[TenTec] Omni VII Speech Processor Question

Fri Mar 15 22:17:15 EDT 2013

In my evaluation of the Omni VII Speech processing system I concluded that 
there are some 9 definable steps found over the displayed range of 0% to 
100%.  These are not exactly linear in their appearance.  In other words, 
changing SP value from 18% to 28% did not actually yield any measurable SP 
level change.  However, changing from 28% to 29% nets a 2.4 dB increase as 
this change transitions from one step to the next step.  In another area of 
the range, a value between 39% and 49% yields no level change but changing 
from 49% to 50% yields a 3.5 dB change in level.

It is also noted that a SP value between 0% and 6%, I find actually adds 4.3 
dB of loss in the audio path.  A value of 18% yields a unity gain 
configuration.  In one case I read where the user was operating the SP at 
10%.  My findings would indicate that this value has added some 3 dB of loss 
to the audio path.

If one chooses to use the Speech Processor effectively, I suggest a value 
between 30% to 50% where 30% nets 2.4 dB of processing gain, 39% nets some 
5.3 dB of processing gain and 50% nets some 8.8 dB of processing gain.  At 
values above 62% the signal has basically has lost most all of its dynamic 
range.

I find the steps, related to the value % displayed are found at: 0%, 7%, 
18%, 29%, 39%, 50%, 62%, 73%.  I was not able to identify the other two 
steps as my method of measurement allowed the radio to enter the ALC range 
thus the output was no longer increasing.

On the down side of any processing application, since the processor raises 
the average level, it will also raise the average level of any background 
noise in the radio room by the same amount.

Successful operation of the internal Speech Processor requires that the 
hardware mike gain be adjusted correctly to attain a DSP MIC gain value of 
between 35% and 65% for proper ALC action, thus occasional flash of the ALC 
LED on voice peaks.  Then adjust the SP value for the desired average 
increase in output power without causing a significant increase in back 
ground noise.  Also, careful attention to the mouth-to-mike distance must be 
employed as working too far from the mike will allow room noise to become 
more apparent.  With the 709A mike, I prefer about 1" distance between the 
mouth and the mike face.  Stating it another way, the more processing 
employed, the closer one needs to be to the mike face and then the DSP MIC 
gain will necessarily need to be reduced.

Notes from my earlier tests:
<snip>
I did find is the following data and the actual SP values vs. power output. 
It is noted that in between the steps there was no power change measured. 
Case and point, at values between 18% and 28% the power remained at 2.0 
watts.  Between 28% and 29% the power changed from 2.0 W to 3.5 W.  The same 
held true for other increment of steps.  Back some time ago, John Henry 
confirmed that there are actually 9 steps in the Speech Processor control 
algorithm.

Power output = 2 W with SP off
With the SP on the following is observed:
 0% = 0.75W   or - 4.3 dB
7%  = 1.0W    or - 3 dB
18% = 2.0W   or   0.0 dB
29% = 3.5W   or + 2.4 dB
39% = 6.7W   or + 5.3 dB
50% = 15W    or + 8.8 dB
62% = 40W    or + 13.0 dB
73% = 70W    or + 15.4 dB  ALC is now on and no further increase in SP gain 
produced any increase in output power

This does concur that the effective SP value should be about 40% to 50% to 
attain some 6 to 8 dB of increase in average output power.
<snip>

73
Bob, K4TAX

----- Original Message ----- 
From: "Richards" <jruing at ameritech.net>
To: "Discussion of Ten-Tec Equipment" <tentec at contesting.com>
Sent: Friday, March 15, 2013 8:25 PM
Subject: Re: [TenTec] Omni VII Speech Processor Question

> HOLD THE PHONE, GUYS.... IT IS AN RF-BASED COMPRESSOR AFTER ALL!
>
> HERE IS WHAT STAN BROCK SENT TODAY, which
> appears to have some input form the infamous
> John Henry... I love to see tag-team
> problem- solving!
>
> Based on this explanation, and what Rick, DJ0IP says, I will keep my Model 
> 715, notwithstanding the Omni VII has an RF-based compressor after all. 
> Sometimes a dedicated device can do better than another device which must 
> handle multiple duties and functions.
>
> Rick usually has good advice... even if I don't
> move on it right away ...   ;-)
>
>
> Thanks for the input !
>
> ----------------------  K8JHR -----------------------
>
>
> On 3/15/2013 8:21 AM, Stan Brock wrote:> James,
> >
> >
> > Sorry, I went back into the code and we indeed have it in the RF
> > processing section in the DSP, Very sorry for the incorrect answer. I
> > need to verify things before I answer.
> > The 712 itself does do a better job because it is taking the af of the
> > voice, converting it to rf, etc back to af, without the other rf
> > processing done inside the dsp, so, it is more effective. Most find the
> > internals work about as good. But some of the compression is lost in the
> > DSP where it is not lost in the 712.
> > My bad, sorry.
> >
> > Thanks, and 73,
> > John Henry, KI4JPL
> > TEN-TEC Engineering
>
> ---------------------------------------------------
>
>
>
>
>
> ..
> _______________________________________________
> TenTec mailing list
> TenTec at contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
> 