That probably would have been a good idea. I have added another leaded 1uF cap to the setup and now the supply noise is down to about .2V, which I think should be on the edge of being workable. However whenever a character is received from the RS232 side (from the PC) the supply dips about .6-.7 volts for about .2ms and then about .1ms later it dips again about half as far and for half as long. Coinciding with these (I assume - I am not set up to measure both simultaneously) there are two dips in the !INVALID output from the buffer which have approximately the same time scale but drop about 3.5V (of a ~5V Vcc) and are much more squared off than the supply noise which gradually dips and rises again (which could just be the effect of the capacitance I suppose). The !INVALID output is supposed to go low whenever the RS232 input has been between +/- 0.3V for more than 30us. This never seems to be the case but I suppose the supply noise could still be the problem since it has to sample such a small voltage range with respect to the still relatively large noise component. How much capacitance can a PIC pin practically drive without causing problems? I know the datasheet says 50pF but that is rated to keep the output rise/fall time within spec, right? I don't care about a slow rise time (I already poll the pin in my code to wait for it to come high and then wait some more just to be sure) I just don't want to toast the chip or cause any weird side effect. I have 3.3uF and 10uF electrolytics I could try in addition to the 1uF chip cap. Think this is a bad idea? Nick ----- Original Message ----- From: "Matt Pobursky" To: Sent: Thursday, January 08, 2004 6:59 PM Subject: Re: [PICLIST] [PIC:] RS232 Driver IC Supply Noise > On Thu, 8 Jan 2004 17:42:29 -0600, James Nick Sears wrote: > > Any ideas? I do have a pad for RC4 (currently unused) right next to a pad > > for the problematic pin. I could parallel the two pins with a short jumper > > wire to reduce the output impedance by around half and this should help, but > > it seems there should be a better solution that should make that > > unnecessary. Opinions? > > It's probably not convenient on your current PCB revision, but I'd add > a P-channel mosfet such as FDV304P (Fairchild) to switch the supply to > the RS232 driver chip. It's small (SOT-23) and inexpensive and will > give you a solid low impedance supply for your driver chip. A gate > pulldown resistor to make sure the FET is off during PIC reset might be > a good idea depending on the requirements of your design too. > > Matt Pobursky > Maximum Performance Systems > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics > -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics