8 channel PCM receiver

[brianpll]

Hi ! guys im trying to build 8 channel PCM receiver using PLL as a FSK demodulator .The PLL is 74HC4046 it works upto 15Mhz (I hope ).Im thinking of using a mixer to obtain an IF of 10.7 Mhz this IF is filtered with 10.7M SAW filter amplified with IF amp and then directly connected to the input of PLL to obtain PCM data .This data is to be subsequently decoded by a PIC (Uc) to obtain PWM data for each servo . Here no 445 khz IF or FM demod IC is to be used . Think itll work ? . Please let me know what you guys think . I could also use some help designing the RF mixer stage . Any bright ideas ?

[Mr.RC-CAM]

This might give you some good ideas: http://home.nordnet.fr/~fthobois/anglais/introrx19.htm

And, you might want to collaborate projects with this eZoner:
http://www.rcgroups.com/forums/showthread.php?t=287775

Lastly, there is no single standard of PCM signalling. Each Mfg uses their own proprietary format(s) with bit frame lengths that can span from 512 to 2048 bits. None of the formats are openly published. So, it might be a stretch to go beyond an intelligent (DSP) PPM method.

RC-CAM

[chaitanya]

AFAIK the PLL output is a square wave whereas that required by the mixer is sine wave

[brianpll]

Yes Chaitanya what you say is correct the PLL output is square wave however there seems to be a little misunderstanding ,the output of the mixer is connected to the input of the PLL and not the other way around .There are only two inputs to the mixer 1) RF signal out from rf amp 2) Local oscillator input from LO both of which are of sinewave nature


Note:**In this case PLL is not used for frequency synthesis.
It is used for FSK demodulation

I hope this clarifies your doubt . Any other suggestions from you are welcome .

[chaitanya]

oh i thought you used it for frequency synthesis...
this might give you an general idea about Futaba PCM coding,
http://www.mh.ttu.ee/risto/rc/electronics/index.htm
-chaitanya

[brianpll]

Found a really simple TX circuit to go with the 8 CH receiver
There are only three inputs to this circuit 1) Vcc 2) Gnd 3) Mod input .
The serial output of the Uc Encoder (in PCM format )can be directly connected to the Mod input of rf deck (may reqiuire a buffer amp ) . The format of PCM used is only dependent on the program code within the Uc encoder and decoder . Hence the same circuit(RX&TX) may be used for various PCM formats by making provision for In circuit prog (ICP) of the Uc or by storing the code in external flash ROM .
The PCM format I have decided to use was published in
elektor mag in PCM radio control project .Uc code was also provided however that code was intended for phillips Uc.Since I will be using PIC I will have to start from scratch .

[Cirip]

Hi brianpll,

I have a bit of exposure to rf so if you don't mind I'd like to add my comments here.
The idea of using the PLL for FSK demod is interesting, however I see a few issues here. Trying to demodulate a narrow band FM signal at 10.7MHz might not be a good idea. You may want to perform another down conversion. The relative frequency deviation is smaller at 10.7MHz than at say 455KHz. This will result in a low signal level at the demod output and fairly poor signal to noise ratio.

Also, I am not sure if the 10.7MHz IF fiters are narrow enough to protect the receiver against adjacent channel interference. This might be another reason why you might need a secondary 455KHz filter. High IF frequency offers good image rejection, but behave poorly with regard to the adjacent channels because of the relative difficulty of obtaining narrow passing bandwith at high frequency.

And by the way, a mixer does not mind square wave. Even supplied with a sine wave, driven into saturation, the mixer will generate harmonics anyways.

Just my 2 cents,
Cirip

[Mr DIY]

brianpll

Looking at post 1, the mixer bandwidth is far less of a problem (if at all) than getting the local osc. to oscillate over numerous bands.

The saw filter .. why?
Saw filters get big as the frequency comes down and generally even bigger as the bandwidth gets smaller (which you need). Have you found a suitable device that is not as big as a miniature receiver?

The PLL discriminator.
I would be far easier to stick with a standard discriminator than to go with a PLL. Better would be to use a suitable IF amp chip that has a suitable discriminator built in. You just adding complexity for little or no gain.

Lastly, there is no single standard of PCM signalling. Each Mfg uses their own proprietary format(s) with bit frame lengths that can span from 512 to 2048 bits. None of the formats are openly published. So, it might be a stretch to go beyond an intelligent (DSP) PPM method.

Yes …use PPM instead. You could try and decode the data stream for PCM if you have the right equipment at your disposal, but that may take a while.

Brian