Hi, i'm looking for an RS232 transceiver with these features
It's easy to accept +-12 as input... just divide it down resistively.
I've blown up two HP ProLiant server serial ports, so far, by applying
12 volts to a serial port input pin. I just swapped the boxes with other people who don't need serial ports. But the chips couldn't stand an honest +12. I'd consider adding some external protection.You can do a decent RS232 transceiver with a 12-cent dual opamp, but you'd need a $4 DC/DC converter SIP to do that from 3.3 volts.
John
TL16C2752 from TI.
On a sunny day (Sun, 30 Jan 2011 10:03:59 -0800 (PST)) it happened Bullwinkle wrote in :
MAX233, AD233 It can work with +-12V input because of an input resistor doing current limiting.
+- 6V out should work for most applications.
True RS-232 must go to +/- 25 volts.
tm
The Specifications table in that datasheet seems to show a max input range of +/- 30 V, so it shouldn't have any problems with +/- 12 V. How much it can grunt out isn't necessarily related to its stage limits.
-- Rich Webb Norfolk, VA
Wrong again. It should not *exceed* an open-circuit range of +/- 25 V.
-- Rich Webb Norfolk, VA
What doesn't it do?
A "true" RS232 driver must provide +-5V into 7K ohm, a receiver must function with +-3V input. A "true" receiver must tolerate +-25V input.
Most if not all of the devices available meet these (e.g. MAX3232)
kevin
Where are you getting your specifications for this and what are they? Do they include, as the last standard I read does, slew rate limits for outputs less than 30V/us (which limits data rate) but also no more than 1ms in the +/-5V output transition region, for example? Active input levels valid and accepted properly from +/-3.0 to +/-15.0V while active output levels always between +/-5.0 to +/-15V into a load of between 3k and 7k ohms (not accounting for capacitance and rate.) Shunt capacitance including cable of less than 2.5nF? Output impedance of more than 300 ohms with the power off? Etc.
Jon
Don't tell me; tell HP!
John
The definitive specification is EIA-232 (D?) (the RS-nnn variants were 'recommended' pre-adoption versions of the standard). If high speed is ever required, RS-423 may be a better choice, and interoperates nicely with EIA-232.
That's impossible in silicon, there has to be voltage-conversion hardware added; the 'single chip' gizmos take a bevy of (usually external) capacitors and such to implement the conversion.
But no handshake? There are some devices that need the handshake signals, or to tap them for power...
TI has TRSF3221 which does this..
Why? Is it painful to type lines of a parts list? Or is it about circuit board area, or cost? Lowest additional components argues for a +/-5V power source
Four Dollars ?! Now that would need some serious muntzing :-)
Moat dunking used to be the usual "remedial action" for screw-ups in medieval Europe. Typically done with bakers, after they were caught selling sub-par rolls.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Because someone will plug an RS-232 device into that which provides an actual +/-12V (or +/- 25V) and the bare minimum unit will go Phut!
-- Paul Hovnanian paul@hovnanian.com ---------------------------------------------------------------------- Have gnu, will travel.
I believe that the inputs can handle +/- 12 with no ill effects.
The output levels were being discussed, because of the +/- 6V charge pumps in these newer chips.
Outputting +/- 6V will not get 1000', but it will have no problem with
6-10'.hamilton
I have two books that gives info for connecting computers, terminals, printers etc via RS-232 with interconnection info listed by model; printers in 1984 and 1988; very useful those daze. Both include RS-232 and CCITT specs and the more recent one includes the parallel Centronic specs. Well...not one word concerning voltage levels or polarities for any of those "specs"; could be anything from a femtovolt to a gigavolt and the currents could be anything from attoamps to tera-amps for all the talk of circuits. So...back to what industry used for a guide. Ref: data sheet MC1488 quad line driver meets RS-232C and CCITT v.24 specs. Supplies +/- 15V (standard in those daze), input rating: -15 to
+7V (not as good on + voltage as other RS-232 parts, but usable), output voltage +/- 15V. Used hysteresis, considered necessary to reject line noise. Seems that signal output range for operation is +/-9V, with min of +/-6V. Ref: data sheet MC1489 quad line receiver (the "companion" part) allows up to +10V supply (no neg V used), +/-30V allowed on input, 1.0V min high threshold, 0.75V min low threshold. Some competitor parts allowed +/-3V for worst case output.Is this of help?
Oh, yes; i understand the following: that the RS-232 spec requires that the driver and/or receiver withstand 25V fault; that it does not say that it must work after that fault. I think that "true" RS-232 required +/-5V minimum so anything powered with 3.3V does not compute. That said, +/-3.3V signaling is not outside what i have seen in industry in the past. What few parts that conformed to the 25V fault spec could be considered to be "inherently" ESD protected (exception: parts made in those daze by Harris - utter crap for reliability). Some required external capacitors to meet rise and fall time specs. It is notable that only the Fairchild "triples" not only met but exceeded the specs and had its own internal capacitance to guarantee rise/fall specs. That was back in the daze of the original Fairchild making analog, digital, RF, diodes, etc & etc. uA070 users UNITE!
That's impossible, unless you've got additional power supplies besides Vcc. See:
Good Luck! Rich
red
.This is uber silly. The part has to work after receiving 25V.
I've used this chip often:
When dealing with I/O you need the real deal. I had some very bad experiences with cheap Sipex devices.
-- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico@nctdevpuntnl (punt=.) --------------------------------------------------------------
The 232 chips I designed just divided down the input with poly resistor, then feed the divided signal to an inverter. I really can't believe this is even a topic for discussion. All the Maxim parts did real 232. I laugh at the notion that the Maxim 232s are unobtainable. That was one of two chips that built the company.
Note one thing to check on any of these 232 parts is the effect of the charge pump or DC/DC on the jitter of the transmit and receive path. Nobody specs this, but the jitter is there. The ATE needs to know the jitter to determine when to give up waiting for a transition. Every vendors has the jitter as an internal spec.
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.