Re: Precision low frequency generation

On Tue, 20 Nov 2007 07:33:19 -0800 (PST), Dirk Bruere at NeoPax
<dirk.bruere@xxxxxxxxx> wrote:

On 20 Nov, 15:28, Jan Panteltje <pNaonStpealm...@xxxxxxxxx> wrote:
On a sunny day (Tue, 20 Nov 2007 07:25:00 -0800 (PST)) it happened
pantel...@xxxxxxxxx wrote in
<2cccb7f8-56d7-4b7e-903f-350c7b819...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>:

Google screwed up my ASCII art, so here again from a real newsreader;

Dirk, to get the difference frequency you need to multiply, not add, the signals.

Here is a cheap man's solution:

0V --potentiometer-- +12V +12V
| |
100kOhm 2200 Ohm
C 1 | | C4
f1---- 100nF----------------g2 drain---------------------------- 100uF---------------- out
BF961 |
C 2 dual gate MOSFET 1uF
f2---- 100nF----------------g1 source-- | C3
| | ///
100kOhm |
| ///
///

C1 and C2 decouple the input.
The potentiometer sets the voltage at gate2 so the drain is at about 7V.
The 2200 Ohm together with C3 form a lowpass.
The second gate in the BF961 dual gate MOSFET modulates the signal caused by the voltage on the first gate.
Here is a similar diagram, note the mixer, it is for higher input frequencies however..
http://panteltje.com/panteltje/lorentz/experiment.ps

Select C3 for adequate low pass filtering.

Does output need to be a precise sinewave?
If so, use an EPROM with a binary counter on the address lines, driven from a 74HC4046 VCO, output
EPROM into a simple DAC.

Thanks.
Initially I need to test with sinewaves, but ultimately I want to use
the computer to generate arbitrary LF waveforms.
Modulating a HF signal with the LF waveform in the computer would seem
the best way to go initially. Then it becomes an AM 'radio' problem.
Cat's whisker?

Dirk

You could generate an arbitrary waveform at, say, 5 KHz, and sample it
at, say, 5001 Hz. The sampling would downconvert the 5K signal to 1 Hz
and preserve the waveform. The circuit needed is a simple
sample-and-hold. It could be done with a resistive mixer and a diode
peak detector, too, 4 passives, but it wouldn't be as elegant.

I bet you could booger a sound card to give a dc-coupled output. It
does have a dac inside somewhere. Or buy a USB dac.

John







.

Relevant Pages

  • Re: Comparing two similiar audio files, FFT?
    ... The waveform will consistently ... frequencies in play are below 5k), the MSE seems to love to approach 0 ... different point in the cycle than on the other (one starts at about ... time signals: ...
    (comp.dsp)
  • Re: NICAM vs MP2 - the final showdown
    ... The waveform can only be "reconstructed" using software which performs ... actual signals is just an approximation. ... no more than troll Steve, ...
    (alt.radio.digital)
  • Re: NICAM vs MP2 - the final showdown
    ... The waveform can only be "reconstructed" using software which performs ... actual signals is just an approximation. ... no more than troll Steve, ... Sorry to disappoint you. ...
    (alt.radio.digital)
  • Re: IM distortion: why related to level?
    ... waveform which is at precisely 1/2 the sample rate, ... The Nyquist criteria is not an "if and only if." ... signals. ...
    (rec.audio.tech)
  • Re: Precision low frequency generation
    ... Dirk, to get the difference frequency you need to multiply, not add, the signals. ... The second gate in the BF961 dual gate MOSFET modulates the signal caused by the voltage on the first gate. ... the computer to generate arbitrary LF waveforms. ...
    (sci.electronics.design)
Quantcast