Post by thread:[PIC] battery and ICSP

I want to ICSP a PIC that runs on two batteries (3V nominal). ICSP requires 5V, which can be supplied by the progger. Any clever suggestions on how to combine the battery and the power from the progger? Using a diode is simple, but limits the useful battery life. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

How about a P-MOSFET ? One or a couple to avoid reverse currents. I use this approach often when i need "almost" zero drop. The drop is given by the Rds and Current through the MOSFET. regards, Sergio Sena On 21 March 2012 12:30, Wouter van Ooijen <spam_OUTwouterTakeThisOuTvoti.nl> wrote: {Quote hidden}> I want to ICSP a PIC that runs on two batteries (3V nominal). ICSP > requires 5V, which can be supplied by the progger. Any clever > suggestions on how to combine the battery and the power from the > progger? Using a diode is simple, but limits the useful battery life. > > -- > > Wouter van Ooijen > > -- ------------------------------------------- > Van Ooijen Technische Informatica: http://www.voti.nl > consultancy, development, PICmicro products > docent Hogeschool van Utrecht: http://www.voti.nl/hvu > C++ on uC blog: http://www.voti.nl/erblog > >

On 21/3/2012 1:50 PM, A. Sergio Sena wrote: > How about a P-MOSFET ? > One or a couple to avoid reverse currents. > > I use this approach often when i need "almost" zero drop. The drop is > given by the Rds and Current through the MOSFET. Can you give a diagram? -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

Something allong this :: http://dl.dropbox.com/u/13600361/BAT%2BMOS.png Please test on your side. It has worked for me so far. regards, Sergio Sena On 21 March 2012 13:07, Wouter van Ooijen <.....wouterKILLspam@spam@voti.nl> wrote: {Quote hidden}> On 21/3/2012 1:50 PM, A. Sergio Sena wrote: >> How about a P-MOSFET ? >> One or a couple to avoid reverse currents. >> >> I use this approach often when i need "almost" zero drop. The drop is >> given by the Rds and Current through the MOSFET. > > Can you give a diagram? > > -- > > Wouter van Ooijen > > -- ------------------------------------------- > Van Ooijen Technische Informatica: http://www.voti.nl > consultancy, development, PICmicro products > docent Hogeschool van Utrecht: http://www.voti.nl/hvu > C++ on uC blog: http://www.voti.nl/erblog > >

On 21 March 2012 13:07, Wouter van Ooijen <wouterKILLspamvoti.nl> wrote: > On 21/3/2012 1:50 PM, A. Sergio Sena wrote: > > How about a P-MOSFET ? > > One or a couple to avoid reverse currents. > > > > I use this approach often when i need "almost" zero drop. The drop is > > given by the Rds and Current through the MOSFET. > > Can you give a diagram? See also "ideal diode" such as http://www.linear.com/product/LTC4413 and related products. -- Dave Twee

> http://dl.dropbox.com/u/13600361/BAT%2BMOS.png thanks! As I see it this uses the FET to conduct in revers mode (D is positive with regards to S), correct? I don't see information (RDson figures) for this mode in the datasheets that I looked in. Is it safe to assume that the D and S behave symmetrical? R72 keeps the gate low when there is no +5V, which makes the FET conduct. When +5V is available, the gate is more positive than D or S, so the FET is open. But what is R68 for? It halves the gate voltage in the battery-only mode, which seems a bad idea? -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

> See also "ideal diode" such as http://www.linear.com/product/LTC4413 and > related products. DFN package :( but knowing the term "ideal diode" enables further googling! -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

>> http://dl.dropbox.com/u/13600361/BAT%2BMOS.png > > thanks! > > As I see it this uses the FET to conduct in revers mode (D is positive > with regards to S), correct? I don't see information (RDson figures) for > this mode in the datasheets that I looked in. Is it safe to assume that > the D and S behave symmetrical? Yes assume that for Rds. > > R72 keeps the gate low when there is no +5V, which makes the FET > conduct. When +5V is available, the gate is more positive than D or S, > so the FET is open. > > But what is R68 for? It halves the gate voltage in the battery-only > mode, which seems a bad idea? Picture that there is no power source other than battery. Taking out R68 and at initial power-up, Source would be 0.7V less than Drain. hence, Bat=3V, VDD=2.3V Immediately FET will close because Vgs beyond threshold. Hence VDD=wee-bit-less-BAT. R68 was there because of Vgs max voltages... anyway, copypaste from other circuits. It may be ommited if you like

Wouter van Ooijen <wouter <at> voti.nl> writes: > > I want to ICSP a PIC that runs on two batteries (3V nominal). ICSP > requires 5V, which can be supplied by the progger. Any clever > suggestions on how to combine the battery and the power from the > progger? Using a diode is simple, but limits the useful battery life. > If you need to do ICSP only, without debugging, maybe just put a jumper to disconnect battery for programming? Alternatively, can you use PIC that can be programmed at low Vpp? Pickit 3 supports Vpp down to 1.8V. Sergey Dryga http://beaglerobotics.com

> If you need to do ICSP only, without debugging, maybe just put a jumper to > disconnect battery for programming? I was thinking of doing this in a thingie that must be used by students. Trust them to do it wrong, and I don't like the idea of leaking (or exploding) batteries. The circuit show by Sergio has one problem: it won't work for PICs that need a power cycle to get into programming mode :( > Alternatively, can you use PIC that can be > programmed at low Vpp? Pickit 3 supports Vpp down to 1.8V. The problem (with the PICs I considered) is Vdd, not Vpp, especially the Vdd required for a chip erase. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

On 22 March 2012 06:29, Wouter van Ooijen <.....wouterKILLspam.....voti.nl> wrote: > > The circuit show by Sergio has one problem: it won't work for PICs that > need a power cycle to get into programming mode :( > How do you plan to make the power cycle? cutting VDD for the PIC to Reset ? -- A. Sérgio Sena http://www.antoniosergiosena.com

> How do you plan to make the power cycle? cutting VDD for the PIC to Reset ? I have no plan. The programmers that I know will cut Vdd and/or short Vdd to ground. But this will result in the battery kicking in, thus preventing the PIC from being power-cycled. Maybe I should reconsider the whole idea. A low-voltage-only PIC or an LPC1114 might be a better idea than trying to combine a 5V PIC with battery and ICSP. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

On 22 March 2012 09:24, Wouter van Ooijen <EraseMEwouterspam_OUTTakeThisOuTvoti.nl> wrote: >> How do you plan to make the power cycle? cutting VDD for the PIC to Reset ? > > I have no plan. The programmers that I know will cut Vdd and/or short > Vdd to ground. But this will result in the battery kicking in, thus > preventing the PIC from being power-cycled. > > Maybe I should reconsider the whole idea. A low-voltage-only PIC or an > LPC1114 might be a better idea than trying to combine a 5V PIC with > battery and ICSP. > You can addd another P-MOS in series with the VPP, just before the micro. Source input, Drain output, Gate pull-down 100k to GND. When needed to power down the micro, just apply Source voltage on the Gate. The solution may be overkill with all this components, but will guarantee a clean power-down->up. -- A. Sérgio Sena http://www.antoniosergiosena.com

> You can addd another P-MOS in series with the VPP, just before the micro. > Source input, Drain output, Gate pull-down 100k to GND. > When needed to power down the micro, just apply Source voltage on the Gate. > > The solution may be overkill with all this components, but will > guarantee a clean power-down->up. Can be done, but it would require a change to for instance a pickit2 or pickit3 programmer: the required signal is not available on the ICSP connector. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

Wouter van Ooijen <wouter <at> voti.nl> writes: > The circuit show by Sergio has one problem: it won't work for PICs that > need a power cycle to get into programming mode :( > Out of curiosity, what PIC you are using? The once I have used so far (16F and 18F) do not require to cycle power to go into programming mode. I do not recall microchip datasheets expressly state that requirement either, it would be good to know to be aware of. Thanks, Sergey Dryga http://beaglerobotics.com

On Thu, Mar 22, 2012 at 12:29 AM, Wouter van Ooijen <wouterspam_OUTvoti.nl> wrote: >> If you need to do ICSP only, without debugging, maybe just put a jumper to >> disconnect battery for programming? > > I was thinking of doing this in a thingie that must be used by students. > Trust them to do it wrong, and I don't like the idea of leaking (or > exploding) batteries. > Can you put the ICSP connector inside / adjacent to the battery compartment, to make it really inconvenient to program while the batteries are in place? Regards, Mark markrages@gmai

> Out of curiosity, what PIC you are using? The once I have used so far (16F and > 18F) do not require to cycle power to go into programming mode. I do not recall > microchip datasheets expressly state that requirement either, it would be good > to know to be aware of. I was thinking generally, but examples are the 12F629 and 16F630. This 'feature' is mostly found on chips that can configure their MCLR pin as input. It is described in the 'programming specifications' document, not in the datasheet. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

> Can you put the ICSP connector inside / adjacent to the battery > compartment, to make it really inconvenient to program while the > batteries are in place? For a newly constructed housing that would be a very good approach. But I want to use an existing (cheap!) toy gun. (The project would be firmware for (parts of) a laser tag game). -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu C++ on uC blog: http://www.voti.nl/erblog

Wouter van Ooijen <wouter <at> voti.nl> writes: >> Out of curiosity, what PIC you are using? The once I have used so far (16F and >> 18F) do not require to cycle power to go into programming mode. I do not recall >>microchip datasheets expressly state that requirement either, it would be good >> to know to be aware of. > > I was thinking generally, but examples are the 12F629 and 16F630. This > 'feature' is mostly found on chips that can configure their MCLR pin as > input. It is described in the 'programming specifications' document, not > in the datasheet. > Thanks, I learned something today :-) Sergey http://beaglerobotics.com

On Thu, Mar 22, 2012 at 12:51 PM, Sergey Dryga <@spam@sergeyKILLspamdryga.us> wrote: > > > Thanks, I learned something today :-) > > Sergey > http://beaglerobotics.com > > Me too!!! I love this list! Carey -- Carey Fisher Chief Technical Officer New Communications Solutions, LLC 678-999-3956 KILLspamcareyfisherKILLspamncsradio.co