HFU HF Underground

Technical Topics => Part 15 AM and FM Station Operation => Topic started by: ChrisSmolinski on April 23, 2020, 2059 UTC

Title: Talking House Transmitter - some spectrum plots
Post by: ChrisSmolinski on April 23, 2020, 2059 UTC
I was curious about the audio bandwidth of the Talking House transmitter, so I did some tests. It is easily 20 kHz, even a bit more. Actually I was generating audio on the computer, so I was limited to 22 kHz. Who knows how wide it really is.

Lots of 60 Hz and harmonics crud around the carrier frequency.

Baseline, TH transmitter off:
(https://i.imgur.com/ENQlnRJ.png)

Carrier on:
(https://i.imgur.com/na9iDRK.png)

White noise:
(https://i.imgur.com/DKkVOop.png)

1000 Hz audio tone, I am fairly confident that I was not feeding in too high of a level and over-modulating:
(https://i.imgur.com/i0LzT7X.png)
Title: Re: Talking House Transmitter - some spectrum plots
Post by: Dave Richards on April 24, 2020, 0349 UTC
For a transmitter whose original intended purpose is simply to play recorded loops/announcements for realtors, and similar simple informational broadcasts, I can't say I'm too surprised. It suffices for that purpose and, being Part 15, is not subject to the audio bandwidth restrictions for licensed stations. Anyone who is keen to limit the audio bandwidth could put an LPF or BPF at the audio input. On the other hand, if you are not required to limit your audio bandwidth by law, why do so? I doubt there'd be a perceptible increase in range by doing so.

It would be nice to be able to get rid of some of the 60 cycle hum though.
Title: Re: Talking House Transmitter - some spectrum plots
Post by: Brian on April 24, 2020, 1817 UTC
Was that into a dummy load?
I suspect the spectrum will be narrower into an antenna.
Title: Re: Talking House Transmitter - some spectrum plots
Post by: ThaDood on April 25, 2020, 1942 UTC
A couple of improvements that can be done to the Talking House / I Am Radio transmitter is to GND the master OSC crystal to chassis GND. That helps to reduce, but not eliminate, the PM (Phase MOD.), coming from that. Then, to smooth out the quality of the modulation and reduce the harmonics, run several GND braids from main PC board GND to the chassis. Mine sounded much cleaner, louder, and fuller, after that. No, it will not put out a more powerful carrier, that will stay a dead-nuts 20-some mW's. But, it will make that TX sound way cleaner. I wish that I could totally eliminate that PM in it. On a C-QUAM receiver, it drives that nuts going in and out of stereo. You can also hear that PM in SSB mode. Other-wise, my I Am Radio TX is a nice sounding AM mono TX.   
Title: Re: Talking House Transmitter - some spectrum plots
Post by: ChrisSmolinski on April 26, 2020, 1306 UTC
Thanks for the grounding tips, ThaDood!
Title: Re: Talking House Transmitter - some spectrum plots
Post by: Josh on April 27, 2020, 1952 UTC
Narrowing the bw will also give a slight increase in range, for those stations that can match that bw on rx.
Title: Re: Talking House Transmitter - some spectrum plots
Post by: Kage on September 26, 2020, 1920 UTC
Narrowing the bw will also give a slight increase in range, for those stations that can match that bw on rx.
I generally agree with this, but I think it has more to do with packing the audio tightly in that bandwidth than it does simply limiting bandwidth.
I've tried both extensively and found that bandwidth restriction alone doesn't seem to have a huge impact, or any that I can tell at all from practical distances. It does give feel good points though so you know you're not interfering with adjacent stations.

What does matter is really cramming the audio into, as you suggested, the bandwidth of the average radio. Placing the NRSC curve to put some punch out to 5kHz makes a ton of difference, beyond that it's up to the station, and my preference is to ride the bandwidth so that up to 5kHz there is strong presence on soft passages or limited audio range (voice), and push the 5-9.5kHz hard to the standard curve, then drop it with the brick wall after that to comply like the big sticks. That presence around 2-5kHz needs to have its own care though or else it will sound awful on music. Professional audio processors actually treat this range specially with analyzing circuitry (or software) so that voice has punch and music remains full and good, and it's an art.

Remember stories of hot audio processors of yesteryear damaging old radio station transmitters? Running tubes too hot or pounding the mod transformer until destruction? That is what gets loudness and to really get that you need 100% modulation peaks and 95-98% neg. Maybe that is what the talking house transmitter needs? Sounds like a fun project :D
Not even sure if the thing can do close to 100% positive, or asymmetrical peak riding even though Stereotool can try its hardest to provide it.
Title: Re: Talking House Transmitter - some spectrum plots
Post by: Kingbear Radio on December 25, 2020, 0110 UTC
Great info guys. The Talking House almost seems like something to play with and not what a serious station would bother using, but maybe tweaking and having it  as a a backup would be a good thing.

It's good for learning's sake though, once in the last decade there was a tech forum about modifying a toy transmitter, it was called the Planet DJ, a gadget for preteens that broadcasted on 20 foot on one signal frequency, yet the guys submitted many modifications and schematics about how to hot rod the circuit for better range and power, they were really into trying to make the planet into a full radio station!
Title: Re: Talking House Transmitter - some spectrum plots
Post by: ThaDood on December 25, 2020, 2316 UTC
What's is nice about The Talking House, a.k.a. I Am Radio, transmitter is that it is PLL agile from the whole AM broadcast spectrum, while the Sean Cuthbert AM Stereo transmitters low-end stops at 1MHz. Still a very usable agile transmitter to do tests with a simplified AM Carrier-Current Coupler, albeit that below 1MHz is more ideal for Carrier-Current. Boomer is working on getting the best effective toroid 50 Ohm to 10 Ohm, or less, RF / AC isolation transformer. At least, find a good compromise that's effective, yet broadband enough. Oh!!!! Check out PLC = Power Line Coupling that companies are using in medium wave for data communications. I guess that we could call this Carrier-Current 2.0?