RS-232 levels to computer

[[email protected]]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

 

[Jim Thompson]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

Yes. (Most RS232 ports don't actually follow the spec.)

...Jim Thompson

 

[John Fields]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

 

[Joel Kolstad]

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

....driven into a 1k load, or something like that.

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

In many (even most) cases, yes, but if you're designing this as as real
product (not just a one-off device for yourself & friends), in a room full of
miscellaneous computers you can assuredly find a machine that *won't* accept
ground as being a logical 1. (It tends to be more older computers.)

For a real product, I would just use a Maxim 3311E and know I'm meeting the
spec.

---Joel Kolstad

 

[John Jardine.]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

All the technical spec's suggest it shouldn't work but on 2 of my PCs
actually works well. For testing purposes saves the ballache of finding a
negative rail.
Would seem those highly specified RS232 Com ports are nicely 5V-logic
compatible.
john

 

[Tim Wescott]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

Not reliably. Fortunately the RS-232 connector gives you both voltages
as part of its signaling (someone please post what they are!). If I
were doing this I'd do a web search on "port powered RS-232" or some
such, then I'd implement a circuit that looks like this:

+12
===
|
|
.-.
| | R2
| | 2k2
'-'
|
| TX
o---------o
R1 |
4k7 Q1 |
TXdata ___ |/
o-------|___|--- ----o------| Q2
v / | |>
--- | |
| .-. |
| | | R3 ===
=== | | 4k7 -12
GND '-'
|
|
===
GND
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)

Q1 is a PNP level shifter, R1 limits the emitter current of Q1 to around
1mA (for a 5V VCC). Q2 is an NPN inverter. R3 guarantees that Q2 will
shut off smartly.

Dunno how well this will work, I've never actually implemented it --
I've just seen similar arrangements elsewhere & thought they looked
clever. There really ought to be a chip out there that does this for
you, but I've never had occasion to really dig for one.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google? See http://cfaj.freeshell.org/google/

 

[petrus bitbyter]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

FAIK the specs want +/-5V at the transmitters and +/-3V at the receivers. I
ever got a circuit that worked well using TTL-levels... most of the time.
The trouble did me decide never to use TTL to RS232 so I always used 1488/89
(later MAX232) ever since.

petrus bitbyter

 

[Meindert Sprang]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Like others said: maybe, maybe not. One simple way of doing it is as
follows: driver the led of an optocoupler. Connect the emitter to RTS,
collector to RxD, a resistor of 2k2 from collector to DTR. Make sure your
software activates RTS and deactivates DTR. this will provide power to the
transistor. Now you'll get a fully bipolar signal AND galvanic isolation for
free.

Meindert

 

[Spehro Pefhany]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Tnx de PDRUNEN

Yes, it will be, almost certainly. The actual thresholds in receivers
are almost always somewhat above zero volts.

This is an okay solution if the connection is short, temporary, and
you have control over the computer used (for example, a setup port on
a product, which is only use in the manufacturing process, would be a
perfect application).


Best regards,
Spehro Pefhany

 

[Don Bowey]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

That agrees with ANSI/EIA/TIA 232-E-1991. It allows for + and - 3 to 15V.

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

That depends on the receiver; some do, some don't. If it doesn't you can
put a DC bias (negative with respect to ground, in your case) on the RD
lead, so the the RD goes slightly neg when the +12 is not present. Uou may
need to insert a few kOhms in series from the SD.

Don

 

[linnix]

Like others said: maybe, maybe not. One simple way of doing it is as
follows: driver the led of an optocoupler. Connect the emitter to RTS,
collector to RxD, a resistor of 2k2 from collector to DTR. Make sure your
software activates RTS and deactivates DTR. this will provide power to the
transistor. Now you'll get a fully bipolar signal AND galvanic isolation for
free.

Relying on other signals are not as clean a solution, since some cables
connect 3 pins (TX, RX and GND) only. Some circuits use the incoming
RX off state (-3V to -12V) to store some charges and feed it back to TX
via some transistors. It works as long as the local TX duty cycle is
less than 50%.

 

[Meindert Sprang]

Relying on other signals are not as clean a solution, since some cables
connect 3 pins (TX, RX and GND) only. Some circuits use the incoming
RX off state (-3V to -12V) to store some charges and feed it back to TX
via some transistors. It works as long as the local TX duty cycle is
less than 50%.

Well, since the OP is designing his circuit, this wouldn't be a problem,
would it?
And for the cables: even the cheap bulk $0.50 cables have all 9 pins
connected.

Oh, and most important: I have actually used this setup many times and it
*always* worked.

Meindert

 

[budgie]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

Tyhe spec requires that a receiver will respond correctly to signals in the +3v
to (IIRC) +15v range as positive, and -3v to (IIRC again) -15V as negative.

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

If you are talking Vce(sat) above ground then IMOextensiveE with RS232
interfaces this will work on 99% of PC clones, ranging from those with 1488/1489
chips to the latest LSI incarnations. As someone lese pointed out, most
receivers seem to actually respond as TTL inputs (with clamping).

Whether you have a +5V or +12V on the collector load is pretty much immaterial.

It is certainly NOT a methodology recommended for production. For that I use
dedicated MAX chips and generate a true bipolar swing, but for your own
quick'n'dirty interface it'll should be fine.

 

[Richard Henry]

Hi All,

I understand that standard RS-232 levels must be bipolar with at least
a -3V for a logic "1" and +3 volts for logic "0".

The micocontroller 232 level is TTL level, I use this signal to drive a
NPN so the the signal is inverted and use +12 on the collector to get
the level. The lowest voltage out of the NPN will be near zero. I am
using this for just collecting some data sent over a small cable lengh
and I don't plan on this being part of a design.

The question is in regards to the computer RS-232 input, would the near
zero voltage be taken as by the com port as the correct level?

Don't count on it. However, many poorly-designed, non-spec-compliant
interfaces will work with those voltages.

Sometimes.

Intermittently.

 

[Richard Henry]

Yes, it will be, almost certainly. The actual thresholds in receivers
are almost always somewhat above zero volts.

I did not know that. Please provide a reference to a EUA-232 receiver with
such a specification.

This is an okay solution if the connection is short, temporary, and
you have control over the computer used (for example, a setup port on
a product, which is only use in the manufacturing process, would be a
perfect application).


Best regards,
Spehro Pefhany

http://www.speff.com

 

[Richard Henry]

Relying on other signals are not as clean a solution, since some cables
connect 3 pins (TX, RX and GND) only. Some circuits use the incoming
RX off state (-3V to -12V) to store some charges and feed it back to TX
via some transistors. It works as long as the local TX duty cycle is
less than 50%.

And as long as the other end isn't trying the same cheap shortcut.

 

[Peter Jakacki]

Well Richard, you seem to have a lot to say to people on this subject,
and with such authority, even though it's actually on topic

The understanding of the spec is correct however the practice is
different. Please tell me which common RS-232 receivers *require* a
negative threshold (I know they exist, but where are they used). The
reality is that the receivers just turn-on at around 1.5V. If you drive
them with less than this they are turned-off (output high).

So, for short runs, without serious noise or ground-loop problems the 0
to +5V method will always work, not intermittently as you state, as if
the thresholds actually vary constantly. Of course this is not within
the EIA-232 spec born from it's archaic requirements, but it is within
the chip spec and actual system parameters.

BTW, I design production equipment with MAX232's but that is mainly for
customer comfort so I can tell them it is "compliant". I don't get
confused with the difference between compliance to a specification vs
hardware design. It is my job to understand the hardware without being
bamboozled by the hype. Remember Y2K?

*Peter*

 

[Spehro Pefhany]

I did not know that. Please provide a reference to a EUA-232 receiver with
such a specification.

Two common types-- the thresholds are listed on the data sheets:

http://www.ee.washington.edu/stores/DataSheets/linear/mc1489.pdf
http://focus.ti.com/lit/ds/symlink/max232.pdf


Best regards,
Spehro Pefhany

 

[budgie]

Don't count on it. However, many poorly-designed, non-spec-compliant
interfaces will work with those voltages.

Richard you are missing the point of the EIA spec. It mandates that compliant
receivers must respond to voltages >+3V as high and to voltages below -3V as
low. Any receiver with a threshold between these points will be compliant.

Sometimes.

Intermittently.

No.

As an example, have a look at the input circuit of the NatSemi:

http://www.national.com/ds/DS/DS1489.pdf

Care to tell us all that it is non-compliant?

It has a positive threshold, with feedback. Care to tell us all that input
transition from +5V to Vce(Sat) above ground (or vice versa) won't switch it?

 

[Mac]

Don't count on it. However, many poorly-designed, non-spec-compliant
interfaces will work with those voltages.

Sometimes.

Intermittently.

Many well-designed ones will work, too. It's just not guaranteed by the
spec to work.

--Mac