Questions about FSK

I am building a multi-drop communication system that will communicate over power lines (12V DC). After some discussion and research, I've decided to use FSK. The system will have 1 master and multiple slaves. I want it to be full duplex (so the master can always be transmiting and 1 slave at a time can be responding). The master will transmit at one frequency and the slaves on another, I want to couple the outputs onto the +12V line. I'm going to use center frequencies of ~100k & ~150k. I'm going to use a CD4046 (74xx4046) PLL chip.

My question is what is the best way to overcome the loads on the power line. The loads I have are the power source (which will most likely be a battery), the voltage regulators (at each of the slaves), the FSK receivers, and the FSK transmitters. From some testing that I've done, it appears that the output impedance from the PLL chip is about 1k. From the data sheets it looks like the input resistance to the receiver is about 400k ohms, and the input sensitivity is about 1/2 volt. (the input resistance goes down and the sensitivity goes up when the supply voltage increases)

So far I have come up with the following solution (this is where I'm open to suggestions). I can isolate the battery with an inductor (not isolate, but increase the impedance looking towards the battery) and I can do the same with the voltage regulators. (but this has a downside if there is a quick demand for more current, and I would need to use large inductors to get any effective resistance). The output of the VCO can be turned off, but I don't know what it's impedance is (if it goes high it would help, but if it stays at 1k, then I might need to add some other kind of buffer).

I would like to be able to have a lot of nodes on the line (maybe as many as 99), so if anybody has any suggestions, please post.

thanks

-jeff

ps. - I've seen other companies that sell a multi-drop FSK modem that can couple on to any lines (power, pbx, etc.) and they don't seem to have any problems with loading, so there must be a way to do this.

Reply to
Jeff Hendrix
Loading thread data ...

Why reinventing the wheel and not just using x10 protocol, etc?

There is no way to overcome the load of the power line. Power line behaves close to the short circuit at 100...150kHz.

In order to be operable, your system has to accept the input signals with the level of 1mV or below. It is not a big problem to achieve that with the analog filter and amplifier.

That does not look like a good solution.

Just increase the sensitivity of your receiver.

Vladimir Vassilevsky, Ph.D.

DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

That's pretty common. There's a whack of a lot of industrial process control stuff in the field that talks Bell-202 (1200 baud FSK) on top of DC supply lines.

FSK on top of DC power is _very_ common. There's way more of it out there than X10.

You use either a real inductor or a "simulated" inductor that's built out of active components so that your power supply and other devices are all high impedance at signalling frequencies.

You never heard of a choke?

Not sure what that has to do with anything. The PLL is the recieving half of it.

You're going to need analog driver/receiver circuitry talored to the type of wiring and the blocking impedance you choose.

Why not? It's done all the time in industrial gear.

You put a blocking inductor in your power supply and in your devices so that they all present a reasonably high impedence at signalling frequencies.

Wrong.

Ah. That explains it.

--
Grant Edwards                   grante             Yow!  Who wants some
                                  at               OYSTERS with SEN-SEN an'
                               visi.com            COOL WHIP?
Reply to
Grant Edwards

Bell-202 implies signaling in the voice band. Good luck fighting interferrence.

The power line is intended to deliver power. Nobody is going to do you a favor of placing chokes everywhere.

Because it increases the impedance. How about a choke which can handle 100A current?

Power line communication works indeed despite of what you said.

You got a lot to learn, dear fellow.

Vladimir Vassilevsky, Ph.D.

DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

Like I said, there are hundreds of thousands of units out there in chemical plants, refineries, sewage plants, breweries and pharmiceutical plants that have been working fine for 10+ years.

Yes they will. You either design them into the power supply and devices themselves, or you spec them in the system schematic and make sure the electricians install them when the devices are installed.

We're not talking about AC mains you know. We're talking about a two-wire 12VDC bus installed specifically for the devices in question.

What on earth are you talking about? These devices typically draw 10 or 20 mA, and there are rarely more than a dozen or so on a line. We're talking an Amp or two max at the power supply.

I don't think you're talking about the topic at hand.

Anybody who thinks he doesn't is a fool.

Never trust anybody outside of academia who insists on putting Ph.D. after his name...

--
Grant Edwards                   grante             Yow!  ANN JILLIAN'S HAIR
                                  at               makes LONI ANDERSON'S
                               visi.com            HAIR look like RICARDO
                                                   MONTALBAN'S HAIR!
Reply to
Grant Edwards

Just to clarify things a bit. This system is isolated from everything else. The master will have the power source that will be used to power the rest of the system. Most of the cable runs will be short (5-15 feet between slaves, entire run probably not more than 100-200 feet. Occasionally maybe longer runs of a few hundred feet).

I was planning on putting a choke between the source (battery) and where the signal will be coupled onto the line. I will probably also put a choke on the slave boxes before the voltage regulator (does anybody know what the input impedance of a linear, say a 7805, regulator is?) I don't expect the current supplied for the entire system to be over a couple of AMPs (each of the slaves will have an AVR microcontroller and a few other chips).

Seems like most of the discussion is things I've already thought about, I guess I should just break down and build a handfull of receivers and make some measurements as I add more to the line.

-jeff

Reply to
Jeff Hendrix

The 8705 requires a cap on the input. Nevertheless, you can add a series inductor before that.

If possible, have the signal feed symmetrical, meaning a set of compensated chokes for isolation and a transformer for coupling the sender.

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

You may want to take a look at the Exar XR2211. I used it for a similar application (many years ago), it has a wide input range, and is pretty cheap.

-Hershel

Reply to
Hershel Roberson

FSK (or other modulation) transceiver at the speeds of up to 2400 baud can be easily implemented with a microcontroller like PIC or AVR with minimum glue parts.

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

If the CPU clock is fast enough. If it's a low-power application, the CPU may only be running at a few hundred KHz.

--
Grant Edwards                   grante             Yow!  I am a traffic light,
                                  at               and Alan Ginzberg kidnapped
                               visi.com            my laundry in 1927!
Reply to
Grant Edwards

The other thing is that you mustn't try using the frequency-phase detector (PD-II) in the 4046 for this. Any time you get a noise pulse that crosses the logic threshold, your PLL will lose lock and have to reacquire. Multiplying phase detectors, e.g. the XOR PD-I in the 4046, are much better at rejecting noise.

Cheers,

Phil Hobbs

Reply to
Phil Hobbs

Hi there,

I have also used the 2211 chip in a FSK system and found it very good, by good I mean a great balance of price/performance. I used it with the XR2206 sine wave generator as the Tx part and a transformer to step it up in amlitude a bit.

The system I used as the network was an electric fence, so I had similar requirments to powerline modems. I was able to get up to 5km distance between Tx and Rx on this lossy wire, not nice copper like house wire, with this EXAR pair. The devices had to appear as high impedance but still allow my communications signal in and out. The key was the coupling circuit. I think you should investigate "resonant" circuits. If you can't seperate, spectrally, your comms signal from other high amplitude signals on the line, in order to filter them, then you have to look at using a type of filter with really steep rolloff.

My favorate powerline modem design is the motorola DSP one, here is a link to the site which has the code and reference circuit.

formatting link

In the document DRM035.pdf there is a section on coupling to the power line. I also think the ST app notes on this topic are not to bad.

good luck, Tony McKay.

Reply to
Tony McKay

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.