[sdiy] Microchip PIC 5 x 4 matrix

[rsdio at audiobanshee.com]

Thank you for pointing that out, Roman.

Figure 19-18 on page 201 show Voh dropping below 3.5 V, and even as low as 1.5 V, when delivering 25 mA at normal room temperature. This is for the PIC16F88 with 5 V supply. It's even worse when running from a 3 V supply, where Figure 19-19 shows that the PIC cannot even deliver 15 mA without the port shutting down completely (i.e. it puts out 0 V).

Brian


On Oct 7, 2015, at 1:59 PM, Roman <modular at go2.pl> wrote:
> I'm not assuming one 100 ohm is used for all 5 LEDs, that would be terrible. I'm talking about bare PIC pin capability to drive that current to GND. 25mA is only marketing talk, look at what output voltage it is achieved.
> 
> Dnia 7 października 2015 22:43 rsdio at audiobanshee.com napisał(a):
>> On Oct 7, 2015, at 1:28 PM, Roman <modular at go2.pl> wrote:
>> > If you want to drive 5 LEDs in a single row from one RBx pin pulled low, that's about 15mA, and definitely RBx will not be at 0V anymore then, but maybe 0.5V or more. That may result in varying brightness depending how many LEDs are on in a row. 5 transistors will fix that. That is irrelevant of course if you plan to light only one of 20 at a time.
>> If there is a separate 100 ohm resistor for each LED, then the brightness will not vary, provided that the total current does not exceed the 25 mA capability of the PIC pin driving the column. Remember that each matrix intersection is in parallel with the others. That's partly why I suggested moving the 100 ohm resistor to the circuit path that is exclusive to the LED, rather than the path that is shared with the button, but it should work either way.
>> 
>> If you need to exceed 25 mA total for any column, then a transistor with a higher current capacity will solve that.
>> 
>> Brian