Just wondering if anyone has used either the 12C508 or 509 when its maxed out.
I'm wondering if it would be able to handle generating bit banged pwm
continuously using all
6 available pins. I've built my proto board using a 84a but thats real
overkill for what its doing.
If the 509 can handle this my cost drops way down. I'm also considering
using the Tiny12
and Tiny15's.
Anyway just curious if someone had used these and how they behaved when
worked hard.
We've used thousands of 12C508s and plan to use just as many 12C671s -
they are great little chips. Can't say that we've "MAXed them out
though". There are also a couple of new flavours coming - the 12F675
and the 12F679 which probably have features you do not need, but the
internal clocks are supposed to be more accurate.
Microchip says that the internal clock on the present 12C series has an
accuracy of +/- 10%, but we find better than +/- 3% - if that is of any
use to you.
> Anyway just curious if someone had used these and how they behaved when
> worked hard.
Ummm... one instruction per instruction cycle (or two cycles, depending),
whether it's a NOP or something else. Why would it behave differently
depending on what it's doing? I'd think that whether it will do
bit-banged PWM on six pins at once depends entirely on your code and the
requirements for PWM period, etc.
>We've used thousands of 12C508s and plan to use just as many 12C671s -
>they are great little chips. Can't say that we've "MAXed them out
>though". There are also a couple of new flavours coming - the 12F675
>and the 12F679 which probably have features you do not need, but the
>internal clocks are supposed to be more accurate.
If it will happily run away full tilt with everything working then it replaces
a bunch of ic's worth about 3x the 509. (this is good). I did the proto
board with
a 84A but that was real overkill and just wasn't being totally utilized.
>Microchip says that the internal clock on the present 12C series has an
>accuracy of +/- 10%, but we find better than +/- 3% - if that is of any
>use to you.
Actually it is. Its close enough I shouldn't have to worry about the timing
aspect of things. The reason I asked about it being able to handle running
flat out was the last time I did that was with a 8739 a few ice ages back
and it
complained rather loudly then farted out some smoke and stopped. It turned out
the output latches didn't like the amount of work and were giving up the
ghost.
We managed to knock the level down a bit and it worked fine.
> I've built my proto board using a 84a but thats real
> overkill for what its doing.
> If the 509 can handle this my cost drops way down. I'm also
> considering using the Tiny12 and Tiny15's.
Maybe consider the 12Fxxx? Flash, so it's easier to use, and price is
roughly equeal to 12c50x. And its a 14-bit core, with 8 level stack.
If the timing is important to you, then hopefully you will get some
more feedback from the list. I can only tell what we have experienced.
However from what I have learnt from the list, Microchip tends to be
quite conservative with their specs, so what I have seen probably does
make sense.
I believe those new flavours I mentioned will also run up to 10MHz or
20MHz, but I am not 100% certain, as I have not actually seen any in
our region yet. There are some people on the list who have actually got
their hands on them, but I can not remember who.
Up to my *#$& in crocodiles at the moment, so have no time to search
for you, sorry. :-(
> I believe those new flavours I mentioned will also run up to 10MHz or
> 20MHz, but I am not 100% certain, as I have not actually seen any in
> our region yet.
12fxxx will indeed run up to 20 MHz, but that uses 2 pins, and the guy
was PWM-ing on 6 pins so he has none to spare.
> Just wondering if anyone has used either the 12C508 or 509 when its maxed
out.
>
> I'm wondering if it would be able to handle generating bit banged pwm
> continuously using all
> 6 available pins.
No. The MCLR pin can not be an output. However, there is no reason it
can't do bit banged PWM on the remaining 5 output pins. It is just a matter
of speed and resolution. I've done a similar thing using only 3 outputs of
a 12C508A for producing the PWM drive pulses for a motor. The PIC did an
open loop slow start algorithm, then kept the motor running at a fixed speed
forever after.
*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
mitvma.mit.edu with SET PICList DIGEST in the body
