modern RS232 ports

The last time I looked into this, a minimum of +- 2.5V was the levels for the outputs and all inputs should be able to handle +/- 25Vs BUt I wouldn't push it that high, I would stick to +/- 12V.

So you shouldn't have any problems detecting a - voltage, just prepare your inputs to handle higher volts..

A level converter chip would be the smartest move :)

Jamie

age

still

if it

Maxim has been making parts that detect if the port is connected for years. This is done with an extra receiver that has a trip point about a vth beneath ground. Otherwise, all real data receivers have positive trip points so ground is good enough.

Maxim added a "plus" mode to detect if there is current flow from the transmitter. Total overkill, but if you want to detect "cheaters" (i.e. just going to ground), that would be the technology to use.

Why piss off the customer with 232?

Lowest swing for RS-232 i have seen is 0V to +3V acceptable on inputs. But outputs in that system was +/- 15V; a different system was..... ... ... ...TTL.

still

it

Nice, except that sane people don't buy from Maxim.

It's a diagnostic/programming port, and RS232 is the easiest way to do it, given a bare-metal ARM processor. The customer seems to like the idea.

John

Probably so. I just tested one that I had around, and it makes +-6 with about a volt p-p of sawtooth ripple.

Too bad laptops don't have serial ports any more. The USB adapters are a nuisance, physically and the thing about the com port numbers jumping all over the place.

I guess I'll include a com port scanner in the PC software, to automatically find the port that the laser controller is hooked to. Nuisance.

John

On a sunny day (Sun, 20 Mar 2011 10:45:10 -0700) it happened John Larkin wrote in :

still

it

I just use the serial port interrupt. If serial port interrupt happens, then somebody is accessing the serial port. Period. You can do auto-baudrate on top of that.

still

if it

If someone plugs a USB-to-serial dongle into a PC, and then connects it to my laser controller, the PC application may not even know which COM port is the path to the laser. When the laser controller boots up, the easiest thing to do is look at static voltages on its RS232 port to decide whether to run in the diagnostic mode or boot and run the laser application.

After the PC application is started, it will probably have to scan the com ports to find the laser. Or something like that.

No reason to autobaud, since we own both ends.

John

On a sunny day (Sun, 20 Mar 2011 11:25:03 -0700) it happened John Larkin wrote in :

message

still

if it

That is a matter of design philosophy. In my PIC projects, all use RS232, booting is always into normal mode, but you can get test and 'diagnostics' by sending some serial command. And a menu by simply typing 'h'. I understand you do not want that laser to be activated automatically at boot up. Use a switch? Because it is absolutely possible, and this happened here, that they plug it into the wrong serial port on the wrong PC, and then it will go into diagnostics and do BAD things perhaps, if software allows, if it gets fed with the wrong baudrate and wrong commands from some other PC application.

Except that next week the low bid manufacturer might be different and the "adapter" may have the same name and package and completely different design.

Isn't the best strategy to grab two or three readily-available USB to RS232 "adapters" from Frys, test them, and write down which are known to work? Then if the customer wants to add to the list, ask them to communicate which ones work back to you so you can expand the list ( or have them have you test them as well ).

-- Les Cargill

message

still

if it

So use DTR/DSR, that's what they're for ;)

The CTS/RTS for hardware handshaking, if you're moving data, for commands too if the widget cannot keep up with the selected data rate. I dislike xon/xoff soft handshaking as it then requires full escaping of control chars if you want to transfer eight bit data.

As far as the RS-232 levels go, the spec wants a minimum +/- 3V at the receiver, with the ability to withstand much higher -- but too many sloppy hardware designs die when connected to -ve voltage on rs232 port unpowered. I think the mechanism is commonly that the on-chip protection diodes are biased on, apply power and the CMOS chip SCRs and shorts power supply until its dead :( Lack of latchup protection?

Over the years I've noticed diodes around the rs232 buffer chips, I think to stop this powerdown current flow simply with series diodes in power lines. No idea whether the problem is still there. I've not done RS232 for some time -- but it doesn't change. There's still a new protocol implementation for each instance ;)

I'm inclined to stay at logic level for serial, unless I'm feeding a PC or something expecting proper serial levels.

These days I think serial USB is more relevant as computers lose the old ports.

Agree.

Grant.

Big nuisance, release the port before disconnect to free the COM port number.

Grant.

A *VOLT*????

Yarg. This being said, it is amazing anything USB works at all.

Yes.

It is like having to write a short story to order dinner :(

-- Les Cargill

Identify them by part number. If that changes...

-- Les Cargill

What frequency, John?

It was about 60 KHz unloaded at the TXD pin, and increased to (as I recall) about 600K with a 1K load to ground. Must be a weird charge pump.

John

Actually, you can still buy laptops with serial ports in them, you just need select the proper model that has the upgradeable slots to have it installed..

Some models of the Panasonic Touch books have all the goodies in them, some dells can get fitted for them etc.. Myself, I use a Toshiba that hsa a serial port. Of course, these units are upper end laptops..

Jamie

Ah, yes, dammit. I now remember your post with that info.

Always about a volt with increasing load?

John

still

if it

Most of the equipment we use at work that has 232, has cables with jumpers in them so that the connected device knows if it should be responding to the port and if it should be allowed to be programmed.

That's the nice thing about RS232, it gives you inputs that can be used for other things and is a standard. Jamie