate the unused inputs. Tying them to GND or Vdd is fine. You don't need to do anything with unused outputs.

Don

One more CMOS comment ...

For all the concentrated effort for ESD control in industry, I have really only blown up 2 devices:
1. A blue LED
2. An A-series (ex CD4011A vs CD4011B) CMOS gate

But it always best to err on the side of caution, so make sure you are grounded to something (safest is one of those ESD wriststraps) when handling any active (transistors, FETs, amplifiers, logic gates, etc.) components. You'd have to be very aggressive to blow up passive components (resistors, capacitors, inductors) with ESD, so you don't have to worry much about handling them.

Don

"I have really only blown up 2 devices" - Don, you have done well! I have managed a few more than that.. and then there are the real problem components - the ones which do not 'blow up' but which are stressed, and fail prematurely.. These are the ones that REALLY cost!

Anyone doing much work with FET's of any kind (transistors, CMOS logic etc) needs to be aware that even if the circuit works, if ESD precautions (construct board on grounded anti-static mat, wear rounded ESD wristband - the basic kit costs about £15) are not taken one is storing up trouble.

With regard to connecting spare pins on logic parts.. Unused AND inputs must be taken High, Unused OR inputs must be taken Low, Other gates - look at the function and tie to whatever gives the output you want.. For higher frequency operation, do not be tempted to tie 2 inputs together (for example, 2 inputs of a NAND or NOR gate to make an inverter) as this increases the capacitance seen on the input - better to tie unused NAND inputs High, and unused NOR inputs Low.

Hi guys!

My friend has the following problem: the frecuency of the output of his antenna's oscillator varies between 1.19MHz and 1.1MHz approximately, being the mixer's output a signal between 10KHz and 100KHz roughly, but he needs a mixer's output between 10KHz and 200KHz. For this reason, the antenna's oscillator has to produce a signal of which frecuency varies between 1.19MHz and 1MHz approximately.

How could he resolve this problem without change the antenna?

Modify the oscillators tank circuit: take a smaller capacitance and a bigger inductance, so that the free run frequency will be the same.

Is this the "Glasgow" Theremin again? If it is, he might need a series trimming capacitor to the antenna, and a trimming capacitor from the original antenna connection point to ground..

BUT - If it is the 'Glasgow' Theremin, one should be able to compensate without needing to mess with the oscillator at all - The opamp buffer has facilities to trim gain and offset, and the monostable pulse width can be varied, and the integrator RC can be varied - Why pick the most awkward place possible to trim this circuit?

If it is not the 'Glasgow' Theremin, what Theremin is it? "he needs a mixer's output between 10KHz and 200KHz" is no normal Theremin requirement! .. My first suggestion would be to adjust the pulse width from the monostable, so that it is close to 10us (near fully on at 100kHz) and adjust the gain /offset on the opamp to give the required CV over the 10kHz-100kHz range.

I have quickly pasted some ROUGH simulations showing the results of heterodyning various waveforms (Ramps and various pulses etc) together -

They can be viewed here. (http://www.therasynth.com/assets/images/autogen/a_Heterodyne_Waveforms.jpg)

My friend wants to know what are the external potentiometers in thh glasgow theremin, the potentiometers which have to be controlled by the user.