Post by thread:AW: [PIC]: PIC "powered a little bit by USART"

Isaac, My level shifter is built with discretes and it is powered by the V24 of the PC, there is no separate power supply for it. No, the level shifter is NOT powered down, it constantly delivers +5V via a 3,3K resistor to my Rx pin of my PIC. Peter -----Ursprüngliche Nachricht----- Von: spam_OUTpiclist-bouncesTakeThisOuTmit.edu [.....piclist-bouncesKILLspam@spam@mit.edu] Im Auftrag von Isaac Marino Bavaresco Gesendet: Montag, 31. Mai 2010 14:57 An: Microcontroller discussion list - Public. Betreff: Re: [PIC]: PIC "powered a little bit by USART" Em 31/5/2010 09:38, Peter Feucht escreveu: {Quote hidden}> > Hello, all, > > During the last days I found, that a 16F628A doesn't reset while > connected to RS232. My test setup is a small board (self made) > containing only the PIC, some caps, 5V regulator (78L05), the reset > pin of the 628A is used as port (= internal reset used). The USART is > constantly connected to my PC via a level shifter (5V to +/-12V). > When I switch off power from my board the pic still runs in irregular > fashion, because it seems to be powered somehow by the still present > +5V idle potential of the Rx line. > I have to disconnect the USART connector, then the power supply, then > restart by applying the power supply first and then connecting USART > which also makes problems due to undefined input to the USART (framing > errors, buffer overrun) Does anybody have a trick on hand how I can > establish a reliable "power down state" while my USART is still > connected? > > Thanks for help and best regards > > Peter > Is your level shifter a MAX232-like device or something with discrete devices? When you switch off power to the board with the PIC, is your level shifter powered-down also? A schematic diagram would help. Regards, Isaac __________________________________________________ Fale com seus amigos de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com/

Em 31/5/2010 10:23, Peter Feucht escreveu: > Isaac, > > My level shifter is built with discretes and it is powered by the V24 of the > PC, there is no separate power supply for it. > > No, the level shifter is NOT powered down, it constantly delivers +5V via a > 3,3K resistor to my Rx pin of my PIC. > > Peter > It seems that your circuit is applying voltage to the PIC's RX pin, and the internal diodes are powering-up the PIC with this voltage. So you should change your circuit so it won't apply voltage to the PIC's RX pin when it is not powered-on. Perhaps using a common-emitter NPN transistor with a pull-up resistor to the PIC's VCC. Regards, Isaac __________________________________________________ Fale com seus amigos de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com/

On Mon, 31 May 2010 15:23:25 +0200 "Peter Feucht" <p.feuchtKILLspampalmed-medizintechnik.de> wrote: > No, the level shifter is NOT powered down, it constantly delivers +5V > via a 3,3K resistor to my Rx pin of my PIC. Well, that's the problem. The input protection diodes of the PIC will take the external +5V, and feed the PIC. Those diodes are not very good, so the PIC will probably have something like 4V left, accounting for the erratic behaviour. The 78L05 and other elements will be loading the +5 line too. You'll have to change the level shifter so its supply comes from regulator. John

part 0 44 bytes his is a multi-part message in MIME format. part 1 989 bytes content-type:text/plain; charset=ISO-8859-1 (decoded 7bit) Em 31/5/2010 11:11, Isaac Marino Bavaresco escreveu: {Quote hidden}> Em 31/5/2010 10:23, Peter Feucht escreveu: > >> Isaac, >> >> My level shifter is built with discretes and it is powered by the V24 of the >> PC, there is no separate power supply for it. >> >> No, the level shifter is NOT powered down, it constantly delivers +5V via a >> 3,3K resistor to my Rx pin of my PIC. >> >> Peter >> >> > It seems that your circuit is applying voltage to the PIC's RX pin, and > the internal diodes are powering-up the PIC with this voltage. > > So you should change your circuit so it won't apply voltage to the PIC's > RX pin when it is not powered-on. Perhaps using a common-emitter NPN > transistor with a pull-up resistor to the PIC's VCC. > > > Regards, > > Isaac > Try this circuit, it works well and I use it extensively. Regards, Isaac part 2 22801 bytes content-type:image/jpeg; (decode) part 3 35 bytes content-type:text/plain; charset="us-ascii" (decoded 7bit)

Em 31/5/2010 20:56, Olin Lathrop escreveu: > Isaac Marino Bavaresco wrote: > >> Try this circuit, it works well and I use it extensively. >> > "Well" for some cases. Note this is a inverting circuit, and as such is > meant to replace the RS-232 receiver, not to be added in line with its logic > output. Accordingly to his original post, it seems to be exactly what he needs. If it were a logic output, he would not need a level translator at all. {Quote hidden}> It is a simple threshold detecting inverter, with the input > switching threshold about 600mV. That means it will correctly detect > correct RS-232 signals, which are from at least -5V to +5V. However, it has > poor noise immunity since it does not make use of the rather large possible > hysteresis range that RS-232 deliberately allows for (-3V to +3V if I > remember right). > > The common MAX232 and similar chips also don't make use of what they could, > but are still better than a pure threshold at 600mV. > There are some MAXxxx devices that detect and signal invalid levels, but the majority just detects any negative voltage as '1' and any positive voltage above some threshold as '0', exactly as this circuit. > Basically, this circuit relies on the other side doing things right so that > you can get away with being sloppy. That will work a lot of the time, but > doesn't make it a good idea. > Works well most of time (not to say all the time). Perhaps it may fail for very long and noisy lines. It will surely work for a short connection directly to a PC serial port. I used this circuit already for networks with more than 10 slaves distributed along lines over 100m long at 19200bps. Isaac __________________________________________________ Fale com seus amigos de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com/

Gentlemen, By the way, are there no girls in this forum? My problem was easily solved. Thanks to various inputs from many people it seems, that my PIC is operated via the protection diodes. For I have this always recommended series resistor of 3,3K in my Rx line I simply shunted my PIC with a 220 ohm resistor, so in case of switching off the 5V power the injected voltage results in such a low operating voltage (about 0,3V) that the 628 resets fine. OK, it is a somehow dirty method, but it works for my test setup. Thanks to everbody! Peter