FONER%MIT-OZ@MIT-MC.ARPA (Leonard N. Foner) (09/05/85)
If you're not a hardware person, you might as well stop reading right now. This is being sent to the above strange combination of lists because I have no access to UUCP's net.electronics. If some kind soul could forward this message there as well (once only!), please do. I'm going to be driving a large hunk of coax with taps off it with FM multiplex stereo. The cable already has normal video RF on it; I intend to modulate in the normal broadcast FM band, between channels 6 and 7 (88 to 108 MHz). The idea is to have a private FM stereo network that anyone with an FM receiver can listen to off the cable (analogous to cable TV, which is already on the coax). Unfortunately, I'm not an RF engineer, and my interest in *analog* high speed circuits was a) as a hobby, and b) quite some years ago. Hence, I'm very out-of-date on how to drive such a beastie. There *must* be a better way than the many-discrete-components ideas I can generate. What I want are references to schematics which would show how to do this. The lower the cost per transmitter, the better. What with stereo FM on videotapes now, it's possible that someone's made a chip to do the modulation, now that consumers might be generating FM. (I *know* the demodulators are a dime a dozen. This is different.) If there does not exist a chip around somewhere, I'd appreciate pointers to recent articles on building cheap stereo FM transmitters. The ideal transmitter would be one whose transmit frequency can be programmed with a DIP switch (i.e., a synthesized transmitter that divides down the carrier and PLL locks it to a crystal). As a part of that last piece, I'd love to hear about divide-by-n chips that are programmable from 1 to 256 via some eight-bit input, and can divide down an input frequency of 110 MHz or lower. If no such exist, I can take a slower part and put a fast flipflop in front of it, hence making even 50 or 25 MHz parts usable but inferior to one that's really fast enough. (This may not be a big problem, though, since 64 different frequencies is plenty given recommended transmitter spacing and the width of the band.) This *must* be a solved problem. I'd hate to reinvent the wheel, whatever I do. Can anyone point me to the relevant literature? Please reply directly to me, since I'm not on all of the above lists, and hold CC'ing the list unless you think it's of general interest. Thanx much! <LNF>
hpk@vax135.UUCP (Howard Katseff) (09/14/85)
I happen to have the address for Panaxis Productions. It is PO Box 130, Paradise, CA 95969. The address is from the back of the FM Atlas catalog, which is a good source for SCA adapters. Their address is FM Atlas, Adolph, MN 55701-0024.
bill@videovax.UUCP (William K. McFadden) (09/17/85)
I tried to send this via mail but it failed. My apologies to the net for
any inconvenience.
--------------------
I have had some experience designing cheap FM stereo
transmitters. (I designed and built one as a college senior project.)
Your project sounds interesting, so I hope some of this will be useful.
First, the technique used to modulate FM stereo onto videotape is
different than the one used in broadcasting. The Beta/VHS Hi-fi systems
use two separate FM carriers, one for each channel. On the other hand,
FM broadcasting modulates the baseband with mono (L+R) and a 38 KHz
subcarrier with L-R information. The subcarrier is AM double sideband,
suppresed carrier (DSB-SC), and a 19 KHz pilot tone is needed to
properly demodulate this signal (also provides a way for your receiver
to tell if an incoming signal is in stereo). To regenerate the L and R
channels, they are mixed in the following way.
(L+R) + (L-R) = 2L
(L+R) - (L-R) = 2R
Thus, a simple matrix will decode the signals. Mono receivers demodulate
only the baseband signal (L+R) and ignore the subcarriers. Thus, the FM stereo
signal is compatible with both types of receivers. (You may already know
all of this, but I thought I'd include it for completeness.)
You are right that PLL is the way to go. As part of my project I looked at
several ways to generate a stable carrier and digital frequency synthesis
was by far the best and cheapest. My design uses a transistor oscillator
with varactor diode tuning to generate the FM carrier. This is divided
down to ~1 MHz with digital counters and fed into a CD4059 modulo-N counter.
The CD4059 can divide by any number between 3 and 9999 inclusive and is
programmed with with a 16-bit input consisting of four BCD digits. This
made it easy to use dip switches or thumbwheel switches for tuning.
I am not aware of any versastile divide-by-N counters that operate up to
100 MHz, but there are high speed decade counters that run this fast.
Therefore it was easy to divide the carrier frequency to the slower rate
needed by the 4059.
The phase comparator was a CD4046 PLL chip with the internal VCO disabled
(it only goes to 1 MHz). The frequency out of the 4059 was 1 KHz, thus
a 1 MHz crystal oscillator followed by a divide-by-1000 counter (MC14559)
generated the reference frequency.
I should stop here before I get carried away. If you would like to know
more about my design (e.g., schematics), let me know.
There are a couple of companies that offer plans, boards, and kits for
FM stereo transmitters, some of which use PLL synthesis and dipswitch
programming. I have included their addresses below.
Panaxis Productions
P.O. Box 130
Paradise, CA 95969
(916) 534-0417
Contact: Ernie Wilson
Stellatron (This company may not still be in business)
4942 Whitsett Av. #205
N. Hollywood, CA 91607
(213) 506-0415
Contact: Wayne Slater
Bill McFadden
Tektronix, Inc.
P.O. Box 500
MS 58-594
Beaverton, OR 97077
(503) 627-6920
Net address:
...{ucbvax,ihnp4,uw-beaver,decvax}!tektronix!bill@videovax.UUCP