Re: DTMF dead?
- From: Tim Shoppa <shoppa@xxxxxxxxxxxxxxxxx>
- Date: Sun, 08 Jul 2007 16:14:37 -0700
Nico Coesel wrote:
"Mook Johnson" <mook@xxxxxxxx> wrote:
"Nico Coesel" <nico@xxxxxxxxxxx> wrote in message
news:468fb22e.156164602@xxxxxxxxxxxxxxxxx
Vladimir Vassilevsky <antispam_bogus@xxxxxxxxxxx> wrote:
Nico Coesel wrote:
I have an upcoming project in which I might want to use dtmf
transceiver chips as a simple means of communication. However, it
seems that dtmf transceiver chips are quite hard to get these days.
Even Digikey stocks none. Anyone knows a source that keeps making dtmf
transceivers for the years to come?
It is possible to encode and decode the DTMF with pretty much any
microcontroller. For example, you can do this:
http://www.abvolt.com/research/Multitone%20detect-IEEE.pdf
However, if the goal is the communication, I would rather implement a
PSK or FSK modem using a microcontroller. It would be much better and
even simpler then the DTMF.
I was hoping to use the analog front-end which is usually quite good
(reception downto at least -30dB). Its hard to beat that with a
microcontroller implementation (been there done that).
PSK / FSK is also an option. I just need something cheap and reliable
that will communicate over a 2 wire system carrying both communication
and power.
If you have a DC return path beside the two wires I'd recommend using
phantom power and going high speed RS-485.
You apply the DC (or AC) to the center tap and hte telemenry to the
secondary. ON the two wires the power is on the common mode and the
telemetry is on the differental mode.
On the other end just pick off the signals in the same order. (power on the
center tap and telemetry on the secondary).
The sort of wiring hasn't been nailed down yet, but at this moment I
would prefer a system with 2 wires which carry both power and
communication. RS485 is also in the picture, but we need to have up to
several hundred devices on a short bus (say 20 meters / 60 feet). I
think this would be stretching RS485 too much to its limits even at
low baud rates.
With that many devices on such a short wire, the limiting number
you're seeing might just be leakage current in the drivers and it's
not really being fair to take them all worst-case unless the
tempurature is really 150C...
Also, the communication comes from a UART. Using a
transformer on DC biased signals is not a good idea.
It also seems an NXP LPC2101 ARM micro is even cheaper ($1.65) than a
NXp LPC935 8051 based micro ($1.82) so I think software FSK -which
I've build before- is not out of the question. DTMF is a whole lot
more complex to decode than FSK.
Again, don't ignore the least-common denominator of OOK. I'm not sure
why you're all that worried about receiver dynamic range and the
advantages of DTMF there. If you're really that price sensitive and
can put enough overhead bits in to overcome noise and collisions, OOK
really does work in bazillions of real world super-cheapo
applications. DTMF made perfect sense for a while in the 60's when the
limitations of no sequential digital logic met the bandwidth of a POTS
circuit met the long-lines of a transcontinental haul. But although
you haven't told us what you're doing or the full details of the
environmnet, with a run of just 20M you're not in the same ballpark
transmission-wise, and you are a half-century ahead in terms of
microcontroller availability and ease. Use those advantages!
Tim.
.
- Follow-Ups:
- Re: DTMF dead?
- From: Nico Coesel
- Re: DTMF dead?
- References:
- DTMF dead?
- From: Nico Coesel
- Re: DTMF dead?
- From: Vladimir Vassilevsky
- Re: DTMF dead?
- From: Nico Coesel
- Re: DTMF dead?
- From: Mook Johnson
- Re: DTMF dead?
- From: Nico Coesel
- DTMF dead?
- Prev by Date: Re: A mobile phone display that's less than 1mm thick [PIC]
- Next by Date: Re: Flexibility of hours for EEs
- Previous by thread: Re: DTMF dead?
- Next by thread: Re: DTMF dead?
- Index(es):
Relevant Pages
|
