Giantkiller moves forward into Full Heterodyning.

[MACEDONIA CD]

can enyone  ansfer  ho is the  frekfency of   noise  of the simple radio recivers <<    _,5 Khz>>
TUNE   CLOUSE  TO  EXSACTLY FREK ..  AND MAKE  SELF DISTROING TPU

NOISE  HAS HMM  0,01 db  constant energy
exsactly raid tune frekfency has   hmm  1db energy 

if tune clouse  and clouse  to  1db 

that is the reasons  way <<sm> dont tune exsactly raid  frek     BEACOUSE  THE  NOISE  FREKFENCY  HAS  STABE    AND CONSTANT  ENRGY
IF  THE S.M  TUNE CLOUSE  TO EXSACTLY  RAID FREK   <<THIS FREK   IS NOT STABLE  AND NOT CONSTANT 
IF  THE S.M  TUNE LIKE THAT     THEN  HIS TPU   WHILL HAVE   VARIABLE OUT  VOLTAGE  AND COURENTS 
THAT IS  S.M  TUNE  THIS   NOISE FRE. FOR  TO HAVE STABLE  V  AND  A
IF YOU LISTEND RADIO  YOU HAVE VARIABLE  db  IN ENERGY  TO YOU RECIVER
IF YOU PUT TO  NOT IN THE STATION  YOU CAN SEE  THE  db is constant

[Feynman]

Success!!

I got the LTC6904 to scan from 1khz up to 4.8 Mhz.  There were some software glitches for the past few hours, but they were resolves when I used 5V common power supply for the I2C bus (rather than two seperate power supplies, one running off 2AA batteries at 3V and the other running off the USB cable).

However, my binary calculations are off!  Oops!  I think I should drink more ale.  Because I was only able to scan up to 4Mhz (instead of full 67Mhz) suggests my math is wrong.  I will post a big post if I can fix this tonight.  Maybe I will go to work late.    hahahaha

<3<3<#
Feynman

EDIT:  Okay so I fixed the math stuff... I've gotten up to 47Mhz stable.  Let me tell you at this frequency the wave starts turning *ugly*!  But the square wave is very beautiful to about 20Mhz. Also when I probe with the scope I can hear the tones rising through my computer speaks up to around 23khz.

[Feynman]

I felt like a should clarify what's going on a bit , so here goes:


I am producing square waves using a small LTC6904 chip.  This chip can go 1khz - 67Mhz.  It is controlled over the I2C bus from an Arduino microcontroller.  The Arduino is controlled from the computer via Serial over USB. 

What was accomplished:
Last night I programmed the Arduino to make the LTC6904 "scan" across frequencies from 1khz to 67Mhz. Every 10 seconds the Arduino would send a new command to the LTC6904 oscillator to move up to a higher frequency band.  There were a total of 32 separate frequencies tested across the 1khz to 67Mhz interval.   Additionally, the Arduino was programmed to report its status over the serial port, so I was able to see in a Linux terminal whenever it changed the oscillator's frequency.

I also watched the output on my frequency counter and scope, and I could see the frequency jump up higher every 10 seconds.  I also did a rough test of close we were to the expected frequency output and this generally was in +- 2% in kilohertz and got much larger (+- 20%) when we exceeded 2Mhz.  The large error was due , in part, to a mistake on my part.  You cannot change both the octave and the frequency at the same time on the LTC6904.  The LTC spec says you need to send two separate commands, but I was doing this in one command which I think was affecting the oscillator stability.

So in summary , the LTC6904 was programmed to scan a large frequency range.  This test was succesful. In fact it was funny because under 22khz I could hear the tones through interference with my speakers.  There were a couple of minor problems with error in higher megahertz, but I think this can be resolved by following the LTC6904 spec sheet. Future plans include testing the accuracy and precision with hopes of improving it and making consistent output.

[zerotensor]

...
For example in classic TTL these tiny currents will make outputs go high....
So it is best to add aditional circuitry which pulls the open connections to absolute ground.
Like a circuit on top of a circuit which simply grounds "unconnected" connections.
...

Back in college when I was building NMR spectrometers, I added a number of tiny capacitors connected to ground on circuit elements near the ic's;  I'm not sure exactly how or why it worked, but this technique somehow helped with stability in the 90-120 MHz range.

[giantkiller]

Just a side note:
The SM17 has inductors and caps placed similar to the LMD/TEM analog computer.
Just rolled up.

This gives a mobius configuration where the coil loops are in parallel and the caps are in parallel. Each pair communicates with itself. Instead of using multiple stages the effect is fedback at the end of each pulse which, by the way, is fired in 2 directions at once. One direction per one coil.

--giantkiller. Totally tanked, horizontally and vertically.