You can only take a few milliamps. The source impedance is quite high, typically 1K, so the voltage falls by (e.g.) 1V for every milliamp drawn.
RS232 can be half-duplex or full-duplex.
You need to use diodes and smoothing capacitors on all the supply inputs. When the transmit line is high, the diode switches-off, and the capacitor provides the negative supply.
DTR ->|---. | | RTS ->|---o-o----- +ve supply | |+ === 10uF /-\\ | | === GND
Tx --|
I better give you the full circuit:
DTR ->|---. | ____ | | 78 | RTS ->|---o-o--------| L05|----o---o +5V | |____| | |+ | | === 100uF | --- 100n /-\\ === --- | GND | | | === === GND GND
Tx --|
The 1489 is _bipolar_ so there's essentially back-to-back diodes to ground.
I'd add a resistor from "RS232 Level Input" to ground, rather than from the inverter input... otherwise you lose noise immunity.
...Jim Thompson
In article , Terran Melconian wrote: [....]
It doesn't work on quite a few laptops. I think one of the low power RS-232 chips has the zero/one boundary below ground.
Take a look here too:
Withe the version4 of the schematics presented here, using a CAT5 cable (Tx-ground, RX-ground, etc) I was able to transmit data at 9600bps 8N1, over 100m with zero errors with the following module:
So it works good and you may grab from almost all RS232 up to 10-15mA. But if you plan such communications outside, better use an optoisolated RS232.
greetings, Vasile
Dan Lenski wrote:
snip
Is it not ok to push current through the ESD diodes as long as you stay below the maximum specified in datasheet? (and make sure you use at least that current on the rail)
I belive Xilinx specifies some of their fpga inputs for max voltage and max current=20 and you can pick what ever is easiest
-Lasse
I don't want to re-start the old argument, but...
I personally never run signal currents thru ESD clamps.
Others do.
Do whatever you want.
Someday your purchasing department will substitute a Chink chip and your circuit will go up in flames on a customer's site.
I have been working with that literally stole a process from a reputable Taiwan company.
Chink company can't find asshole using both hands, flashlight and a mirror, let alone cope with the ART of designing ESD structures.
Buyer beware ;-)
...Jim Thompson
most of them use the MC1489 receiver chip, and as Jim will tell you anything below about +2V is considered as a low by that chip.
yeah, if you want to send the signal through a mile of doorbell wire having
+/- 12V signals will help.If it's going across the desktop in a shielded cable +5/0V will be fine.
Bye. Jasen
Have something to add? Share your thoughts — no account required.
Ask the community — no account required