DSBSC    demodulation by the squaring means     diode doubler technique

Here is yet another technique for the demodulation of double sideband suppressed carrier signals . This is still  the squaring technique, where by a pair of diodes configured in a frequency doubling configuration are used to generate the missing  carrier that would appear  half way between the upper and lower sidebands .
The technique i tried  is based on  patent No. 2979611   "Synchronous demodulation system  ".  issued to Joseph .W Halina by the United States Patent office .                                    .
My experiment involved a simple connection of components to achieve this .I tested the circuit at 1.8 Mhz as my operational synchronous oscillator was centered around 912 KHz   from a previous experiment .  The (2)  Hybrid was a simple 50 ohm  resistive splitter , (7)  the modulator  or product detector ,was  an SBL-1   connected straight to an amplified computer speaker system .  The (3) full wave rectifier was the standard  twin,  push push diode frequency doubler  . I did not use a clipper amplifier  but relied on the good sensitivity of the (6) synchronous oscillator,  sub harmonically locking to the half frequency output of the  diode frequency doubler .

I constructed the diode frequency doubler  using at first a 1. 84 MHz centre tapped tuned circuit  but for the life of me could not get the output frequency after the diode summing,  to be double the input frequency . so I replaced the tuned circuit with a trifilar wound transformer , and achieved the double frequency output . . I measured the input power level and doubled output power level and obtained the conversion loss . this is seen in the table underneath .  Immediately I can see some sort of active doubler  amplification needs to be used to allow the use of radio receiver  Intermediate frequency 455 Khz power levels to enable good  reliable lock by the synchronous oscillator . It is a toss up as to just place a 20 dB  IF amp stage before the diode doubler or incorporate the active device in the doubling process as is in the  Electronic design sept 1970 "simple RF doubler"  circuit by C  Andren   also in RSGB, Am ,Radio Techniques  Ed 5,  pg 31 .
The diode doubler ,Input   trifilar transformer was  24 turns of fine twisted wire on a half inch core ( found in the junk box)  the diodes where BAT62 schottky diodes   the output transformer was a bifilar winding half inch , no real attempt at input/output  matching was achieved to optimise the doubler performance as I was just interested in testing basic circuit principles .

input  912 Khz                  output  1824 Khz                    isolation at 912Khz
         +3  dBm                                 -13 dBm                          -45 dB
           0                                           -16.4                                 -43
         -3                                            -19.2                                 -50
         -6                                            -24                                    -53
        -10                                           -31                                    -59
        -13                                           -40                                    -67
        -16                                           -52                                       -
        -20                                           -71                       
        -22                                           -78
                                                                                                                -25                                           -84
                                                                                                                -30                                           -96

The main reason I measured the isolation of the fundamental wrt to the output is that I want the synchronous oscillator to lock onto the  processed  frequency of 1824  KHz ( its  subharmonic of 912  KHz )  and not the bleed through of the fundamental  912 KHz     , the 1824 KHz  output is  many db stronger in all cases  so that is not a problem .

The synchronous oscillator was  built up using a BF494  Rf oscillator transistor  biased with a 100K resistor and a BC548  injection transistor  biased with a 10 Meg resistor . the BC548 has a much higher  Hfe than the BF494  so its biasing requirements are much less. I link coupled off the synchronous  osc tank coil with a couple of turns , this produced a 3 dBm RF output level ,more than enough to operate the SBL-1  DBM.

The synchronous oscillator

This oscillator can solidly lock to a  -50  to -40dbm  signal and be pulled many KHz plus and minus  before loosing lock with the injection signal . the locked stability is such that the dsb demodulated audio is clear and undistorted 

I measured the fundamental suppression of the active  " Andren low cost full wave frequency  doubler"  as tried in my  other DSB demod circuit and found the suppression woefully inadequate using unselected transistors( as would be expected )   I measured +3 dB  !!    no doubt this could be improved by placing a 1 k ohm trimmer between the push- push  emitters and the wiper to the collector of the common current transistors , one may have to RF bypass the  emitters with capacitors ???  then maybe one could trim for maximum fundamental suppression .
 I will place a tuned IF amplifier stage ahead of the passive diode doubler and thus make use of its excellent fundamental suppression  and improve the doubled " rf output"  at the same time

I rebuilt the circuit below on double sided  board to investigate the concept


DSB  input to  circuit  from HP8904A       at -28 dBm                                                        spectrum  output of diode frquency doubler        -42 dBm into board     

The push push diode doubler enables the synthesis of the wanted 10 Khz  component of the summation of the upper and lower sidebands   , despite the input levels
the 910 Khz component is not strong  and could be improve by narrow filtering of the diode output and subsequent amplification  . The doubler set up is also susceptible to intermodulation though I didn't test to see if this was a process of the pre doubler amplifier  buckling  under or the  diode doubler itself ,  either way if one overdrives the setup ..... wide band intermod appears   at the output of the doubler  .  see  -28 dBm into board image  The synchronous oscillator has to look at this  "junk" and decide which sidband to lock to !  better to keep the input levels down  (see the -42 dBm  image ) to make life somewhat easier for the locking oscillator .  i noticed when testing the demod board with just the 1 Khz tone  DSB the oscillator could be tuned to lock onto either of the three signals visible on the -42 dBm into board  image .


   spectrum  output of diode frquency doubler      -28 dBm into board                20 sec composite image  comms audio at -42 dbm  into board   note audio notches

I replaced the HP 8904a  dsb source  with a signal generator SRA-1 and  the audio coming  out of my yeasu VX5 on the airband channel  , i set the spectrum analyser to max hold mode on the display and let the system run for 20 seconds   , the comms audio built up the composite image above , note the audio notches either side of the synthesised carrier  between 0 and 300 Hz   ( actually 0 to 600 Hz because of doubling)       

This circuit works very well  i have yet to play with the gains and losses of various parts of the circuit to make the principle easily applied to the output of a 455 Khz transistor  I.F.   I rebuilt up the complete circuit  in double sided  pc board  at 455   and made the circuit above  ,  The loss through the doubler is quite  drive dependant but I found by biasing the two schottky diodes with a meg resistor from +12v   improved the conversion by almost  20dB at such low 455 Khz I.F  levels  . If I was going to rebuild the circuit I would introduce some filtering at 910 Khz with one or two ceramic filters and some post filter amplification before the synchronous oscillator . to improve its locking range as the levels it has to work with in the above circuit provide only a narrow lock range to compensate for drift ..

more to come

   Double sideband demodulator  mk3  squaring loop

Double Sideband suppressed carrier demodulation  ,  MC1496 as squaring means

back to home page                                                                Hit Counter

Double Sideband   suppressed carrier Demodulation   Patents

These are all the References I can find on Double sideband suppressed carrier demodulation , there are bound to be more ,  they are all US patent Office registration No..  The E  ones are European Patent office ,
Have a look on the application of File dates and the Granting dates and some people had to wait a long time before their patents  were granted .

2924706 ,2992326, 2999155, 3047659,  3060383, 3088069,  3108158, 3141066, 3151217, 3252094, 3286183, 3345571, 3457513, 3537017, 3593149,  4253066

 4821322, 4475218, 4642573,  4862098, 2979611,2397961, 4130839, 4430620, 4485487, 4510467, EP0291826B1, EP0293828B1, 2193801, 2276863, 2784311,

3375453 , 3430151,

3719903, a  novel dsb modulator ,