Post by thread:[PIC] 16F914 using MCLR pin as input

Like many PIC's, the 16F914 that I am using can implement the mclr/vpp pin to be a digital input as well. Typically I'd shy away from using it, but im forced as I need ONE more input... So, programming it in the config word is easy, and I assume that defining it as the alternate mode, that MCLR works in conjunction with the powerup timer to reset the device on power up. But what about the ICD....does it but 12V on this pin during programming still? I assume it does as the ICD doesnt know anything about that. So, I believe I need to isolate the digital input connection from the VPP signal of the ICD during programming to protect the chip that is driving it. Comments? --------------------------------- Yahoo! Shopping Find Great Deals on Holiday Gifts at Yahoo! Shopping

On December 12, 2005 10:50 pm, alan smith wrote: > Like many PIC's, the 16F914 that I am using can implement the > mclr/vpp pin to be a digital input as well. > > Typically I'd shy away from using it, but im forced as I need ONE > more input... > > So, programming it in the config word is easy, and I assume that > defining it as the alternate mode, that MCLR works in conjunction > with the powerup timer to reset the device on power up. > > But what about the ICD....does it but 12V on this pin during > programming still? I assume it does as the ICD doesnt know anything > about that. > > So, I believe I need to isolate the digital input connection from > the VPP signal of the ICD during programming to protect the chip that > is driving it. Comments? Sounds like you have it fairly well figured-out, since it's the same dillema with tiny chips such as the 12Fxxx series. Yes, expect hi-voltage required to program the chip, which means watch-out what you connect to this pin. Since it's an input-only, you might want to look at trading one of the other pins with this one, for example, suppose you have pushbuttons going to another pin, how about switching it with this one, then all you really have is a pull-up resistor & pushbutton affected by 12v on this pin, instead of a digital circuit which like you say, needs to be protected. If you can trade another pin's function with this one, great, but if you have no luck there, then you'll need to protect the external circuit one way or another.

Besides on the potential I/O conflicts, there is also the issue with Vpp-before-Vdd. I know that the mentioned 12F's have this "problem", don't know about the 16F914. Some other chips seems to not be affected by this, 16F688 and/or F88, if I'm not wrong... This is generaly a problem with ICSP where the programmer can't control Vdd. Anyway, probably better check it up. Regards, Jan-Erik.

On Mon, 12 Dec 2005 22:50:57 -0800 (PST), you wrote: >Like many PIC's, the 16F914 that I am using can implement the mclr/vpp pin to be a digital input as well. > > Typically I'd shy away from using it, but im forced as I need ONE more input... > > So, programming it in the config word is easy, and I assume that defining it as the alternate mode, that MCLR works in conjunction with the powerup timer to reset the device on power up. > > But what about the ICD....does it but 12V on this pin during programming still? I assume it does as the ICD doesnt know anything about that. > > So, I believe I need to isolate the digital input connection from the VPP signal of the ICD during programming to protect the chip that is driving it. Comments? Yes, but as it's input only, you can just put a resistor (e.g. 10K) in series with the pin. to isolate your circuitry. If the signal into the pin changes frequently, locate the resistor near the pin to minimise capacitance.

alan smith wrote: > But what about the ICD....does it but 12V on this pin during > programming still? I assume it does as the ICD doesnt know anything > about that. > > So, I believe I need to isolate the digital input connection from the > VPP signal of the ICD during programming to protect the chip that is > driving it. Comments? You should expect programmers to drive this line to 13V, and you have to design your circuit to tolerate that and not interfere with it either. Similarly, your circuit must allow PGC and PGD to be driven between 0 and the PIC Vdd even though the rest of the circuit may not be powered. The only way to avoid 13V on MCLR is to use low voltage programming. That has its problems and requires a dedicated pin, so that's not a solution for you anyway. For more info on circuit design for ICSP, see my writeup at http://www.embedinc.com/picprg/icsp.htm. ****************************************************************** Embed Inc, Littleton Massachusetts, (978) 742-9014. #1 PIC consultant in 2004 program year. http://www.embedinc.com/products

That might be the best solution. Boy I hate it when I find these right after I send the board out to be made.... But I am using PORTB for switch inputs (external to the board via a header), intending on using the built in pullups, so swapping this signal for a switch....could work and since there already is a pullup on this pin for MCLR... But of course, I have the problem that you either enable all the pins to have weak pullups or not. Sure would be nice if there was a mask register for this so you could choose which ones, else need the external pups. Otherwise, either a series resistor or just a jumper block for now could do the trick. Jose Da Silva <spam_OUTDigitalTakeThisOuTJoesCat.com> wrote: Since it's an input-only, you might want to look at trading one of the other pins with this one, for example, suppose you have pushbuttons going to another pin, how about switching it with this one, then all you really have is a pull-up resistor & pushbutton affected by 12v on this pin, instead of a digital circuit which like you say, needs to be protected. If you can trade another pin's function with this one, great, but if you have no luck there, then you'll need to protect the external circuit one way or another.

>> So, I believe I need to isolate the digital input connection from >> the VPP signal of the ICD during programming to protect the > chip that >> is driving it. Comments? > > Sounds like you have it fairly well figured-out, since it's the same > dillema with tiny chips such as the 12Fxxx series. > Yes, expect hi-voltage required to program the chip, which means > watch-out what you connect to this pin. OR: forget ICSP. At least for production: for development you could consider a development rig with one (or maybe more) switches (or even relais). Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

yes for production, isp will be dropped. Its just for development. Wouter van Ooijen <.....wouterKILLspam@spam@voti.nl> wrote: >> So, I believe I need to isolate the digital input connection from >> the VPP signal of the ICD during programming to protect the > chip that >> is driving it. Comments? > > Sounds like you have it fairly well figured-out, since it's the same > dillema with tiny chips such as the 12Fxxx series. > Yes, expect hi-voltage required to program the chip, which means > watch-out what you connect to this pin. OR: forget ICSP. At least for production: for development you could consider a development rig with one (or maybe more) switches (or even relais). Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu