What effect does the Processor have in FM and AM?

KD0RC

From the SmartSDR Software User Guide page 188:

Processor (PROC) button: Clicking on this button will enable the W9GR Controlled Envelope Single Sideband (CESSB) DSP algorithm resulting in additional talk power. The speech processor may be on or off and has three different settings when on. In the NOR or normal setting, the processor provides minimal additional gain and simply prevents audio peaks from clipping or producing power greater than the set level. In the DX setting, some compression is provided to the audio to increase the overall sideband envelope which results in a stronger signal that may be more readily heard at a distance. The DX+ setting adds even more compression increasing your talk power or “punch” without incurring significant audio distortion. DX+ is most effective if you increase the low-cut TX filter to between 200-400 Hz to concentrate your talk power in the audio frequency range that has the greatest intelligibility. This feature is available only in SSB, AM and FM modes.

The PROC button enables CESSB, so I understand this when in SSB mode (the W9GR video was excellent at explaining what is really happening). In FM and AM, however, I find it hard to believe that CESSB is part of the equation. So I wonder if the PROC actually does anything in FM and AM (if so, what does it really do?).

Mike-VA3MW

CESSB is for SSB Voice only.

Ken Wells

Hello,

I just ran some tests. -- The PROC is indeed active on AM & FM.

If the TX Hi/Lo-Cut and TX EQ are adjusted properly for FM, then the PROC adds considerable additional talk power and intelligibility to the FM signal. but may overdeviate the signal if overused.

Set Low-Cut to 250-300 and High-Cut to between 2700 and 3000. Don't add too many highs between 1000-4000 Hz, or it will get "hissy" on FM. Don't set the TX EQ for really sharp audio as you would on a SSB signal.

On AM, however, it is a bit dicey. Adding the PROC at anything beyond NORM can make it sound pretty nasty pretty quickly. Don't overdo the bass frequencies, and don't hit the highs quite as hard as you would on a SSB signal. Listen to your own signal using the Full-Duplex method. Or do a quick-record and play back so you can tell what your audio sounds like.

I have yet to experiment with reducing the overall mic gain and making additional changes to TX EQ on AM, because I have a Mic Profile for AM that sounds great without any PROC. This radio can sound superb on AM without the PROC.

Your mileage may vary.

John KB4DU

Since the CESSB is a form of analog peak limiting, and FM doesn’t have peaks, no effect.

KD0RC

Hi John, while it could be implemented using analog circuitry, I am sure that in the Flex it is all done in one of the processors.

My real question is, in the Flex, does the CESSB PROC code get executed in FM and AM, or is it bypassed, or is there a separate code path for FM and AM?

Jim KJ3P

CESSB is a form of peak clipping and mitigation, similar to that which has been used in virtually every broadcast (analog) FM station since 1980 or so.

Audio contains quick short peaks which contribute nothing to intelligibility, but which push down average modulation levels. If those useless peaks are clipped, average modulation levels can go up, but distortion remains due to the clipping products. The "magic" is to process the output of the clipper to greatly reduce those clipping artifacts. I'd guess it could be applied in some form to any audio.

--Jim KJ3P

Dan Trainor

Does it process AM modulation?

KD0RC

Thanks Mike, that is what I suspected. The manual mentions that this is active in AM and FM as well, so a little re-write is probably in order here.

KD0RC

Thanks Ken, very insightful. I almost never use AM, so I will just leave PROC off there. For FM, I will follow your lead and use FDX to experiment with the settings to get good sounding audio.

Thanks for your quick response to my recent help desk ticket. I got my radio back and the transverter now works perfectly! To fully appreciate a Flex radio, use something else for a couple of weeks...

Ken Wells

Hello Len, I would also suggest listening to your Flex FM signal with another dedicated FM radio - differences in receive Filters and de-emphasis may alter your results. It is good to test with the type of FM receiver that most of the people you are contacting will also be using. I have been having great results with my Q5 transverter on FM, but confirmed it on my HT as well.

KD0RC

Great idea, Ken. I will give that a go as well.

I have the little Ukrainian transverter and it is working great. I have not put it on the spectrum analyzer yet, but will give that a try this week. I am currently driving it to 10 W out feeding a Mirage 80 W amp. I rarely use the amp on FM as I can hit all the repeaters around here full quieting at 10 W. I do want to be sure that it is reasonably clean - I may wind up dropping the drive depending on what the spectrum analyzer tells me...