Post by thread:[PICLIST] PIC 16F84 & serial port on a PC

Hello, I am designing a board that has to exchange some data between a PIC16F84 and a serial port on a PC, in both ways. This will be done for about 1 minute, every week. So, my question is : Do I have to use a TTL<->RS232 converter, or can I simply use resistors as it is explained sometimes on PIC web pages. Could you give me your advices ? Thanks a lot, Best regards, Matt. -- http://www.piclist.com hint: To leave the PICList spam_OUTpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

Me too. I think the general opinion is that it works for one-offs and small stuff, but it may not be wise to depend on it in a large-run commercial product. For the PIC receiving RS232 it seems to work quite well, and I've han no trouble at all. For sending, you'll get a non-compliant signal that *most* RS232 receivers will see but no guarantees. Dale On Fri, 13 Oct 2000, Leo wrote: > I have used the direct connection between PC (EIA RS232) and the PIC as > showed in PBP manual. It works very well. You just take attention on the > polarity. The TTL serial have logical level inverted. > > > Ciao > > > Leo > {Original Message removed}

If I were building something that: 1: I knew was going to be connected _only_ to a PC's serial port. 2: Was _very_ cost sensitive Then I might consider using the direct-TTL trick. Bob Ammerman RAm Systems (contract development of high performance, high function, low-level software) -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam@spam@mitvma.mit.edu

On Fri, 13 Oct 2000, Bob Ammerman wrote: > If I were building something that: > > 1: I knew was going to be connected _only_ to a PC's serial port. > > 2: Was _very_ cost sensitive > > Then I might consider using the direct-TTL trick. I dunno. I looked at the specs for the 148x and MAX232 type line receivers, and all of them I saw would see 0 and 3.5V as valid RS232 levels, at least as far as they were concerned. It surprised me that even the 1489(88? haven't used one in years) didn't need to see a negative voltage. I seem to recall the cutoff was 0.7V. Has anyone had an actual experience with something that would *not* work with the simple resistor interface, I wonder? I've had them talking to PCs, modems, and terminal servers. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam.....mitvma.mit.edu

>I dunno. I looked at the specs for the 148x and MAX232 type line >receivers, and all of them I saw would see 0 and 3.5V as valid RS232 >levels, at least as far as they were concerned. It surprised me that even >the 1489(88? haven't used one in years) didn't need to see a negative >voltage. I seem to recall the cutoff was 0.7V. Has anyone had an actual >experience with something that would *not* work with the simple resistor >interface, I wonder? I've had them talking to PCs, modems, and terminal >servers. The reason for this is a requirement in the RS232 spec where an open circuit input should behave as though it has a voltage more negative than -3V. The easiest way to do this is set the schmitt trigger threshold at a slightly positive voltage. For this reason chips like the 1489 make excellent TTL line receivers, especially as you can slug the frequency response with a capacitor to slow them down. Check the spec sheets for this function. it is also possible to adjust the threshold level of the 1489 using a resistor to a positive or negative voltage. If the resistor goes to a positive voltage then the threshold goes more negative. This is also covered in the data sheets. -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestspam_OUTTakeThisOuTmitvma.mit.edu

> I am designing a board that has to exchange some data between a PIC16F84 > and a serial port on a PC, in both ways. > This will be done for about 1 minute, every week. > So, my question is : > Do I have to use a TTL<->RS232 converter, or can I simply use resistors as > it is explained sometimes on PIC web pages. > Could you give me your advices ? You can get away with a resistor and diodes for a quick hack for data from PC to PIC, but the voltage has to go from 0 -> 5 to -12 -> +12 the other way around. Maxxim makes a whole line of RS-232 converters, including some that run just off of +5. I've used these many times with PICs. ***************************************************************** Olin Lathrop, embedded systems consultant in Devens Massachusetts (978) 772-3129, olinspam_OUTcognivis.com, http://www.cognivis.com -- http://www.piclist.com hint: To leave the PICList @spam@piclist-unsubscribe-requestKILLspammitvma.mit.edu

On Fri, 13 Oct 2000, Olin Lathrop wrote: > > I am designing a board that has to exchange some data between a PIC16F84 > > and a serial port on a PC, in both ways. > > This will be done for about 1 minute, every week. > > So, my question is : > > Do I have to use a TTL<->RS232 converter, or can I simply use resistors as > > it is explained sometimes on PIC web pages. > > Could you give me your advices ? > > You can get away with a resistor and diodes for a quick hack for data from > PC to PIC, but the voltage has to go from 0 -> 5 to -12 -> +12 the other way > around. Maxxim makes a whole line of RS-232 converters, including some that > run just off of +5. I've used these many times with PICs. The MAX series is a great way to get standards-compliant RS232 signal levels, and it's only one chip plus four resistors. If a couple of bucks and another square inch of board space won't kill your project, definitely use one. I can tell you, however, that the PC will receive data from the PIC without one. Just remember you need to send the data inverted, which means you can do it with a bit-banged interface but not with the PIC's internal UART. To get around the inverter thing, I have used a 74LS04 with high value series resistors. It worked (for me) both ways, but I certainly wouldn't recommend it. I just ran out of MAX232's and needed to test something quickly. I was surprised it worked. quite frankly. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: PICList Posts must start with ONE topic: "[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads

I can tell you, however, that the PC will receive data from the PIC without one. Just remember you need to send the data inverted, which means you can do it with a bit-banged interface but not with the PIC's internal UART. A long time ago, after observing that the SDK-86 requiring +5 and +/-12 used the +/-12V ONLY for the rs232 drivers, I connected the +12 supply to +5, and the -12V to GND, and it worked fine (talking to a CPM system, IIRC. (yes, that long ago)) Since then, I've had a very "liberal" view of rs232 requirements. :-) There have been a couple of ciruits published (here?) that do clever things when at least one side has "real" rs232 voltages and "typical" rs232 signal leads. Ie, rather than switch between +/-PWR, you can go between +5 and "whatever is on DTR (which is supposed to be -V, and I can damn well MAKE it be -V since I control the PC side software too.)" (This is slightly more clever, IMHO, than using the modem signal leads to create a "power supply" for a traditional level converter circuit.) BillW -- http://www.piclist.com hint: PICList Posts must start with ONE topic: "[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads

William Chops Westfield wrote: {Quote hidden}> I can tell you, however, that the PC will receive data from the > PIC without one. Just remember you need to send the data > inverted, which means you can do it with a bit-banged interface > but not with the PIC's internal UART. > > A long time ago, after observing that the SDK-86 requiring +5 and +/-12 > used the +/-12V ONLY for the rs232 drivers, I connected the +12 supply to > +5, and the -12V to GND, and it worked fine (talking to a CPM system, > IIRC. (yes, that long ago)) > > Since then, I've had a very "liberal" view of rs232 requirements. :-) > > There have been a couple of ciruits published (here?) that do clever things > when at least one side has "real" rs232 voltages and "typical" rs232 signal > leads. Ie, rather than switch between +/-PWR, you can go between +5 and > "whatever is on DTR (which is supposed to be -V, and I can damn well MAKE > it be -V since I control the PC side software too.)" (This is slightly > more clever, IMHO, than using the modem signal leads to create a "power > supply" for a traditional level converter circuit.) > > BillW You can use a diode off RxD (perhaps other pins as well, coming from PC) to an electrolytic's - side. i.e. Cathode on PC side, anode to Cap's - side, to generate -12V (or so.) And same trick to generate +12v off the same COM port. Or, harder, run a charge pump voltage tripler off some PIC pins to make +- 14V or so. Schottky's can be a good idea here if you need more voltage (Seen BAT85's sometimes used.) http://uq.net.au/~zzdkeena/AVR/DCKProgSchem.GIF for one example, can find others. Mark -- http://www.piclist.com hint: PICList Posts must start with ONE topic: "[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads

BillW wrote: ........ >There have been a couple of ciruits published (here?) that do clever things >when at least one side has "real" rs232 voltages and "typical" rs232 signal >leads. Ie, rather than switch between +/-PWR, you can go between +5 and >"whatever is on DTR (which is supposed to be -V, and I can damn well MAKE >it be -V since I control the PC side software too.)" (This is slightly >more clever, IMHO, than using the modem signal leads to create a "power >supply" for a traditional level converter circuit.) > I have an appnote describing a simple RS-232 ckt using 2 transistors that translates 0 & 5v to +5v and -12v levels. The latter is "pilfered" from the Rx line of the PC's port. This ckt costs < $1, and has the great advantage over a MAX232 of less $$$$, and also that it has 0 mA quiescent current requirements [actually transistor leakage] versus the MAX's 2+ mA. Personally, I think you need a -voltage for good reliability and generality. See an-ecd02.zip at: http://www.users.uswest.net/~oricom/appnote.htm - Dan Michaels Oricom Technologies =================== -- http://www.piclist.com hint: PICList Posts must start with ONE topic: "[PIC]:","[SX]:","[AVR]:" =uP ONLY! "[EE]:","[OT]:" =Other "[BUY]:","[AD]:" =Ads

Mark Willis wrote: >> >> See an-ecd02.zip at: >> >> www.users.uswest.net/~oricom/appnote.htm >> > >This URL redirects to http://www.users.qwest.net/~oricom/appnote.htm >which 404's from here. I think I'm using the exact same circuit and >mentioned part of it (2 diodes to the PC's serial output, one to the >cap, the other is on the other side of a resistor and prevents an NPN >transistor's base from being pulled below -0.7V, etc.) It's a pretty >good solution really, all in all. I've seen MOSFET implementations of >this, and ones with NPN and PNP transistors, prefer the PNP / NPN >implementations as they're cheaper for production <G> > Oops, sorry, try: http://www.users.uswest.net/~oricom/appnotes.htm Mark, you are correct that including a diode from the base of the Rx tansistor to gnd [anode to gnd] is helpful in preventing E-B reverse breakdown. I typically now use a 2:1 voltage divider in the base to prevent this. 10K Rx-to-base, 4.7K base-to-gnd. Guess I should update the appnote. - Dan Michaels Oricom Technologies =================== -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.