Re: [Repeater-Builder] Re: Audio Reproduction

9 Apr 2004 15:51:18 -0000

Coy Hilton wrote:
First I would like to read each of your personal definitions of FM and PM, then your dissertations on exactly why one is better than the other, Complete.

I apologize,  this is long.


Definitions:
FM:  Modulation of a carrier that its instantaneous frequency differs from its resting (no audio signal) frequency by an amount proportionate to the instantaneous amplitude of the modulating waveform.

PM:  Modulation in which the angle of the carrier frequency differs from the resting (no audio signal) angle by an amount proportional to the instantaneous amplitude of the modulating waveform.

Here is my dissertation on the two, based on commonly available radio equipment used to build repeaters from.  While it is necessary to present the theory, I'll go into the more practical sense and let the theory to S-COM Bob and others:

It is true that FM and PM (in theory) are equally linear, in practice FM is much easier to keep linear.  It is also true there is nothing inherently wrong with Pre-De emphasis, however they are rarely reciprocal to the strictest degree.  The shortcoming is not because of filter slope, but rather where the filters start and stop (called the corners).  In FM, we commonly clip/amplitude limit the audio applied to the modulator, this is called 'processing'.  It was found that adding pre-emphasis to the audio before its application to the processing allowed for a more balanced level of clipping since the human (and especially male) voice spectrum emits more lows than highs.  Most modern PM two-way radios pre-emphasize audio applied to the processor, and then low pass filter it and de-emphasize it before applying it to the modulator.  In this case the de-emphasis stage was needed because the PM modulator automatically pre-emphasizes the applied modulation as a function of the modulating scheme. 

In a repeater, audio recovered from the discriminator is pre-emphasized, pre-emphasized by the users radio. 

In repeaters that use a PM modulator and pre and de-emphasis stages, the discriminator audio is de-emphasized, likely run through a controller, pre-emphasized prior to the processing, processed, low pass filtered, and de-emphasized again, then pre-emphasized again by the function of the modulator.  Wow, hardly a chance there for things to go wrong huh?  Even if the emphasis stages were 100% reciprocal, cascading effects will narrow the bandwidth of the audio as the corners will be pulled in from both sides.  This is why links often sound worse the further down the chain.  To add to the problem, often times the processing "blob" will not only provide low pass filtering to remove harmonics created by the clipping, but roll off everything below 300 cycles for two reasons.  First, the reduction of  frequencies below 300 cycles is important because their harmonics will fall *into* the bandwidth of the audio being repeated making the audio raspy and distorted.  Second, it was found that reduction of these lower frequencies allowed the PL decoders to operate better without being "talked off". 

In a repeater built under the rule of "Flat Audio Modification" where a FM modulator is used, audio is taken from the discriminator, run through the controller, applied to the processing, low pass filtered and applied to the modulator.  Since the emphasis stages were eliminated, the chance of their inconstancies are eliminated.  On the ham bands, some folks go one step further and eliminate the processing and low pass filter.  While this results in the best retransmitted audio, it likely also results in occupying too much bandwidth for a given channel.  Audio should be at minimum low pass filtered.  Since in many cases the processing is eliminated thus eliminating LPF'ing, CTCSS or PL tones can be passed right through the repeater.  Obviously care of building and set-up of the 'flat audio' repeater is an absolute necessity, otherwise you'll be clobbering your neighbor, especially on a 15 kc. 2 meter pair.

In conclusion,  repeaters can be made to sound better no matter the modulation method, and one isn't likely better than the other.  It's simply easier to make an FM exciter sound better because most commonly available FM exciter are already more linear than their PM counterpart and using one allows the system to be built without the emphasis stages thus eliminating their possible inconstancies.

It should also be noted that with crystal based radios, as frequencies increase, usually so does the multiplication factor.  The limitations of present common PM modulators become less and less as the multiplication factor increases.  A GE Mastr II UHF radio with PM modulation will sound nearly as good as its sister with true FM ICOM.  Why?  Because the operating frequency is the result of the oscillator frequency being multiplied 36 times.   In this case the crystal frequency only needs swung + and - 139 cycles to achieve the same 5 kc of deviation on the UHF channel.  Low Band and High Band radios really suffer with PM modulation methods found in common radio equipment we convert to repeaters, and for that reason FM becomes more attractive on these bands.

Scott and I recently acquired the processing units that were designed by Joe Montierth, K7JEM.  These were designed to replace stock processing modules "blobs" that are partly the limiting factor for high quality audio reproduction.  Scott and I will be building these units and evaluating them after implementation in GE and Motorola exciters.  These units *may* need to be refined before they are made available to the public.  After we are satisfied with its performance, the product will be made available from Repeater Builder.  Oh yes,  they will work with either an FM or PM exciter  :)

Kevin Custer

























Yahoo! Groups Links

Reply via email to