Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
11/27/07 17:55
Read: times


 
#147457 - Broken COM port suggestions
Responding to: ???'s previous message
While current isolation between equipment isn't a bad idea it isn't necessary in this case. I'm going to lean towards the fellow who suggested testing the ground potential between the PC and and the proto board. I have been using an AT89C52IC2 44PLCC on a protoboard for several years. The circuit contains all the necessary Communication circuitry I need to develope programs that operate all kinds of external devices. I use the internal RS232 port as well as the SPI and I2C ports. 7 ports in all; 4 RS232 (one is the internal port and the others are on the SPI ports using UARTs), 2 RS485's, and 1 I2C port. The port hooked to the internal RS232 port is plugged directly to my PC for On-Board-Programming purposes. The proto-platform is sacrosanct and has only been modified for newer devices a couple of times. It has been quite dependable and stalwart...The point is I had to solve this ground potential difference my self. Next, another person noticed that you don't have a bypass cap on the power pin of the Max device. You should always have one on those. Could be the main reason the Max232 device heats up. I noticed that you use 1uF Tants, on the charge pump pins, instead of the 0.1uF's recommended by Maxim. I have had to on other devices also, but, be careful that you aren't charging much higher than 10V. You may need to reduce the value down to 0.1uF. Now, Proto boards are notoriously noisy (basically they resemble "fuzzy antennas") and the test leads that hook it up to power collect noise, so, get a couple of 400uF 25V, or larger, E'Lytics and attach them to the power strips of the proto board. Also connecting the ground strips of each column together near the circuit will help clean things up a bit. Change the Max232 it's probably broken. Even if it isn't, start with a fresh circuit. You should also install TVS diodes, such as the SM712, on the Tx and Rx lines going off board. The amount of energy you pick up on the cable can be surprisingly large. Another fellow suggested grounding the case of the DSUB connector to pin 5, good advice, both ends. I have a wire that connects the PC case to the Proto board's power supply because I use a 3 pin friction-lock molex connector on my proto board instead of the DSUB just to save space and because it plugs directly to the board. I only use pins 2, 3, and 5 anyway. If your PC com ports are still operable, and they probably are, this circuit won't harm them. After that only your software will be holding you back. Remember, Ground, Bypass, Diodes, clean power on proto board. Good luck.

List of 60 messages in thread
TopicAuthorDate
PC COM Ports Damaging !!!            01/01/70 00:00      
   Earth loops?            01/01/70 00:00      
      Why Now and this circuit ?            01/01/70 00:00      
   Damage?            01/01/70 00:00      
      I am Sure ...            01/01/70 00:00      
   Common mode            01/01/70 00:00      
      You haven't mentioned            01/01/70 00:00      
   do you have the shell of DB9 grounded?            01/01/70 00:00      
   Serial port damage            01/01/70 00:00      
      for serial port damage            01/01/70 00:00      
         No            01/01/70 00:00      
      No, that can't be right!            01/01/70 00:00      
         re:3.3V            01/01/70 00:00      
            are you sure about that, Andy?            01/01/70 00:00      
   Are You SURE?            01/01/70 00:00      
      I'm sure Dear ...            01/01/70 00:00      
         not to be argumentative ... but ...            01/01/70 00:00      
   check it            01/01/70 00:00      
      Checks            01/01/70 00:00      
         We haven't seen how the DB-9 is wired            01/01/70 00:00      
            It\'s probably not a DB-9!            01/01/70 00:00      
               sure, but            01/01/70 00:00      
               Sloppy terminology            01/01/70 00:00      
   My Burned out ports            01/01/70 00:00      
      What?            01/01/70 00:00      
         What I mean is            01/01/70 00:00      
   a common misunderstanding is            01/01/70 00:00      
      So problem is in "modern" PCs?!            01/01/70 00:00      
      So problem is in "modern" PCs?!            01/01/70 00:00      
         maybe i'm wrong            01/01/70 00:00      
   try with opto coupler and +9 volt supply            01/01/70 00:00      
      Back when folks cared whether things work ...            01/01/70 00:00      
      Assuming He does not have one            01/01/70 00:00      
         There is a way to be sure.            01/01/70 00:00      
            Standard            01/01/70 00:00      
               Power Supply ...            01/01/70 00:00      
                  Dwhatever9 shell connection to GND...            01/01/70 00:00      
                     about that shell-to-shell connection            01/01/70 00:00      
                        generally...            01/01/70 00:00      
                  Truth emerges            01/01/70 00:00      
                     developed?            01/01/70 00:00      
                  that\'s the problem!            01/01/70 00:00      
         HUH?            01/01/70 00:00      
      why??            01/01/70 00:00      
         Broken COM port suggestions            01/01/70 00:00      
            he is not wrong - if            01/01/70 00:00      
               ...he is not wrong... only said watch it            01/01/70 00:00      
                  BULL!!!            01/01/70 00:00      
                     Bull?            01/01/70 00:00      
                        yes, genuine male cow manure            01/01/70 00:00      
                           Eric, you are a sad man            01/01/70 00:00      
                           Amateurs and pros            01/01/70 00:00      
                              of course I do            01/01/70 00:00      
                     Well, it has to do...            01/01/70 00:00      
                        Kai, I now you know this stuff ... but            01/01/70 00:00      
                           Different CMOS design?            01/01/70 00:00      
   that's right, when you can not argue tha facts ...            01/01/70 00:00      
      My Dear Friends ...            01/01/70 00:00      
         look in the datasheet            01/01/70 00:00      
            Y E S . . .            01/01/70 00:00      

Back to Subject List