Searching \ for '[EE] NaPiOn IR sensors revisited' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/io/sensors.htm?key=sensor
Search entire site for: 'NaPiOn IR sensors revisited'.

Exact match. Not showing close matches.
PICList Thread
'[EE] NaPiOn IR sensors revisited'
2007\03\28@225058 by Debbie

flavicon
face
PICers ... I have a bit of a dilemma selecting an IR sensor.  They're NaPion
motion sensors that are made by Matsushita. They come in both digital and
analgue o/p: it's digital that I need. I've used them before and they're pretty
damn good.

The one's I'm talking about are here -->
http://pewa.panasonic.com/pcsd/product/sens/select_motion.html

The dilemma -->
They come in a standard version that works on 5V but its quiescent drain is
170uA to 300uA

However, they now offer an ultra-low standby current version that's rated at 46
to 60 uA quiescent.

That's MUCH better for a battery operated device. BUT, the ultra-low Iq version
only works on 3V.

Okaaaay, I have a 12V, 7.5Ah system and the PIC's (16F84) voltage is regulated
down to 5V using a 78LS05. Not sure what its standby drain is but must be
pretty low as the battery lasts for weeks.

The QUESTION: what ultra-low Iq 3V regulators come to mind?

If I go with the 3V sensor, I'd have to drop 2V off the 5V rail or 9V off the
12V rail. This when I'm pulling, hopefully, total standby current of <150uA
(inc. both regulators, the PIC in sleep mode and the IR sensor)

What would people say is the best way to go?

These sensors aren't all that cheap and I need about 6, so I probly want to get
it right first time. :(

Thanks in advance for advice.

Best - Debbie :)

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\28@225113 by Debbie

flavicon
face
PICers ... I have a bit of a dilemma selecting an IR sensor.  They're NaPion
motion sensors that are made by Matsushita. They come in both digital and
analgue o/p: it's digital that I need. I've used them before and they're pretty
damn good.

The one's I'm talking about are here -->
http://pewa.panasonic.com/pcsd/product/sens/select_motion.html

The dilemma -->
They come in a standard version that works on 5V but its quiescent drain is
170uA to 300uA

However, they now offer an ultra-low standby current version that's rated at 46
to 60 uA quiescent.

That's MUCH better for a battery operated device. BUT, the ultra-low Iq version
only works on 3V.

Okaaaay, I have a 12V, 7.5Ah system and the PIC's (16F84) voltage is regulated
down to 5V using a 78LS05. Not sure what its standby drain is but must be
pretty low as the battery lasts for weeks.

The QUESTION: what ultra-low Iq 3V regulators come to mind?

If I go with the 3V sensor, I'd have to drop 2V off the 5V rail or 9V off the
12V rail. This when I'm pulling, hopefully, total standby current of <150uA
(inc. both regulators, the PIC in sleep mode and the IR sensor)

What would people say is the best way to go?

These sensors aren't all that cheap and I need about 6, so I probly want to get
it right first time. :(

Thanks in advance for advice.

Best - Debbie :)

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\28@225118 by Debbie

flavicon
face
PICers ... I have a bit of a dilemma selecting an IR sensor.  They're NaPion
motion sensors that are made by Matsushita. They come in both digital and
analgue o/p: it's digital that I need. I've used them before and they're pretty
damn good.

The one's I'm talking about are here -->
http://pewa.panasonic.com/pcsd/product/sens/select_motion.html

The dilemma -->
They come in a standard version that works on 5V but its quiescent drain is
170uA to 300uA

However, they now offer an ultra-low standby current version that's rated at 46
to 60 uA quiescent.

That's MUCH better for a battery operated device. BUT, the ultra-low Iq version
only works on 3V.

Okaaaay, I have a 12V, 7.5Ah system and the PIC's (16F84) voltage is regulated
down to 5V using a 78LS05. Not sure what its standby drain is but must be
pretty low as the battery lasts for weeks.

The QUESTION: what ultra-low Iq 3V regulators come to mind?

If I go with the 3V sensor, I'd have to drop 2V off the 5V rail or 9V off the
12V rail. This when I'm pulling, hopefully, total standby current of <150uA
(inc. both regulators, the PIC in sleep mode and the IR sensor)

What would people say is the best way to go?

These sensors aren't all that cheap and I need about 6, so I probly want to get
it right first time. :(

Thanks in advance for advice.

Best - Debbie :)

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\28@225452 by Debbie

flavicon
face
PICers ... I have a bit of a dilemma selecting an IR sensor.  They're NaPion IR
sensors that are made by Matsushita. They come in both digital and analogue
o/p: it's digital that I need. I've used them before and they're pretty damn
good.

The one's I'm talking about are here -->
http://pewa.panasonic.com/pcsd/product/sens/select_motion.html

The dilemma -->
They come in a standard version that works on 5V but its quiescent drain is
170uA to 300uA

However, they now offer an ultra-low standby current version that's rated at 46
to 60 uA quiescent.

That's MUCH better for a battery operated device. BUT, the ultra-lowQ version
only works on 3V.

Okaaaay, I have a 12V, 7.5Ah system and the PIC's (16F84) voltage is regulated
down to 5V using a 78LS05. Not sure what its standby drain is but must be
pretty low as the battery lasts for weeks.

The QUESTION: what ultra-low Iq 3V regulators come to mind?

If I go with the 3V sensor, I'd have to drop 2V off the 5V rail or 9V off the
12V rail. This when I'm pulling, hopefully, total standby current of <150uA
(inc. both regulators, the PIC in sleep mode and the IR sensor)

What would people say is the best way to go?

These sensors aren't all that cheap and I need about 6, so I probly want to get
it right first time. :(

Thanks in advance for advice.

Best - Debbie :)

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\28@231602 by Debbie

flavicon
face

--- Debbie <spam_OUTcyberia429-piclistTakeThisOuTspamyahoo.com.au> wrote:
Please accept my apols for the dupes everyone. Yahoo's giving weird error
messages.
D.

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\29@004642 by Marcel Duchamp

picon face
Debbie wrote:
> PICers ... I have a bit of a dilemma selecting an IR sensor.  
>
> What would people say is the best way to go?

Well.... depends. What about other low current methods? Like turning on
for 10% and off for 90%? Or some such ratio?

This might let you reduce current by a similar percentage. This has
worked well in the past whilst searching for the dreaded Thylacoleo beast.

2007\03\29@011835 by Bob Blick

face picon face


Debbie wrote:

{Quote hidden}

Doesn't the 78LS05 consume about 3 milliamps? You could save some
current right there using a better regulator on the 5V side.

Cheerful regards,

Bob

2007\03\29@014037 by Robert Rolf

picon face
Bob Blick wrote:

{Quote hidden}

Just about ANYTHING is better than a 78L05 for Iq.
Look at the LM2931-5 LDO regulators.

I think they also have a 3.3V version.

If you have nothing for current, why not just use a series 2V zener or
3 silicon diodes to get you the needed drop.

And can you strobe the sensor power (power it off a PIC pin, with diodes)
to save even more power?

Robert

2007\03\29@041117 by Alan B. Pearce

face picon face
{Quote hidden}

My thought too. I've been playing with the Linear technology LT3470 switcher
to get 5V off a high voltage supply. With Debbie's case I would then use a
linear LDO to get the 3.3V from the 5V - but does the 16F84 have suitable
input thresholds to use the signal from a 3.3V device? It may be better to
use a later PIC that will operate at 3.3V and run the whole thing at 3.3V -
without knowing what else the PIC is attached to.

2007\03\29@050848 by Jinx

face picon face
> but does the 16F84 have suitable input thresholds to use the
> signal from a 3.3V device ?

According to the datasheet it will run from 2V to 6V and almost
all I/O thresholds are relative to Vcc. If it were run slowly, eg
32kHz, that would help with consumption. I don't think a newer
one, eg F88, is any better in LP mode, still around 15-20uA. If
the PIR is generating an IRQ then the PIC will be asleep anyway,
and typically new nW parts are much better in that department,
less than a uA. It would make sense to go for one Vcc rather
than waste power having two

2007\03\29@052850 by Daniel Dourneau

flavicon
face
I would suggest a low drop out regulator from your PIC 5 Volts supply.

You can look at on semiconductors or linear technology websites. On
semiconductors has a selection tool that is very useful.

Selon Debbie <.....cyberia429-piclistKILLspamspam@spam@yahoo.com.au>:

{Quote hidden}

> -

2007\03\29@074609 by Dario Greggio

face picon face
Robert Rolf wrote:

> If you have nothing for current, why not just use a series 2V zener or
> 3 silicon diodes to get you the needed drop.

I'd enforce this too.

--
Ciao, Dario

2007\03\29@104121 by Timothy Weber

face picon face
Debbie wrote:
> The QUESTION: what ultra-low Iq 3V regulators come to mind?

I've been using Microchip's MCP1702 for a battery-powered 3V project.
Quiescent 2 uA, output up to 250 mA.  And you can get samples in TO-92
for prototyping.
--
Timothy J. Weber
http://timothyweber.org

2007\03\29@150309 by Debbie

flavicon
face

--- Marcel Duchamp <marcel.duchampspamKILLspamsbcglobal.net> wrote:

> Debbie wrote:
> > PICers ... I have a bit of a dilemma selecting an IR sensor.  
> >
> > What would people say is the best way to go?
>
> Well.... depends. What about other low current methods? Like turning on
> for 10% and off for 90%? Or some such ratio?

Wake it up from sleep periodically? Yeah, that's an idea.

>
> This might let you reduce current by a similar percentage. This has
> worked well in the past whilst searching for the dreaded Thylacoleo beast.

Heh, never know what's out there in the dark, hey? :)

Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\29@150901 by Debbie

flavicon
face

--- Jinx <.....joecolquittKILLspamspam.....clear.net.nz> wrote:

> > but does the 16F84 have suitable input thresholds to use the
> > signal from a 3.3V device ?
>
> According to the datasheet it will run from 2V to 6V and almost
> all I/O thresholds are relative to Vcc. If it were run slowly, eg
> 32kHz, that would help with consumption. I don't think a newer
> one, eg F88, is any better in LP mode, still around 15-20uA. If
> the PIR is generating an IRQ then the PIC will be asleep anyway,
> and typically new nW parts are much better in that department,
> less than a uA. It would make sense to go for one Vcc rather
> than waste power having two

Yep, the '64 would do. The system is activated by an interrupt on RBo, which is
edge triggered. So i guess absolute levels aren't decisive. The Napion digital
are excellent cuz you have the pulses appear @ RB0 already preconditioned.

>I've been using Microchip's MCP1702 for a battery-powered 3V project.
>Quiescent 2 uA, output up to 250 mA.  And you can get samples in TO-92
>for prototyping.

Gentlemen, thanks for the tips. I'm thinking 3V is the way to go. Definitely
don't like the idea of 2 regulators. I guess my trouble is I'm somewhat hung up
on 5V-TTL as "the" logic level.

I'll stick with the trusty old 16F84 @ 3V for now cuz the CCT design is tried
and true. I'm using periph chips for RS232 voltage conversion and
light/dark/battery voltage sensing but the PIC switches them on/off rail bmo a
relay as per needed. At the moment I'm bit banging the gizmo's serial IO but am
developing some I2C code using the 16F877's modules.

QUESTION: if you guys were considering a 3V PIC with maybe I2C and SPI modules,
and seriously low Iq sleep mode, which one would you go for?

D.



Send instant messages to your online friends http://au.messenger.yahoo.com

2007\03\29@155656 by Timothy Weber

face picon face
Debbie wrote:
> Gentlemen, thanks for the tips. I'm thinking 3V is the way to go. Definitely
> don't like the idea of 2 regulators. I guess my trouble is I'm somewhat hung up
> on 5V-TTL as "the" logic level.

I was dragged into it myself by a part that's not 5V tolerant.  Not too
bad.  You give up ICSP, though.  (I mean, I'm sure there's some
brilliant way around it, but I just grab the ZIF socket.)

> QUESTION: if you guys were considering a 3V PIC with maybe I2C and SPI modules,
> and seriously low Iq sleep mode, which one would you go for?

16F916 is my 'usual' now, or 917 for more pins - though there's now one
that's cheaper and has PWM instead of the LCD interface (which I've
never used), don't know the number off hand.
--
Timothy J. Weber
http://timothyweber.org

2007\03\29@160645 by Herbert Graf

flavicon
face
On Thu, 2007-03-29 at 15:56 -0400, Timothy Weber wrote:
> Debbie wrote:
> > Gentlemen, thanks for the tips. I'm thinking 3V is the way to go. Definitely
> > don't like the idea of 2 regulators. I guess my trouble is I'm somewhat hung up
> > on 5V-TTL as "the" logic level.
>
> I was dragged into it myself by a part that's not 5V tolerant.  Not too
> bad.  You give up ICSP, though.  

I apologize in advance if you're not talking about ICSP with PICs, but
if you're inferring ICSP with PICs doesn't happen below 5V, that's
simply wrong.

The only reason you'd need 5V for ICSP is if you are code protecting
your part and need to do a bulk erase. Otherwise, any ICSP programmer
that allows for the target to supply power can theoretically support
ICSP below 5V. I know for a fact the ICD2 works perfectly fine
programming (and debugging) a part running at 3.3V without issues.

TTYL

2007\03\29@170939 by Timothy Weber

face picon face
Herbert Graf wrote:
> I apologize in advance if you're not talking about ICSP with PICs, but
> if you're inferring ICSP with PICs doesn't happen below 5V, that's
> simply wrong.
>
> The only reason you'd need 5V for ICSP is if you are code protecting
> your part and need to do a bulk erase. Otherwise, any ICSP programmer
> that allows for the target to supply power can theoretically support
> ICSP below 5V. I know for a fact the ICD2 works perfectly fine
> programming (and debugging) a part running at 3.3V without issues.

Hey, that's great!  I guess there must be some other reason I've never
been able to make it work in the context I have that uses 3V (PICkit 2,
various PICs, XBees, not much else).  I'll look into it in more detail.
--
Timothy J. Weber
http://timothyweber.org

2007\03\29@172643 by Herbert Graf

flavicon
face
On Thu, 2007-03-29 at 17:09 -0400, Timothy Weber wrote:
> Herbert Graf wrote:
> > I apologize in advance if you're not talking about ICSP with PICs, but
> > if you're inferring ICSP with PICs doesn't happen below 5V, that's
> > simply wrong.
> >
> > The only reason you'd need 5V for ICSP is if you are code protecting
> > your part and need to do a bulk erase. Otherwise, any ICSP programmer
> > that allows for the target to supply power can theoretically support
> > ICSP below 5V. I know for a fact the ICD2 works perfectly fine
> > programming (and debugging) a part running at 3.3V without issues.
>
> Hey, that's great!  I guess there must be some other reason I've never
> been able to make it work in the context I have that uses 3V (PICkit 2,
> various PICs, XBees, not much else).  I'll look into it in more detail.

I will add that it does depend on the ICSP programmer, it must "support"
programming the PIC at voltages other then 5V. The ICD2 certainly
supports that, and I'd hope that most better ICSP programmers out there
would support it as well. I don't know if the PICkit 2 supports it,
anyone here know for sure?

TTYL

2007\03\29@184336 by olin piclist

face picon face
Herbert Graf wrote:
> Otherwise, any ICSP programmer
> that allows for the target to supply power can theoretically support
> ICSP below 5V.

Or the other approach is to let the programmer supply power.  All mine do
this so that they can set Vdd to 5V to do a bulk erase.  You do have to
design your power supply to be tolerant of this.  It might also mean
separating the PIC supply with diodes.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2007\03\29@184409 by Mark Rages

face picon face
On 3/29/07, Herbert Graf <EraseMEmailinglist3spam_OUTspamTakeThisOuTfarcite.net> wrote:
>
> I will add that it does depend on the ICSP programmer, it must "support"
> programming the PIC at voltages other then 5V. The ICD2 certainly
> supports that, and I'd hope that most better ICSP programmers out there
> would support it as well. I don't know if the PICkit 2 supports it,
> anyone here know for sure?
>
> TTYL
>

Yes, the PICkit 2 does, with the latest firmware.

Regards,
Mark
markrages@gmail

--
Most of the time,
for most of the world,
no matter how hard people work at it,
nothing of any significance happens.
    -- Weinberg's Law

2007\03\29@202435 by Timothy J. Weber

face picon face
Mark Rages wrote:
> On 3/29/07, Herbert Graf <mailinglist3spamspam_OUTfarcite.net> wrote:
>> I will add that it does depend on the ICSP programmer, it must "support"
>> programming the PIC at voltages other then 5V. The ICD2 certainly
>> supports that, and I'd hope that most better ICSP programmers out there
>> would support it as well. I don't know if the PICkit 2 supports it,
>> anyone here know for sure?
>
> Yes, the PICkit 2 does, with the latest firmware.

I may not have tried it since the old firmware.  Will give it another go.
--
Timothy J. Weber
http://timothyweber.org

2007\03\29@203636 by Timothy J. Weber

face picon face
Olin Lathrop wrote:
> Herbert Graf wrote:
>> Otherwise, any ICSP programmer
>> that allows for the target to supply power can theoretically support
>> ICSP below 5V.
>
> Or the other approach is to let the programmer supply power.  All mine do
> this so that they can set Vdd to 5V to do a bulk erase.  You do have to
> design your power supply to be tolerant of this.  It might also mean
> separating the PIC supply with diodes.

I've kept that sort of thing in mind, but I'm trying not to be that
clever.  The dumb solution (pop the PIC off the breadboard and into the
ZIF programmer) seems more reliable than "OK, have I been clever enough?
 Because if not, I'm about to fry this $32 module now."

I guess another issue for me is that some of my current projects have 3
different PICs on different breadboards.  So if I went with ICSP, I'd be
forever moving one set of programming leads around on the breadboards -
or I'd have to make three sets of flying-lead connectors to plug the
programmer into.  Then I'd have to pull them out anyway, because the
breadboards get taken elsewhere for testing.  Again, pulling chips and
plopping them in the ZIF socket is dumb but it's more reliable.

If I get this project to the point where I'm making PCBs, adding a few
headers for ICSP (and getting it working with self-power or adding
diodes) makes a lot more sense.  Thanks for the reminder that it Can Be
Done.
--
Timothy J. Weber
http://timothyweber.org

2007\03\30@015927 by wouter van ooijen

face picon face
> I guess another issue for me is that some of my current
> projects have 3
> different PICs on different breadboards.  So if I went with
> ICSP, I'd be
> forever moving one set of programming leads around on the
> breadboards -
> or I'd have to make three sets of flying-lead connectors to plug the
> programmer into.

Why not use 3 programmers? (and of course a makefile that donwloads
changed code automatically) My original programmer (WISP) even had the
ability to work with multiple programmers on one serial line, exactly
for this purpose. I abandoned that idea lateron because it was difficult
to make it reliable, and ubs-serial converters were getting cheap. Some
programmers (pickit2 for instance) can't do this beacause they have no
identity, but certainly for serial port programmers it should not be a
problem if you have enough ports (or when the progger supports
usb-to-serial: enough converters). One client bought a box of Wisp628's
(I don't recall the exact number) to build two batch production rigs. He
used PC cards with 32 serial ports.

One thing always baffeles me about people who want to have a ZIF in
their programmer: they never seem to have ZIF;s in their target
circuits!

For me the only valid reasons for ZIF sockets are:
- you sell programmed chips
- you want to test your circuit with different (types of) chips
but in both cases I would still use ICSP so I can test the programmed
chip immediately after programming.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2007\03\30@084421 by olin piclist

face picon face
Timothy J. Weber wrote:
> I guess another issue for me is that some of my current projects have 3
> different PICs on different breadboards.  So if I went with ICSP, I'd be
> forever moving one set of programming leads around on the breadboards -
> or I'd have to make three sets of flying-lead connectors to plug the
> programmer into.

Or three separate programmers.  You can have multiple USBProgs connected to
the same USB.  They come with unique names based on their serial numbers,
and the programming software can be told to use a particular programmer by
name.  You can also set the names yourself to be more relevant to your
project.

Note that I'm pointing out this is possible, but I'm not actually
recommending this in your case.  There are cases of particular lab or
production setups where a programmer per PIC does make sense.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2007\03\30@120713 by Timothy Weber
face picon face
wouter van ooijen wrote:
>> I guess another issue for me is that some of my current
>> projects have 3
>> different PICs on different breadboards.  So if I went with
>> ICSP, I'd be
>> forever moving one set of programming leads around on the
>> breadboards -
>> or I'd have to make three sets of flying-lead connectors to plug the
>> programmer into.
>
> Why not use 3 programmers?

(and so did Olin)

Interesting question.  Let me put some context on it: This is Compulsory
Electronics Project #7: Making my obsessively ideal living-room remote
control.  It has one PIC16F916 on the battery-powered remote that reads
buttons and talks to an XBee radio; one 16F688 on the mains-powered base
that talks to the receiving XBee; and an additional 12F675 there that's
dedicated to driving IRLEDs.  All are currently on small breadboards.

So every evening I bring either the remote, the base, or both up to the
third-floor workshop, tweak some aspect of it, then bring them back down
to the first-floor living room to test in situ while watching the
nightly DVD with the wife.  Usually, say, one to six programming
operations between trips, often on one chip, sometimes on two, seldom on
three.

Leaving flying leads in place would be messy.  Hooking them up for two
programmers might take more time than I spend pulling and replacing
chips (and seems more error-prone to me).  Buying a third programmer
probably wouldn't address that.

> One thing always baffeles me about people who want to have a ZIF in
> their programmer: they never seem to have ZIF;s in their target
> circuits!

I use solderless breadboards, and I have a good chip puller.  So for me
it's quite quick to pull and re-install.

Oh - and I often use small chips, 8- or 14-pins.  I think this changes
the ICSP/ZIF tradeoff some, because (A) you have fewer pins to dedicate
to ICSP, and if you can't dedicate them you have to think harder about
what's going to happen to them, and (B) smaller chips are easier to pull
and replace than bigger ones.

> For me the only valid reasons for ZIF sockets are:
> - you sell programmed chips
> - you want to test your circuit with different (types of) chips
> but in both cases I would still use ICSP so I can test the programmed
> chip immediately after programming.

I generally agree; I prefer ICSP.  I think this is an exception because
of these unusual factors:

- Multiple chips
- Must be transported to do system tests
- One-off, so no PCB yet (if ever), so must use flying leads for ICSP.

These factors combined mean lots of plugging and unplugging whether I
use ICSP or ZIF (either wires or chips).  And ZIF takes less thought,
especially in a design with devices that will fry if exposed to 5V by
accident.  Hypothetically speaking.  (Heavy sigh, wistful look.)
--
Timothy J. Weber
http://timothyweber.org

2007\03\30@131821 by wouter van ooijen

face picon face
> Interesting question.  Let me put some context on it: This is
> Compulsory
> Electronics Project #7: Making my obsessively ideal
> living-room remote
> control.  It has one PIC16F916 on the battery-powered remote
> that reads
> buttons and talks to an XBee radio; one 16F688 on the
> mains-powered base
> that talks to the receiving XBee; and an additional 12F675
> there that's
> dedicated to driving IRLEDs.  All are currently on small breadboards.

I am lazy. IMHO this is good in a software (systems) engineer :)

In your place I would either
- use chips that can write themselves, and arrange for a bootloading
scheme,
- make sure I could ICSP the chips in-situ (maybe with a very long USB
or  RS232 cable), or
- duplicate the setup on my desktop (and arrange for multiple
programmers and/or bootloading)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2007\03\30@154836 by Timothy Weber

face picon face
wouter van ooijen wrote:
>> Interesting question.  Let me put some context on it: This is
>> Compulsory
>> Electronics Project #7: Making my obsessively ideal
>> living-room remote
>> control.  It has one PIC16F916 on the battery-powered remote
>> that reads
>> buttons and talks to an XBee radio; one 16F688 on the
>> mains-powered base
>> that talks to the receiving XBee; and an additional 12F675
>> there that's
>> dedicated to driving IRLEDs.  All are currently on small breadboards.
>
> I am lazy. IMHO this is good in a software (systems) engineer :)

Yes!  Laziness is gospel!

> In your place I would either
> - use chips that can write themselves, and arrange for a bootloading
> scheme,

The savings would be two flying leads (async and ground) instead of
five, yes?  And the cost would be code space (two of these are small
chips), as well as complexity (I haven't used a bootloader before)?

> - make sure I could ICSP the chips in-situ (maybe with a very long USB
> or  RS232 cable), or

I don't have a computer in the living room, so those cables would have
to be what, 50-100 m to snake throughout the house!

> - duplicate the setup on my desktop (and arrange for multiple
> programmers and/or bootloading)

Which is something I've considered.  But the hardware is still in flux,
so that's a little extra time to build it plus extra headaches
diagnosing the things I forget to copy over to the other board.  And
anyway, even if I duplicated the breadboards, I'd have to duplicate the
DVD player and stereo amp in order to do a *full* system test!

I think the moral here is that the laziest option often depends on what
you've done before.
--
Timothy J. Weber
http://timothyweber.org

2007\03\30@155537 by Carl Denk

flavicon
face
I hope you are able to share the final (or near) result. The home A/V
remote control situation is getting out of hand, with some of the
commercial equipment expensive and then may not do what you need. :)

Timothy Weber wrote:
{Quote hidden}

2007\03\30@161715 by Timothy Weber

face picon face
Carl Denk wrote:
> I hope you are able to share the final (or near) result. The home A/V
> remote control situation is getting out of hand, with some of the
> commercial equipment expensive and then may not do what you need. :)

Yup, that's the motivation!  Seemed like fun, and will do exactly what I
need.  (For some definition of 'need.')  :)

I'll certainly post a link to the details when I get a chance to post them.
--
Timothy J. Weber
http://timothyweber.org

2007\03\30@163934 by Vasile Surducan

face picon face
On 3/29/07, wouter van ooijen <@spam@wouterKILLspamspamvoti.nl> wrote:
> > I guess another issue for me is that some of my current
> > projects have 3
> > different PICs on different breadboards.  So if I went with
> > ICSP, I'd be
> > forever moving one set of programming leads around on the
> > breadboards -
> > or I'd have to make three sets of flying-lead connectors to plug the
> > programmer into.
>
> Why not use 3 programmers?

Why not using one HV programmer connected to all ICSP on those three
breadboards and switch just the MCLR and SDA with a simple three
position two pole switch between those breadboards...

Or use a LVP programmer for all breadboards and a simple multiplexer

Or...

2007\03\30@174206 by Timothy Weber

face picon face
Vasile Surducan wrote:
> On 3/29/07, wouter van ooijen <KILLspamwouterKILLspamspamvoti.nl> wrote:
>>> I guess another issue for me is that some of my current
>>> projects have 3
>>> different PICs on different breadboards.  So if I went with
>>> ICSP, I'd be
>>> forever moving one set of programming leads around on the
>>> breadboards -
>>> or I'd have to make three sets of flying-lead connectors to plug the
>>> programmer into.
>> Why not use 3 programmers?
>
> Why not using one HV programmer connected to all ICSP on those three
> breadboards and switch just the MCLR and SDA with a simple three
> position two pole switch between those breadboards...
>
> Or use a LVP programmer for all breadboards and a simple multiplexer
>
> Or...

Many possibilities!  I guess those would get it down to reconnecting two
sets of leads rather than three (the base and remote units can't be
wired together during operation, so at least two sets).
--
Timothy J. Weber
http://timothyweber.org

2007\03\31@031608 by wouter van ooijen

face picon face
> > In your place I would either
> > - use chips that can write themselves, and arrange for a
> bootloading
> > scheme,
>
> The savings would be two flying leads (async and ground) instead of
> five, yes?  And the cost would be code space (two of these are small
> chips), as well as complexity (I haven't used a bootloader before)?

Use larger chips, arrange for your existing comms to piggyback a
bootloader comms.

> > - make sure I could ICSP the chips in-situ (maybe with a
> very long USB  or  RS232 cable), or
>
> I don't have a computer in the living room, so those cables
> would have to be what, 50-100 m to snake throughout the house!

yes, and?

> I think the moral here is that the laziest option often
> depends on what you've done before.

true. and my lazyness wants to eliminate the steps that are most error
prone when I execute them. like bent pins, exchanged chips, backwards
chips, programming the wrong hex file, etc. So I like a fully automated
scheme!

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu




'[EE] NaPiOn IR sensors revisited'
2007\04\03@134835 by alan smith
picon face
and thus the reason for JTAG......wouldnt it be nice if they supported JTAG? course it eats up 4 pins but maybe use the lower 4 port B pins? cant use that much silicon to throw on the JTAG state machine.

"Timothy J. Weber" <RemoveMEtwTakeThisOuTspamtimothyweber.org> wrote:  Olin Lathrop wrote:
> Herbert Graf wrote:
>> Otherwise, any ICSP programmer
>> that allows for the target to supply power can theoretically support
>> ICSP below 5V.
>
> Or the other approach is to let the programmer supply power. All mine do
> this so that they can set Vdd to 5V to do a bulk erase. You do have to
> design your power supply to be tolerant of this. It might also mean
> separating the PIC supply with diodes.

I've kept that sort of thing in mind, but I'm trying not to be that
clever. The dumb solution (pop the PIC off the breadboard and into the
ZIF programmer) seems more reliable than "OK, have I been clever enough?
Because if not, I'm about to fry this $32 module now."

I guess another issue for me is that some of my current projects have 3
different PICs on different breadboards. So if I went with ICSP, I'd be
forever moving one set of programming leads around on the breadboards -
or I'd have to make three sets of flying-lead connectors to plug the
programmer into. Then I'd have to pull them out anyway, because the
breadboards get taken elsewhere for testing. Again, pulling chips and
plopping them in the ZIF socket is dumb but it's more reliable.

If I get this project to the point where I'm making PCBs, adding a few
headers for ICSP (and getting it working with self-power or adding
diodes) makes a lot more sense. Thanks for the reminder that it Can Be
Done.
--
Timothy J. Weber
http://timothyweber.org

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...