Post by thread:16F84, surface mount, ICSP...

Hello all, I have decided in my application to use half a dozen 16F84's to interface to a bunch of components. However, space is very limited, so I want to use the surface-mount version of the chip. I plan on programming them in circuit, and will provide a four-pin header for each chip on the board. I plan on using RB6 & RB7 only for programming, and so will not need to isolate them from the circuit (since they will only go to the pins on the header). The question is, how do I isolate Vdd? I was thinking along the lines of running a trace from the Vdd pin directly to the power header pin, and to the rest of the circuit through a diode. The diode would be mounted to allow current to pass into the chip from the circuit power supply, but not from the chip to the rest of the circuit when powered from the programmer. Presumably, since the voltage drop would be about 0.6 volts, the power running into the PIC from the 5 volt circuit supply would end up being 4.4 volts, which is well within the limits. Will this work? Is there a better way? Thanks in advance... Later, Jon -------------------------------------------------------------- Jon Hylands spam_OUTJonTakeThisOuThuv.com http://www.huv.com/jon Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com

At 05:32 AM 28/04/98 GMT, you wrote: {Quote hidden}>Hello all, > >I have decided in my application to use half a dozen 16F84's to interface >to a bunch of components. > >However, space is very limited, so I want to use the surface-mount version >of the chip. I plan on programming them in circuit, and will provide a >four-pin header for each chip on the board. I plan on using RB6 & RB7 only >for programming, and so will not need to isolate them from the circuit >(since they will only go to the pins on the header). > >The question is, how do I isolate Vdd? I was thinking along the lines of >running a trace from the Vdd pin directly to the power header pin, and to >the rest of the circuit through a diode. The diode would be mounted to >allow current to pass into the chip from the circuit power supply, but not >from the chip to the rest of the circuit when powered from the programmer. >Presumably, since the voltage drop would be about 0.6 volts, the power >running into the PIC from the 5 volt circuit supply would end up being 4.4 >volts, which is well within the limits. > >Will this work? Is there a better way? > >Thanks in advance... > >Later, >Jon > >-------------------------------------------------------------- > Jon Hylands .....JonKILLspam@spam@huv.com http://www.huv.com/jon > > Project: Micro Seeker (Micro Autonomous Underwater Vehicle) > http://www.huv.com > > I don't understand the need to isolate VDD. Is the circuit in such a way that there be no pull ups or downs from pins being driven by the PIC, remember that the PIC should be in a RESET type state (All pins as inputs) during the programming cycle. Cone to think of it, you have control of the reset line to all the PICs, if this is the case, then you should be able to hold the PICs not bing programmed in RESET, and place RB6 and RB7 in //. You do indicate a 4 pin header for each chip, so from this can we assume that the reset lines are common? If you use the above, then control and progamming should be simple. However! A seperate header for each will speed up development as a stand alone programmer could be used. Hope that this helps Dennis

Jon, VDD is not the programming voltage pin; MCLR is, a 5 pin connector is required. Regards, Graham Daniel Jon Hylands wrote: {Quote hidden}> > Hello all, > > I have decided in my application to use half a dozen 16F84's to interface > to a bunch of components. > > However, space is very limited, so I want to use the surface-mount version > of the chip. I plan on programming them in circuit, and will provide a > four-pin header for each chip on the board. I plan on using RB6 & RB7 only > for programming, and so will not need to isolate them from the circuit > (since they will only go to the pins on the header). > > The question is, how do I isolate Vdd? I was thinking along the lines of > running a trace from the Vdd pin directly to the power header pin, and to > the rest of the circuit through a diode. The diode would be mounted to > allow current to pass into the chip from the circuit power supply, but not > from the chip to the rest of the circuit when powered from the programmer. > Presumably, since the voltage drop would be about 0.6 volts, the power > running into the PIC from the 5 volt circuit supply would end up being 4.4 > volts, which is well within the limits. > > Will this work? Is there a better way? > > Thanks in advance... > > Later, > Jon > > -------------------------------------------------------------- > Jon Hylands JonKILLspamhuv.com http://www.huv.com/jon > > Project: Micro Seeker (Micro Autonomous Underwater Vehicle) > http://www.huv.com

Hi Jon (Jon Hylands), in <.....354667ad.11930787KILLspam.....mail.huv.com> on Apr 28 you wrote: > The question is, how do I isolate Vdd? I was thinking along the lines of > running a trace from the Vdd pin directly to the power header pin, and to > the rest of the circuit through a diode. The diode would be mounted to > allow current to pass into the chip from the circuit power supply, but not > from the chip to the rest of the circuit when powered from the programmer. > Presumably, since the voltage drop would be about 0.6 volts, the power > running into the PIC from the 5 volt circuit supply would end up being 4.4 > volts, which is well within the limits. > > Will this work? Is there a better way? I don't think this is a good idea. You wrote you have several PICs interfacing to lots of peripherials. When these peripherialy are powered at 5.0V and output strong highs to the PIC, the PIC will be far beyond spec (max input high voltage = VDD, see DC specs). Probably it'll work, as the protection diodes are 0.7V, too. But it's definately not a good design. Better: Don't connect the 5V supply to the ISP header at all. Require the _circuit_ to supply power to the PICs also during programming. Connect MCLR and RB6 and RB7 and GND to the ISP header. Have all VDDs at 5V as normal in your circuit. Tie each MCLR to its own PICs' VDD with a 5K1 resistor (as long as you don't have a power supply monitor chip in your circuit).

On Wed, 29 Apr 1998 11:07:49 CET, Marc Heuler <EraseMEmarcspam_OUTTakeThisOuTAARGH.MAYN.DE> wrote: > I don't think this is a good idea. Right, hadn't considered that. Thanks. > Connect MCLR and RB6 and RB7 and GND to the ISP header. Have all VDDs at > 5V as normal in your circuit. Tie each MCLR to its own PICs' VDD with a > 5K1 resistor (as long as you don't have a power supply monitor chip in your > circuit). Hmmm, is that going to be okay when the programmer pumps 13 volts through MCLR? I guess that's what the 5K resistor is for... Thanks for your input, I appreciate it. Later, Jon -------------------------------------------------------------- Jon Hylands Jonspam_OUThuv.com http://www.huv.com/jon Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com

Hi I have used a 5pin header for in circuit programming . I have used the blocking diode configuration for all five programming pins and it seems to work ok. I only program when the power to the target board is disconnected. ASH - UK - on 28 Apr 98, KILLspamg.daniel.invent.designKILLspamxtra.co.nz wrote... {Quote hidden}>Jon, >VDD is not the programming voltage pin; MCLR is, a 5 pin connector is >required. >Regards, >Graham Daniel > >Jon Hylands wrote: >> >> Hello all, >> >> I have decided in my application to use half a dozen 16F84's to interface >> to a bunch of components. >> >> However, space is very limited, so I want to use the surface-mount version >> of the chip. I plan on programming them in circuit, and will provide a >> four-pin header for each chip on the board. I plan on using RB6 & RB7 only >> for programming, and so will not need to isolate them from the circuit >> (since they will only go to the pins on the header). >> >> The question is, how do I isolate Vdd? I was thinking along the lines of >> running a trace from the Vdd pin directly to the power header pin, and to >> the rest of the circuit through a diode. The diode would be mounted to >> allow current to pass into the chip from the circuit power supply, but not >> from the chip to the rest of the circuit when powered from the programmer. >> Presumably, since the voltage drop would be about 0.6 volts, the power >> running into the PIC from the 5 volt circuit supply would end up being 4.4 >> volts, which is well within the limits. >> >> Will this work? Is there a better way? >> >> Thanks in advance... >> >> Later, >> Jon >> >> -------------------------------------------------------------- >> Jon Hylands RemoveMEJonTakeThisOuThuv.com http://www.huv.com/jon >> >> Project: Micro Seeker (Micro Autonomous Underwater Vehicle) >> http://www.huv.com > >

I am attempting the same thing, and am following the design at: http://www.ccc.nottingham.ac.uk/~cczsteve/pic/pic84.pdf and a similar design at: http://www.ebar.dtu.dk/~c888600/icsp.htm The diode from MCLR to Vcc (through the MCLR pullup) limits the voltage during programming when MCLR is 13v. On some dev. boards, the header has been implemented as an RJ-45 connector. {Quote hidden}> > On Wed, 29 Apr 1998 11:07:49 CET, Marc Heuler <spamBeGonemarcspamBeGoneAARGH.MAYN.DE> wrote: > > > I don't think this is a good idea. =20 > > Right, hadn't considered that. Thanks. > > > Connect MCLR and RB6 and RB7 and GND to the ISP header. Have all VDDs = > at > > 5V as normal in your circuit. Tie each MCLR to its own PICs' VDD with = > a > > 5K1 resistor (as long as you don't have a power supply monitor chip in = > your > > circuit). > > Hmmm, is that going to be okay when the programmer pumps 13 volts through > MCLR? I guess that's what the 5K resistor is for... > > Thanks for your input, I appreciate it. > > Later, > Jon > > -------------------------------------------------------------- > Jon Hylands TakeThisOuTJonEraseMEspam_OUThuv.com http://www.huv.com/jon > > Project: Micro Seeker (Micro Autonomous Underwater Vehicle) > http://www.huv.com > Jonathan Cline RemoveMEjclineTakeThisOuTvision.calpoly.edu

Hi Jon, I use in circuit programming only, even with DIP packaged PICs. I hate to extract/insert the chip milion times (16F84s are ideal for this, does not even need UV erase). You have to think about ICSP during the circuit desing though. I use following setup: VDD and GND connected directly to programming header, MCLR connected to header and thru 39K resistor to VDD, RB6 & RB7 connected to header. The last two pins can be used as outputs in my design, since my programmer tristates its clock and data lines when the PIC is not in programming mode (I am not familiar with other programmes, mine is a home-brewed one). There is a slide swith in the programmer for +5V, so VDD can be supplied either from the circuit or from the programmer. There are several drawbacks in this approach. Nothing can be connected to the MCLR pin, which does not withstand 13V, RB6 & 7 are limited to output use only, the circuits connected to them have to tolerate the pulses during programming (usually not a problem) and should not draw excess current (like 20 mA LED or optocoupler). You may also run into difficulties, when verifying the PIC at different voltages (my programmer does not support this). If your circuit can supply some 20-40 mA to the VDD pin during programming, you need not to connect VDD to the header at all. Connect only GND, RB6, RB7 and MCLR instead of VDD. Josef > However, space is very limited, so I want to use the surface-mount version > of the chip. I plan on programming them in circuit, and will provide a > four-pin header for each chip on the board. I plan on using RB6 & RB7 only > for programming, and so will not need to isolate them from the circuit > (since they will only go to the pins on the header). ====================================================================== Electronical devices for chemical laboratory, custom electonics design ---------------------------------------------------------------------- Snail Instruments Josef Hanzal Vojanova 615 phone/fax: +420-311-24433 266 01 Beroun e-mail: euroclassEraseME.....pha.pvtnet.cz Czech Republic URL: http://www.vitrum.cz/snail/ ======================================================================