Unsolved Technical Problems

transmitter?  Can you make a pre-emphasis

network with discrete parts that would be

low cost, and just put it in the line between

the 9030 and the CC transmitter?  Maybe not,

or you would have already tried it.  Oh I forgot -

do you still feed a Part 15 FM transmitter to get

your signal downstairs to the 9030?  How do

you get the signal downstairs to the LPB rig?

Bruce

Here's a link to a discussion about this.

http://www.antiqueradios.com/forums/viewtopic.php?t=128055

Though more opinion than fact, two things to note are that AM radios have narrow IF bandwidth and the treble control can be used to de-emphasize the highs. But, again, this is opinion but a simple experiment could help you develop your own opinion.

The pre-emphasis is a simple roll up which can be done with a couple of Rs and a C but the problem is at the top of the audio band. A "brick wall" filter is needed which cuts about 50 dB from 9.6 to 10.5 kHz. This is not trivial.

My station does not use pre-emphasis and on portables and old radios it sounds fine. On my 2010 model truck radio it does sound a bit dull due to lack of highs but I can compensate with the treble boost.

Neil

Hi Carl,

Here's a simple solution. Move the preemphasis jumper in the AMT... transmitters to the OFF position. Turn on preemphais in Sterieo Toools for your Output 2. Feed all three transmitters from that output.

PhilB, my Winamp Stereo Tool combination also feeds three online audio streams, so the NRSC Pre-Emphasis in Stereo Tool is not an option.

Yes Bruce, the link to the LPB is via the Technics FM Tuner, and building a R-C network is an option, Neil.

Thing is Stereo Tool has a bandpass filter which I set 80 Hz to 5kHz.

Sounds fine.

This thread worked.

What is your next "Unsolved Technical Problem?"

This morning I reset the bandwidth of the Stereo Tool audio processing to 80 Hz - 10 kHz, as specified in the NRSC Pre-Emphasis Specification.

ALSO, forgotten in the above discussion, is the fact that in addition to driving three web stream servers, the Stereo Tool output also feeds a C.Crane FM Transmitter STL-to-LPB channel, another reason NRSC cannot be sent from Stereo Tool, but must be set within the AMT3000/5000 transmitter pair.

Neil's idea of a C and 2x R simple EQ network for the LPB will, as indicated above, solve this first Unsolved Technical Problem.

Your audio distribution is complex and conflicting.

If I understand correctly you are feeding from one source (your computer with Winamp and Stereo Tools to a total of 3 servers (internal to the computer) and 3 actual physical audio connections (C. Crane FM, AMT3000 and AMT5000 transmitters). The various destinations call for different conflicting Stereo Tools settings.

The current Stereo Tools setting of 80-10K bandwidth is a needless limitation for the stream servers. The 10K upper limit is good for the AM transmitters, but you can't turn on NRSC preemphasis because of the stream servers. It really seams like it would be best to feed the stream servers separately from a separate instance of Stereo Tools. You could then set one instance of ST tailored to the streams and set the other instance of ST to full AM mode with the 10kHz limit and NRSC preemphasis. I don't know if you can do that without having two separate computers. You would know that better than I.

Otherwise, Neil's suggestion to put a simple RC network at the input to the LPB transmitter is good, but calculation of the actual R and C values depends on the actual audio input impedance of the LPB. Whatever is specified in the LPB manual may or may not be in the ballpark. You would need to calculate the R and C values to match the NRSC curve and then do some measurements and experimentation to get the final values. Or, more practical, would be to just use the calculated values and accept the possible inaccuracies.

Another solution would be to insert an inexpensive audio graphics equalizer at the input of the LPB and adjust it until you get an approximation of the NRSC preemphasis curve. Search the web. There are some that are only around $25. See for example: http://www.amazon.com/s/?ie=UTF8&keywords=graphic+equalizer+home+audio&tag=googhydr-20&index=electronics&hvadid=45258147182&hvpos=1t1&hvexid=&hvnetw=g&hvrand=14278739713019759797&hvpone=&hvptwo=&hvqmt=b&hvdev=c&ref=pd_sl_6ufqf45qbn_b

Either the RC filter or a graphics equalizer will take care of the NRSC preemphasis requirement and the Stereo Tools setting you selected will limit the upper audio bandwidth to 10kHz to take care of the NRSC "brick wall" cutoff.

It bothers me a bit that you are limiting you bandwidth to 10kHz on the stream servers, but if you are OK with that, then it's OK.

Setting aside for a moment the matter of audio bandwidth for AM transmission, we look at the specifications for audio bandwidth of streaming servers.

My server configurations are based on a belief that much of the world still suffers from low speed internet connections, older-slower computers, and perhaps a lot of dialup accounts.

While many streaming stations today send high bit-rates like 192kbps, I think they are limiting their audience to the most fortunate internet subscribers.

For a long time I ran a 16kbps stream, reasonable usable by dialup listeners, but turned it off due to non-use; no one listened to it.

Now my minimum stream rate is 32kbps,11,025kHz, stereo.

If the audio bandpass were set for that stream it would cut off at 5,512.5kHz sample rate (half of 11,025kHz).

This morning I connected and listened to that stream, and the audio remained clear without high-end artifacts from mismatched bandwidths. But the programming at the time was "old-time radio" followed by speech with a telephone guest. Likely there was no high frequency content.

Later today I will monitor the 32kbps stream again, during the broadcast of a Richard Strauss opera. This will be an ultimate test, because the Strauss is loud, harmonically very rich, and recorded by the world's best acoustic engineers for the WFMT Network of Chicago.

Is my thinking reasonable, or am I missing something?

Having now heard close to an hour of sonically splended streaming coming across at 32kbps 11.025 kHz stereo, driven by Stereo Tool with a bandpass of 80 to 10 kHz, there is nothing to complain about.

The music quality is similar to an FM station sound a few years back using a 15 kHz equalized telephone line.

I must assume that the internet stream bandpass-limits itself... there is no high frequency disturbance detectable by this listener.

Each one of the mp3 audiofiles comprising a day's radio playlist come from different producers that use their own setups, and for that it's amazing how most of it is "broadcast quality."

The comment line remains open.

As the only person available to respond to himself, because everyone else is downloading Oompah Hours, Free Minds Radio and Blare OnAir, here is an FM antenna post made last Wednesday on another thread, without response:

THIS website provides an example of what it will be like when the human population reaches near extinction.

But for now, it isn't humans near demise, it's AM radio and Part 15 Hobby Radio, owing to zombie culture... people walking around with massive cell-phone induced tumors.

AS a person without a cell phone, I am still able to post, and bring another technical issue to this "problem unsolved" thread:

News websites.

Alright, what do news websites have to do with Part 15?

EVERYTHING!

Not to mention that all computer components contain a Part 15 Boiler Plate Tag. AND also many Part 15 radio stations have websites EVEN IF THEY DO NOT STREAM.

Then there's the matter of news gathering for radio reporting, which requires visiting news websites.

FINALLY we reach the point of all this...

The news websites I visit tend to repeat and repeat the same news stories over and over so-as to appear more content filled, which causes fatigue.

AND news websites seem to have video advertisments that start playing automatically, which clutters the experience and causes distraction.

NOT ONLY THAT BUT many link choices lead off and off into a land of memory glut causing freezes and crashes.

Perhaps malware is throwing a picnic.

IN THE END it's just good to get out of there, without ever knowing the news.