Forum Replies Created
-
Posts
-
[quote author=”Leif”]Wow, I will definitely take you up on that! I’ve never been to Greece — this could be a lot of fun!
What airport is closest to you?
///Leif[/quote]
The Thessaloniki International Airport (SKG) is the closest.
When and if you decide it, just give me an early warning. I have some thoughts… 😉[quote author=”Leif”]Sigmacom, that looks amazing!
Hey, I’m in europe (Twente, Netherlands) all summer. Maybe I should come visit you, show you the Omnia.9 and talk more about future cooperation. 🙂
I’d also like to buy one of those exciters from you. Heck, it sounds like Greece’s economy can use it right now 😉.
///Leif[/quote]
Hahaha! 😆
Make your visit happen, and you will go with a DDS-30 Exciter under your armpit – for free.Actually there are no more exciters left (except one I kept for myself).
The first production was quite small, and prototypes went to local stations.
I liked that, because I could have them back for modifications and return them fast and easy.
That’s why I decided not to send any prototype outside my city – at this phasePrototypes are operating on mountains since September 2010, and I had feedback and suggestions, like as:
– Internal RDS (although many radio automation systems do not support UECP)
– SFN support in basic configuration (audio buffer for delay sync, GPS clock)
– Separate connector for A-MPX (currently used analog Left channel connector)
– Ability to reduce RF power by time schedule (e.g. lower power in night hours)Lack of time, is my root problem… 😳
It depends on the performance you want.
The DDS FM technology is considered as a professional cutting edge solution.
Many ordinary exciters with VCO+PLL, can’t compete at the same time the noise and modulation performance of a DDS.
Also, none analog exciter can process digital audio directly. The "digital" PLL exciters with AES/EBU input, are actually using DAC internally – and then they do FM modulation.I can only see 20 years of arrogance from someone who never actually tried BBP and doesn’t know who’s behind it…
(or he does, and has a purpose)It’s not ready to be announced yet, but it will be A LOT cheaper than any equivalent DDS exciter.
In fact, it will be much cheaper than some good analog exciters, like RVR PTX-30LCD/S+AES/EBU option (catalog price).Hello to everyone! The progress so far, is:
- Re-designed to fit into 1U chassis
- Replaced buttons with rotary encoder
- Added extra features in UI menu
- Fixed some software bugs
- Added extra functions (DC blocker, etc)
- Output RF power is 30W (80W unit, s/w locked)
The last photo is from a mountain installation, operating since September.
Next steps for the next model are:
- Upgrade to a larger DSP chip
- Internal RDS encoder
- SFN support (D-SFN for D-MPX)
- On board Ethernet
I decided to keep this model as a "low-end", and provide the next in 2011.
My promise to some guys here will be fulfilled soon, but currently I need local feedback to correct fast any mistakes found…
😀*It is* relative, indeed! 🙂
But you can set as much bass as you like, as long as you don’t play clipped songs (otherwise enable HPF at BBP).
Also use a DC Blocker (or 1Hz HPF) at the exciter, and you’re done with the low frequency effect.Otherwise, GIGO happens…
Please, let me explain from scratch what I call as "the very low frequency effect".
Will take some time, sorry for that:Imagine that you have an FM modulator at 100MHz, which deviates +/-75kHz when audio is +/-1V (2Vpp).
If you give 20 Hz 2Vpp square wave audio, a perfect linear modulator will force the carrier to be at 100,075kHz for 25mS and at 99,925kHz for another 25mS.
If you try to measure the output frequency, you will find it mostly at 100 MHz, but sometimes you may see a slightly drifted frequency.
This depends on how fast your frequency meter is. The faster the meter is -> larger drifting is shown.Now imagine a very low frequency (1Hz) square wave audio 2Vpp, which results in 500mS at 100,075kHz and another 500mS at 99,925kHz.
Any frequency meter instrument is obviously faster than that, and for sure you will see very large frequency drifting.
Remember, it is not an error, this is what your modulator is supposed to do with that kind of audio!At that time, if someone is auto-scanning the FM band with his radio, there is a chance for a very strict receiver to skip you, because at the time it tuned at 100,00 MHz, your carrier was at 100,075 MHz (or 99,925 MHz). It may stop at 100,05 MHz, or it may skip you at all – especially if your RF signal is poor.
We are talking for VERY implausible conditions! What are the odds for someone to have *that* kind of receiver, auto-scanning at *your* frequency at the *exact* time where you carrier isn’t exactly there for a long time?!!! WOW!
Well, it seems camclone has achieved all of them. His settings to produce tremendous amount of bass on air, produce constantly very strong low frequency components that cause his exciter to be for a long time away from it’s specified center frequency. This is readable with a frequency meter as "frequency drift", and he found some receivers that don’t stop at his station while scanning the band.With analog audio path, the series capacitors block or attenuate such LF components. Also, the PLL functionality will try to correct such errors, so the effect is minimal.
With digital audio path, there are no capacitors to help. With DDS there is no PLL to help. In this configuration, the "very low frequency effect" is fully revealed – unless you put an HPF at 2Hz.That’s why camclone is complaining about digital HPF at me (for DDS) and Leif (for BBP).
Even with the best audio card in the universe, as long as he creates that kind of audio and there is no HPF to stop it, the "very low frequency effect" will not go.P.S.
Sorry for the long post, I hope you understood exactly what’s happening.Not in production phase yet. 🙄
Three prototypes are circulating for evaluation, and expect them to return soon.
As promised, they will go to some guys here in BBP community…I know what’s happening with camclone’s system. I’ve been traveled 750km to see it my self, and then I was able to reproduce it in my lab.
The settings he uses to achieve the amount of bass he wants to have on air, causes a lot of subsonic components – sometimes even DC!
If a very-very low frequency signal is clipped for e.g. 10mS, shifts the FM carrier and holds it there steady for 10mS. This amount of time is enough to trick a frequency meter instrument, and report that the carrier is not at the desired frequency (can show about 1kHz offset or more). This is what camclone have seen at Pira75 gadget.When he switched to digital audio and DDS, the problem was fully revealed, since DDS has completely linear modulation and no capacitor or PLL obstacles.
As long as he sends strong and clipped subsonic components or DC, the frequency measurement instruments will show sporadic carrier shifting "error".
Of course an exciter must modulate what you give, but I have regretted not to put an HPF or a DC blocking filter at the digital input of DDS (the analog input has 1Hz HPF).Now you know why camclone is complaining about HPF… 🙂
[quote author=”Leif”][quote author=”Boki”]Question is,
Does Leif wants to make such STL software?[/quote]Having this exciter in my possession will certainly increase motivation and thus scheduling priority 🙂.[/quote]
OK, I got the allusion… 😆[quote author=”Diekgait”]The serial port could be used to read the status of the SFN adaptar, to see how close it comes to running out of audio in case of a internetstream for example.
Using this together with Leif’s idea of streaming mpx at 128khz 12bit using Flac(Which will result in a bitrate somewhere between 1.1 and 1.3mbit) could make it able to use the internet to create a SFN! Make sure you add quite some memory to the SFN adaptar, so that it can create a large buffer.
And for the delay, 100ms is already to much to monitor of air. So if you can’t monitor of air, what’s the difference between 100ms and 3s? The listener won’t notice:)[/quote]
Buffer size is not a big issue for me, but SFN over internet scares me a bit…
Anyway, it will be a very interesting experiment. 🙂[quote author=”Diekgait”]We’ll maybe a serial interface will do, because you can use a pc to feed the digitall mpx anyway. So that pc can control the box using rs232/usb as well. I hope Leif can find time to write the ip-stl software, this will be so damn interesting to fieldtest![/quote]
Serial port, for what purpose? SFN synchonization? Not needed! 😀Okay… In a few words, the concept is:
Digital L/R audio or D-MPX in AES/EBU format, must be fed into the SFN controller. There, an integrated GPS receiver will provide internally a timestamp, which will be inserted into the original audio AES/EBU frames. Also, a system parameter "MaxNetworkDelay" will be injected periodically. This transforms the original AES/EBU frame into "SFN frame".At each DDS TX site, the received "SFN frame" from an IP or digital radio STL, is passed into the SFN adapter which also has an integrated GPS receiver.
Then, the frame timestamp and the "MaxNetworkDelay" parameter are extracted, and the original audio payload is buffered.
Now, the job of each SFN adapter (besides the reference clock generation for the phase synch of the DDS RF carrier) is to hold each audio frame in the buffer, until the most-distant member of the SFN has received the same frame. When this time comes, all the SFN members will modulate on air the same audio frame.The most important issue here, is the estimation of the maximum network delay. That is, the time that every audio frame needs to reach the most-distant TX site.
With a time-unstable transport media (like IP STL), the SFN operator may have to define a big value for "MaxNetworkDelay" parameter – like 2 or 3 seconds.
That means, if you say "Hi" on the microphone, you will hear it after 3 seconds on your radio receiver, even if there is a transmitter on the roof of the studio.
In a more time-accurate and fast transport media (like digital radio STL), the total delay could be reduced to a few hundreds of mS – or less.[quote author=”Diekgait”]Sounds very interesting! Can you please explain the idea a bit more. Wil it be a combination of tcp-ip to connect the boxes and gps for sync?[/quote]
There must be phase, frequency and modulation syncronization of all RF carriers that constitute the SFN.
I need AES/EBU, no matter if it’s over IP or any other media! At the moment I don’t know if the boxes will have IP interface by default.
Synchronization will be based on GPS receivers and will follow the basic concept of ETSI TS 101 191 specs.
I am dying to describe in detail how it will work, but I don’t think it is wise to reveal this… 🙄
