The new WB7FID repeater:

"It is our belief that the aural subcarrier has no business being modulated in the video exciter... at least on a repeater..."
 

Aural Carrier transmitter controller/reference board 

Adding the aural subcarrier to the baseband video causes several problems: Any nonlinearities result in additional intermodulation products, possibly in the infamous 'triple-beat.' If triple-beat isn't enough, then sync compression can result in vertical sync buzz on the audio. Perhaps the worst effect of modulating the audio with the video at a repeater is the myriad of low-level products scattered all around the spectrum, modulated with the audio program, probably getting into every nearby receiver. A plus of having a separate transmitter is that there isn't the VSB image audio carrier to contend with...
 

Aural Carrier transmitter VCO assembly

This transmitter is capable of over 25 watts output power (adjustable) to allow one to vary the aural/visual power carrier ratios. In our installation, the aural carrier will be combined with the visual carrier with a directional coupler (See the Xmit Filters and Combiners page for more info.) This coupler offers less than 1db of insertion loss to the visual carrier but about 9db to the Aural carrier. Keep in mind that "FM Power" is cheap compared to linear AM power so enough power is pushed to provide the proper power ratios.
 

Internal view of aural transmitter

This transmitter generates its carrier on-frequency using an MCL POS-535 VCO module. The output of the VCO is amplified by a MAV-11 and then by an MRF-559 to several hundred milliwatts and applied to a Mitsubishi M57729 power module.

A dual-modulus divide-by-N counter is used to divide the carrier frequency down to the 62.5 KHz reference frequency. This is compared to the reference derived from a 5 MHz crystal oscillator.
 
 

The power amplifier module assembly

The output of the phase/frequency comparator is applied to a loop filter/integrator. The audio is bandpass filtered for a 60 Hz to 13 KHz bandpass and combined with the output of the integrator. It is important that the lowest modulated frequency is above the corner frequency of the PLL's loop filter:  Failure to do this will result in modulation causing loop instabilities. Because the VCO is only being frequency-controlled (that is, the loop filter wasn't being used to eliminate phase noise, etc.) extraordinary measures had to be taken to keep extraneous noise from the VCO: The VCO has it's own, private voltage regulator. It is also mounted in its own die-cast enclose which is insulated mechanically from the main enclosure with foam, to prevent microphonics.  A note:  The noise produced by the VCO's own 7812 regulator resulted in very audible noise on the received signal, so additional active (e.g. resistor, capacitor, and pass transistor) as well as passive (large-value capacitor) filtering needed to be added to eliminate this problem.

The controller provides an interlock such that if the transmitter's frequency control loop is unlocked the transmitter will not key (or will unkey if it unlocks during transmission.) Because of the long time-constant of the loop filter/VCO chain, it can take as long as 15 seconds for the PLL to lock up. To key the transmitter during this lock time could cause a signal to appear who-knows-where... Because of the long initial lock time, power is applied continuously so that the synthesizer remains locked at all times, ready to be keyed on a moment's notice.  This circuit was complicated a bit by the fact that when the amplifier is enabled, the PLL goes out of lock momentarily (due to 8-12 amps suddenly flowing around the enclosure causing a 200-500 Hz shift in frequency.)  Additional logic had to be added so that the "unlock" condition would be ignored for a short time after key-up, or else the transmitter would continuously key, unlock, unkey, then relock, key, unlock, unkey...

If I were to build this transmitter again, there are a few things that I would do differently:

• I would not use a POS-535 VCO.  There is nothing wrong with this VCO module in general, but in this situation - where it is being used on a single frequency to generate a (relatively) narrowband FM signal - it is not the best choice.  Having an oscillator with a tuning range from 300-something to about 500 MHz means that the amount of audio imposed on the VCO tuning line is microscopic - only a few microvolts of noise can result in audible degradation.  A better choice would have been to use a homebrew VCO with a much smaller tuning range (i.e. a tuning range 2-3 times that of the expected temperature and aging drift.)
• A simple microcontroller to do the loop control/transmit keying logic.  The existing design uses a number of R/C timers and latches to prevent "illegal" states (i.e. transmitting when the PLL is unlocked, etc.)  While this circuit works well, it would have been nice to be able to get rid of a few ICs and gain the added flexibility of having the behavior of the system determined by the program.  At the time I built this transmitter, I just didn't have the resources to drop a PIC into it - or I would have.

Keep watching this page, as it will be updated as time goes on...

Do you have any questions/comments about what you have just read? If so, please email me and make an ask of yourself...

Return to the New WB7FID ATV Repeater Page...

Updated 991213