SW UART and a nightmare problem

Inventor-George · Joined: 28 Mar 2006 Posts: 74

I have two processors [on seperate boards] that need to communicate via a SW UART. I am using

enniom · Joined: 20 Oct 2009 Posts: 537

hello mr. inventor,

from the BASCOM Help:

AdrianJ · Joined: 16 Jan 2006 Posts: 2483 Location: Queensland

Lucky for you that Mark provided the Inverted function on a SW UART. And that is probably easiest for you. But of course you can easily invert an RS485 signal in hardware too, just swap the A and B wires somewhere on the transmission link.

W4GNS · Joined: 03 Dec 2009 Posts: 11 Location: (Virginia)

Inventor-George · Joined: 28 Mar 2006 Posts: 74

It isn't that the inverted didn't work.
Truth time:
I have used a SW UART many times in the past. I never needed to invcert the signal. So I had no reason to go to help for a function I was quite familiar with.

When I realized the problem, I poured over the device data sheets , but didn't look up help.
I am guilty of hitting the panic button before checking out all options. I considered it a one-in-a-million shot to put the problem up here.

Thank you guys!!!

AND THANK YOU MARK!!!!!!!! for thinking ahead.

AdrianJ · Joined: 16 Jan 2006 Posts: 2483 Location: Queensland

I know this is probably academic, since you already have hardware in production, and a way to solve the problem, but I wonder how you got into this situation in the first place.

The standard method of doing RS485 has a RS485 driver/receiver at each end of a transmission line. Those drivers do the level conversion between processor voltage levels and RS485 levels, and both 3 and 5 volt versions are available, depending on your processor levels. So you should not need yet another level conversion ( and its inversion !!??!! ) at either end.

Evert :-) · Joined: 18 Feb 2005 Posts: 2156

Take for example the LTC1480 as rs485 driver. It runs at 3.3V so you don't need a level convertor.

MWS · Joined: 22 Aug 2009 Posts: 2262

Evert :-) · Joined: 18 Feb 2005 Posts: 2156

Please keep politics and religion out of this forum.
The Mod has remove one word form 1) I will be ... Maxim - they deserve it.

For us in Holland it's no problem to say or read words like that, but in other parts of the world you can have a serious problem.

MWS · Joined: 22 Aug 2009 Posts: 2262

albertsm · Joined: 09 Apr 2004 Posts: 5921 Location: Holland

ADMIN, that is Mark in this case. I do not want this forum spoiled with off topic discussions that can be held somewhere else. So i either edit or remove a topic. A joke for one, can be offensive to others. Please let this rest or write me at admin@mcselec.com

MWS · Joined: 22 Aug 2009 Posts: 2262

Paulvk · Joined: 28 Jul 2006 Posts: 1257 Location: SYDNEY

Now with my work here with MAX3232s they are inverters I had a problem with a magazine project that inverted the serial output as it fed another micro directly so when I added a MAX3232 the output was inverted so I was getting rubbish on my PC trying to look at the output then after some logic probing I saw that it was inverted thankfully the software writer sent me a hex file with it not inverted so I did not have to change the hardware.

Regards Paul

AdrianJ · Joined: 16 Jan 2006 Posts: 2483 Location: Queensland

AFAIK all RS232 type drivers invert the logic state. That is intentional, and users who try to drive or receive raw RS232 signals on processor ports should be well aware of it. IMHO its very bad practice to do this, since the RS232 spec allows for at least +/- 12 volt signals on the lines. Not many processors would like that on their port lines. Of course if you choose to do this because your device will never ever be connected to a real RS232 line, then it does not matter.

In the case of RS485 signals, as the OP was discussing, the situation is entirely different. The transmission is on nominally +5 volt signal lines, but in differential pairs, so its trivial to swap the polarity, just swap the differential pair at one end. Matching each processor supply voltage with its driver at its own end of the line is totally irrelevant to the RS485 transmission line signals.

Personally I really like RS485, its so easy to debug in the field. Just hang an IR LED across the line pair, and point it at a receiver in a laptop or PDA running a terminal program, and you can see the traffic, fully isolated from the transmission lines. If you get garbage, just reverse the LED connection ( or get the baud rate and settings right ).

MWS · Joined: 22 Aug 2009 Posts: 2262