Searching \ for '[EE] PCB traces as fuses?' 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/pcbs.htm?key=pcb
Search entire site for: 'PCB traces as fuses?'.

Exact match. Not showing close matches.
PICList Thread
'[EE] PCB traces as fuses?'
2006\01\22@021704 by Peter Todd

picon face
Just thought I'd run this idea by you guys...

I had another post a few days ago about building a PROM from out of a
PCB in visible scale. The critical part of that project would be to
design a fuse from PCB tracks, ideally one that can be blown using very
little actual energy. The last requierment is especially important, as
I'm trying to determine the feasibility of a long term counter, powered
by lithium coin cells, that would be designed to count for at least 10
years, perferably 50-100. Basically it would run in very slowly, using
nanowatts normally, but every, say, week, it would blow one PCB fuse in
an array of a few hundred. This assembly would be encased and the fuses
would be user visible through glass or plastic. I'd then call it art and
sell it to people and pray the batteries don't die and my customers
don't come bothering me when I'm 75....

Anyway I'm going to send off some gerbers to AdvancedCircuits tomorrow
for another project and I was planning to put some test fuses on them.
Right now I'm thinking I'll make each fuse be two large traces, say 250
mils, with the smallest allowable trace joining them, 6 mils. I'll do a
few different ones with different fuse trace lengths, I'm assuming
shorter fuses will be easier to blow, but thermal heatsinking effects
may of course make that quite wrong. I'm also thinking of trying two
pointed triangles, with the tips touching each other.

Testing this would then be a matter of connecting each fuse up to a
low-impedance power source and and firing. I think a DSO measuring the
voltage rise across a shunt resistor should get me my current, and
integrating the current waveform and voltage should get me roughly the
power used, I think...

Comments? I think my reasoning for this project is sound, but if you
guys think I'm missing anything, I'd love to hear! I did some google
searches on "PCB fuses" and variations, and wasn't able to find any
references. The before mentioned PCB trace width calculators don't seem
to be too much good either, they list safe currents, not fusing
currents.

--
spam_OUTpeteTakeThisOuTspampetertodd.ca http://www.petertodd.ca

2006\01\22@030841 by Jose Da Silva

flavicon
face
On January 21, 2006 11:31 pm, Peter Todd wrote:
> I had another post a few days ago about building a PROM from out of a
> PCB in visible scale. The critical part of that project would be to
> design a fuse from PCB tracks, ideally one that can be blown using
> very little actual energy. The last requierment is especially
> important, as I'm trying to determine the feasibility of a long term
> counter.

Take a look at a glass 10amp fuse and you will note the wire is rather
thin already.
You may be testing the boundaries of what can be considered reliable PCB
trace etching if you attempt to produce anything thinner, which is what
you need to do if you are attempting to use less power to blow a trace.

Just my 2 cents

2006\01\22@032223 by Peter Todd

picon face
On Sun, Jan 22, 2006 at 12:13:39AM -0800, Jose Da Silva wrote:
> On January 21, 2006 11:31 pm, Peter Todd wrote:
> > I had another post a few days ago about building a PROM from out of a
> > PCB in visible scale. The critical part of that project would be to
> > design a fuse from PCB tracks, ideally one that can be blown using
> > very little actual energy. The last requierment is especially
> > important, as I'm trying to determine the feasibility of a long term
> > counter.
>
> Take a look at a glass 10amp fuse and you will note the wire is rather
> thin already.
> You may be testing the boundaries of what can be considered reliable PCB
> trace etching if you attempt to produce anything thinner, which is what
> you need to do if you are attempting to use less power to blow a trace.

That's a rather good point... I should just go and measure one of those
wires...

Still I'm assuming that if a fab company like Advanced Circuits says
they can do 6mils for cheap, then they can do 6mils! As it is they say
they can do 4mil on special order, if I can combine that with a
non-standard 0.5oz/in2 plating thickness I may be set, I hope.

--
.....peteKILLspamspam@spam@petertodd.ca http://www.petertodd.ca

2006\01\22@034948 by Richard Prosser

picon face
10 years at 1 fuse blow a week would be 500 fuses. 100years is 5000
"fuses". You can't use binary coding because you can't unblow a fuse.

Otherwise, you should be able to make a track narrow enough toblow at
under 10 A - provided it is long & thin enough not to conduct the heat
away too quickly.

Add some inductance to the circuit and the stored energy will addd to
the effect. Add some capacitance & you may be able to recover some of
it back - with extra circuitry.

Mostly guessing though although I have used pcb track as a fusable
elemnt - until we tried for UL with the product!

RP

On 22/01/06, Peter Todd <petespamKILLspampetertodd.ca> wrote:
{Quote hidden}

> -

2006\01\22@040718 by Xiaofan Chen

face picon face
On 1/22/06, Richard Prosser <EraseMErhprosserspam_OUTspamTakeThisOuTgmail.com> wrote:

> Mostly guessing though although I have used pcb track as a fusable
> elemnt - until we tried for UL with the product!
>
> RP

In a word, they do not trust "PCB track" as a "fusable element", right?
That is what I know. And I will not trust that as well.

Regards,
Xiaofan

2006\01\22@073731 by Jinx

face picon face
> by lithium coin cells, that would be designed to count for at least 10
> years, perferably 50-100

At nA you'll be expecting around the shelf-life, which is ~ 10 years
for lithium to drop to 70%. Voltage output is quite stable for most
of its life though

2006\01\22@103014 by olin piclist

face picon face
Peter Todd wrote:
> Anyway I'm going to send off some gerbers to AdvancedCircuits tomorrow
> for another project and I was planning to put some test fuses on them.
> Right now I'm thinking I'll make each fuse be two large traces, say
> 250 mils, with the smallest allowable trace joining them, 6 mils.
> I'll do a few different ones with different fuse trace lengths, I'm
> assuming shorter fuses will be easier to blow, but thermal
> heatsinking effects may of course make that quite wrong.

You want to make them long enough so that the center gets close to the
exepected temperature for an infinitely long fuse with the same current.  I
don't know what the length/width ratio needs to be, but I'd knee jerk pick
around 10:1 for starters.

Of course if you can deliver the energy very quickly, then shorter shouldn't
matter much since there won't be time for heat flow away from the heated
area.  You probably want to slowly charge up a large electrolytic cap (or
probably a bunch in parallel) and then discharge it as quickly as possible
accross the fuse.  In that case you want the fuse resistance to roughly
match the cap ESR.

In any case, making fuses from PCB copper is going to be rather
unrepeatable.  A "6 mil" track is probably anywhere from 4-8 mils, check the
specs.  Also most board houses only guarantee the minimum coppper width, not
the maximum since for most applications more is better.  I think you should
probably expect a factor of 2-4 to 1 in resistance accross production lots.

> I'm also
> thinking of trying two pointed triangles, with the tips touching each
> other.

That will be even more prone to resistance variation as a function of track
width variation.

> Testing this would then be a matter of connecting each fuse up to a
> low-impedance power source and and firing. I think a DSO measuring the
> voltage rise across a shunt resistor should get me my current, and
> integrating the current waveform and voltage should get me roughly the
> power used, I think...

You're going to have a tough enough time blowing them without a shunt
resistor in there.  For testing take a large cap and measure the before and
after voltage.  From that you can calculate the energy used in blowing the
fuse.  Since that will be how the real circuit will likely do it, it will be
a relevant measure of circuit power drain to blow one fuse.  Also make sure
you connect the cap to the fuse with a transistor or FET, not a mechanical
switch.  The characteristics of a switch will otherwise get in the way.  You
may also find that the transistor makes a better fuse than the copper trace.

How about a bunch of the smallest and cheapest surface mount diodes?  I
think they can probably take less abuse than a 6 mil copper trace.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\22@104346 by David VanHorn

picon face
Getting them to blow in such a way that you can SEE them might be
interesting.
A dot of smokeless powder in an adhesive might work.

2006\01\22@105211 by olin piclist

face picon face
Olin Lathrop wrote:
> How about a bunch of the smallest and cheapest surface mount diodes?
> I think they can probably take less abuse than a 6 mil copper trace.

I was just thinking about this some more.  Maybe even the smallest cheapest
surface mount NPN transistors you can find.  Emitter to ground, collector to
cap, and the base becomes the self destruct input line via a 2Kohm resistor.

Or a bunch of cheap 0603 or 0402 LEDs arranged in a X-Y matrix.  The forward
voltage will guarantee just about all the energy will go into the LED.
Since they are diodes, the X-Y grid drive arrangement should work.  This
would make for a cool flash whenever you did blow a fuse.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\22@110508 by Spehro Pefhany

picon face
At 10:43 AM 1/22/2006 -0500, you wrote:
>Getting them to blow in such a way that you can SEE them might be
>interesting.
>A dot of smokeless powder in an adhesive might work.

'Caps' (the toy gun kind, not the capacitive kind) are cheap.

>Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speffspamspam_OUTinterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff


2006\01\22@112000 by David VanHorn

picon face
> Or a bunch of cheap 0603 or 0402 LEDs arranged in a X-Y matrix.  The
> forward
> voltage will guarantee just about all the energy will go into the LED.
> Since they are diodes, the X-Y grid drive arrangement should work.  This
> would make for a cool flash whenever you did blow a fuse.


Easy to make them fail shorted, harder to make them fail open.

2006\01\22@113511 by olin piclist

face picon face
David VanHorn wrote:
>> Or a bunch of cheap 0603 or 0402 LEDs arranged in a X-Y matrix.  The
>> forward
>> voltage will guarantee just about all the energy will go into the
>> LED. Since they are diodes, the X-Y grid drive arrangement should
>> work.  This would make for a cool flash whenever you did blow a fuse.
>
> Easy to make them fail shorted, harder to make them fail open.

If you blast enough current thru them, like discharging a large cap, they
pretty much have to fail open I would think.  I admit I haven't tried this.
Hmm, I do have a roll of 0603 LEDs around here somewhere...


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\22@114716 by Spehro Pefhany

picon face
At 11:35 AM 1/22/2006 -0500, you wrote:
>David VanHorn wrote:
> >> Or a bunch of cheap 0603 or 0402 LEDs arranged in a X-Y matrix.  The
> >> forward
> >> voltage will guarantee just about all the energy will go into the
> >> LED. Since they are diodes, the X-Y grid drive arrangement should
> >> work.  This would make for a cool flash whenever you did blow a fuse.
> >
> > Easy to make them fail shorted, harder to make them fail open.
>
>If you blast enough current thru them, like discharging a large cap, they
>pretty much have to fail open I would think.  I admit I haven't tried this.
>Hmm, I do have a roll of 0603 LEDs around here somewhere...

The bonding wire will probably act as the fuse, but being embedded in epoxy
it will take more energy to fuse than if the wire was hanging in mid-air.

Here's a chart of fusing currents for various types of wire:

http://www.speff.com/fusing.htm

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
@spam@speffKILLspamspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff


2006\01\22@114827 by David VanHorn

picon face
>
> If you blast enough current thru them, like discharging a large cap, they
> pretty much have to fail open I would think.  I admit I haven't tried
> this.
> Hmm, I do have a roll of 0603 LEDs around here somewhere...


You'd have to blow the bond wire, or crack the die.
Seems reasonable, and I like the idea for the light flash.

According to hollywood, all you have to do is bash a villain's head into it,
and the whole thing should go up in a shower of sparks, smoke, and flame. :)

2006\01\22@184551 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 01:32:53AM +1300, Jinx wrote:
> > by lithium coin cells, that would be designed to count for at least 10
> > years, perferably 50-100
>
> At nA you'll be expecting around the shelf-life, which is ~ 10 years
> for lithium to drop to 70%. Voltage output is quite stable for most
> of its life though

I've seen that figure a few times, and one of 0.5% capacity loss per
year.

I'm curious though... Does that 70% figure apply to the total capacity,
or voltage? IE if I can tolerate a lot of capacity wasted from
self-discharge, would my actual usable shelf-life be longer?

--
KILLspampeteKILLspamspampetertodd.ca http://www.petertodd.ca

2006\01\22@191300 by Jinx

face picon face
> I'm curious though... Does that 70% figure apply to the total
> capacity, or voltage? IE if I can tolerate a lot of capacity wasted
> from self-discharge, would my actual usable shelf-life be longer ?

I believe it would be capacity - voltage seems stable. See Fig 2

http://www.maxim-ic.com/appnotes.cfm/appnote_number/1793

http://www.73.com/specs/th693.pdf

2006\01\22@202909 by Peter Todd

picon face
On Sun, Jan 22, 2006 at 10:30:11AM -0500, Olin Lathrop wrote:
> You want to make them long enough so that the center gets close to the
> exepected temperature for an infinitely long fuse with the same current.  I
> don't know what the length/width ratio needs to be, but I'd knee jerk pick
> around 10:1 for starters.

Good idea, I'll try them around that level.

> Of course if you can deliver the energy very quickly, then shorter shouldn't
> matter much since there won't be time for heat flow away from the heated
> area.  You probably want to slowly charge up a large electrolytic cap (or
> probably a bunch in parallel) and then discharge it as quickly as possible
> accross the fuse.  In that case you want the fuse resistance to roughly
> match the cap ESR.

Sounds about right. Lithium coin cells are only rated for a maximum of
anywhere from 1ma to 20ma discharge from what I've seen in my research.

The shelf-life of the capacitors themselves could be an issue I think.
Electrolytic caps loose their electrolyte... This will be a
room-temperature application, but it's a very, very long time frame.

Still gotta figure out how many joules of energy I need to blow a fuse
first of course. :)

> In any case, making fuses from PCB copper is going to be rather
> unrepeatable.  A "6 mil" track is probably anywhere from 4-8 mils, check the
> specs.  Also most board houses only guarantee the minimum coppper width, not
> the maximum since for most applications more is better.  I think you should
> probably expect a factor of 2-4 to 1 in resistance accross production lots.

Thanks for pointing that one out, according to Advanced Circuits:

Trace Width/Air Gap
The greater of +/- 20% or +/- 0.002"

This is from their pcb tolerances section.

So I assume if I went with 4mil tracks, they could be 2mil, or 6mil.

So, roughly, the minimum fusing current is something like a third of
the maximum fusing current.

Here's a question... When they say "impedance controlled traces" do they
mean that they maintain consistant trace thicknesses? (amoung other
things)

{Quote hidden}

Good idea. So I would charge the cap up to maximum, disconect it from
the charger, blow my fuse, and measure how much more current is needed
to charge it again?

> a relevant measure of circuit power drain to blow one fuse.  Also make sure
> you connect the cap to the fuse with a transistor or FET, not a mechanical
> switch.  The characteristics of a switch will otherwise get in the way.  You
> may also find that the transistor makes a better fuse than the copper trace.

A very real issue... One problem I forsee, is that I need to have around
32 to 128 of these transistors in the final array for the x and y. They
will need to be very low inductance, but also be physically small enough
to fit on the perifery of the array.

I've got plenty of beefy FET transistors with really low ESR's, but
they're big!

> How about a bunch of the smallest and cheapest surface mount diodes?  I
> think they can probably take less abuse than a 6 mil copper trace.

Good idea. I'm not familiar with surface mount diodes, know any part
numbers I should buy to try experimenting with?

--
RemoveMEpeteTakeThisOuTspampetertodd.ca http://www.petertodd.ca

2006\01\22@203014 by Jinx

face picon face
> or voltage? IE if I can tolerate a lot of capacity wasted from
> self-discharge, would my actual usable shelf-life be longer?

As you're using such low quiescent, what will probably be the
dominant factor re life-time is electrolyte evaporation through
the seal

If you choose a battery that is hermetically sealed (welded) you
should be able to improve on the nominal 10 years for a crimped
version

http://www.sonnenschein-lithium.de/articles/04AutomotiveApplications.pdf

"The hermetically sealed case guarantees a long shelf life as well as
a long operating life under extreme temperature profiles. Unlike
crimp seals, the hermetic sealing prevents leakage and thus capacity
loss over the life time and damage to the application"


2006\01\22@205501 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 01:12:13PM +1300, Jinx wrote:
> > I'm curious though... Does that 70% figure apply to the total
> > capacity, or voltage? IE if I can tolerate a lot of capacity wasted
> > from self-discharge, would my actual usable shelf-life be longer ?
>
> I believe it would be capacity - voltage seems stable. See Fig 2
>
> http://www.maxim-ic.com/appnotes.cfm/appnote_number/1793

Yes, the above technote, and it's graphs going to 75 years, gave me much
hope. :)

> http://www.73.com/specs/th693.pdf

Thanks! That link is very complete. The bit about the internal
resistance is interesting, looks like at the end of the batteries
lifetime the internal resistance, which is already high, goes way up.
Could make charging my fuse blowing capacitors tougher.

Come to think of it... If I'm charging up a large capacitor bank, I'm
essentially shorting out the batteries. This would be bad, for one I'll
likely disrupt my power source for my uC. So given that, would PWM work
as a form of current control? I don't want to waste energy with a
current limiting resistor.

--
spamBeGonepetespamBeGonespampetertodd.ca http://www.petertodd.ca

2006\01\22@210422 by Peter Todd

picon face
On Sun, Jan 22, 2006 at 11:58:32AM -0500, Spehro Pefhany wrote:
> >If you blast enough current thru them, like discharging a large cap, they
> >pretty much have to fail open I would think.  I admit I haven't tried this.
> >Hmm, I do have a roll of 0603 LEDs around here somewhere...
>
> The bonding wire will probably act as the fuse, but being embedded in epoxy
> it will take more energy to fuse than if the wire was hanging in mid-air.
>
> Here's a chart of fusing currents for various types of wire:
>
> http://www.speff.com/fusing.htm

Thanks!

Geezus, that's really scary, 5.12 amps for a 6mil dia wire...

The chart looks accurate to me though. I've got some #32 stainless steel
wire for another project. According to your chart, iron fails at 2.21
amps. In my experience, SS is about 2.5A Not the exact same material of
course, but that's close.

I wonder if it's possible to get circuit boards that have *just* tin as
the conductive layer? Tin's fusing current is one-fifth that of copper.

--
TakeThisOuTpeteEraseMEspamspam_OUTpetertodd.ca http://www.petertodd.ca

2006\01\22@211905 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 02:29:28PM +1300, Jinx wrote:
> > or voltage? IE if I can tolerate a lot of capacity wasted from
> > self-discharge, would my actual usable shelf-life be longer?
>
> As you're using such low quiescent, what will probably be the
> dominant factor re life-time is electrolyte evaporation through
> the seal

Indeed. I'll be running my PIC chip at the absolute lowest possible
speed. It doesn't take much processor power to count... I'm sure a 32khz
crystal will be overkill in this case.

{Quote hidden}

Thanks! I'll look into those.

--
RemoveMEpetespamTakeThisOuTpetertodd.ca http://www.petertodd.ca

2006\01\22@212336 by Richard Prosser

picon face
Can you figure out a way to make solder traces ? Might be lower  still.

RP

On 23/01/06, Peter Todd <peteEraseMEspam.....petertodd.ca> wrote:
{Quote hidden}

> -

2006\01\22@213138 by Peter Todd

picon face
On Sun, Jan 22, 2006 at 11:16:23AM -0500, Spehro Pefhany wrote:
> At 10:43 AM 1/22/2006 -0500, you wrote:
> >Getting them to blow in such a way that you can SEE them might be
> >interesting.
> >A dot of smokeless powder in an adhesive might work.
>
> 'Caps' (the toy gun kind, not the capacitive kind) are cheap.

They are, and they remind me of my childhood!

I'm hoping to hermetrically seal the whole unit and have the insides
purged with, say, argon or something to ward off oxidisation and
condensation in the long run. So caps might mess up my pristene
atmosphere a bit.

That said... You made me think of thermal paper. Thermal paper over the
fuses might be enough to make a mark, even if the fuses don't blow.
Though I just tried some tests with thermal paper around my studio, and
the stuff melted without it changing color!

Also this link says it's not archival:

http://www.naa.gov.au/recordkeeping/rkpubs/advices/advice29.html

--
RemoveMEpeteEraseMEspamEraseMEpetertodd.ca http://www.petertodd.ca

2006\01\22@213928 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 03:16:06PM +1300, Richard Prosser wrote:
> Can you figure out a way to make solder traces ? Might be lower  still.

Definetely, higher resistance and they melt at much lower temperatures.

A quick test shows that 6V is quite enough to melt 3 inches of 0.015"
solder. Didn't measure how many amps though.

I think the big problem would be trying to mount the stuff to the board.
You can't exactly solder it in place. :) Maybe a very carefully
controlled heating process from below, maybe.

--
RemoveMEpetespam_OUTspamKILLspampetertodd.ca http://www.petertodd.ca

2006\01\22@214748 by Jinx

face picon face
> > http://www.maxim-ic.com/appnotes.cfm/appnote_number/1793

> Yes, the above technote, and it's graphs going to 75 years, gave me
> much hope. :)

Ah, but.... notice the cut-off line at 10 years. You can calculate
for many years past the physical life of the battery. eg throw in
10Ah D cells and you'd think, oooh, at 10nA Iq I could get more
than 114,000 years out of those. Except the batteries will leak
and rot before the end of Hilary's second term

> Come to think of it... If I'm charging up a large capacitor bank,
> I'm essentially shorting out the batteries. This would be bad

I've used the LM317 it in a similar situation to what you want to
do. My application had four solenoids that needed to be activated
as quickly as possible to fast-forward a four-face tower clock
(each solenoid drives a ratchet and pawl). To do this without a
large cap for each would have collapsed the 0.25A PSU I wanted
to use. I had the luxury of enough reserve to overcome serial
losses and Iq of the LM317 though. You'd have to find a low-
current way to limit both losses and capacitor charging if you want
as long a life as possible out of those coin cells. Any chance at
all of including some supplemetary supply ? Solar cell ? Except
they have a use-by date too ;-((  Induced charge ?

2006\01\22@220031 by Spehro Pefhany

picon face
At 09:45 PM 1/22/2006 -0500, you wrote:
>On Sun, Jan 22, 2006 at 11:16:23AM -0500, Spehro Pefhany wrote:
> > At 10:43 AM 1/22/2006 -0500, you wrote:
> > >Getting them to blow in such a way that you can SEE them might be
> > >interesting.
> > >A dot of smokeless powder in an adhesive might work.
> >
> > 'Caps' (the toy gun kind, not the capacitive kind) are cheap.
>
>They are, and they remind me of my childhood!
>
>I'm hoping to hermetrically seal the whole unit and have the insides
>purged with, say, argon or something to ward off oxidisation and
>condensation in the long run. So caps might mess up my pristene
>atmosphere a bit.
>
>That said... You made me think of thermal paper. Thermal paper over the
>fuses might be enough to make a mark, even if the fuses don't blow.
>Though I just tried some tests with thermal paper around my studio, and
>the stuff melted without it changing color!

That's what I thought of first, but it turns yellow and fades pretty
quickly (check any thermal receipts you may have in your tax files, for
example). 'not archival' may be an understatement.

>Also this link says it's not archival:
>
>http://www.naa.gov.au/recordkeeping/rkpubs/advices/advice29.html

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffTakeThisOuTspamspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff


2006\01\22@220930 by Jinx

face picon face
> ll be running my PIC chip at the absolute lowest possible speed. It
> doesn't take much processor power to count... I'm sure a 32khz
> crystal will be overkill in this case.

It does take "some" power though, and you'll have to decide which
part of the PIC you keep awake to do that

I haven't hit the books, but things to consider include -

Does 32kHz LP use less power than 1kHz RC ?
Do you keep the PIC asleep and use WDT to count ? (if so, the
suggestion is that you use a HS mode to count - awake shorter)

If you asssume X years is what you'll get from the battery,
that means working backwards to find the consumption to
get to X years

eg 87,600 hours in 10 years, 1uA Iq = 87.6mAh capacity, ignoring
voltage drooping at cell end-game. If your circuit can run on eg 2V
that would give you extra time

2006\01\22@234023 by Jinx

face picon face
How are you going to address these traces / get power to
them BTW ? (just causing trouble ;-))) )

2006\01\23@004226 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 05:40:32PM +1300, Jinx wrote:
> How are you going to address these traces / get power to
> them BTW ? (just causing trouble ;-))) )

x-y grid... Ahh, yeah, I knew I was forgetting something. x-y grids work
with diodes, like LEDS, just fine, not so well with fuses...

Hmm...Guess I get to blow up diodes then.

Well I did some tests with 1N4148's... Tested about 50 of them at
different voltages all supplied by a very beefy 25A power supply.

At 15V they always blow open. They blow extremely quickly with a little
blue flash. You can't actually see if they have blown, there are no
visible signs.

At 10V they normally blow open in the above manner. Otherwise they blow
a little differently as below...

At 5V they always blow "a little differently" In this mode of failure
they blow closed. What happens next is dependent on the amount of
mechanical force present on the glass envelope. If I'm carefull to exert
no stress on the envelope, it's a dead short and eventually one of the
leads heats to red hot and melts, yes, the leads, not the diode.

If there is stress on the envelope, ie if I have the leads bent, it
tends to visibly crack, which opens the circuit. It may stay in a
cylindrical shape, or if there is enough stress present they break into
two peices.


I also tried some standard low-intensity red LEDs.... They are very
boring to say the least. At 10V they blow for about 10 seconds then die.
At 15V they die instantly. In both cases you can see a bit of a black
spot by the die. No fun.

--
EraseMEpetespamspamspamBeGonepetertodd.ca http://www.petertodd.ca

2006\01\23@014347 by Jinx

face picon face
> I also tried some standard low-intensity red LEDs.... They are
> very boring to say the least. At 10V they blow for about 10
> seconds then die. At 15V they die instantly. In both cases you
> can see a bit of a black spot by the die. No fun

How inconsiderate of them.......

If you have enough I/O you could drive the "fuses" by x/y, similar
to the simplest keypad arrangement

Could you perhaps use thermochromic paint in any way ? You can
get both reversible and irreversible types. I know the irreversible
stuff is used to record if eg refrigerated food has been exposed to
too high a temperature. I think it goes blue or black, don't know
what colour it starts at though

2006\01\23@023430 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 07:42:57PM +1300, Jinx wrote:
> > I also tried some standard low-intensity red LEDs.... They are
> > very boring to say the least. At 10V they blow for about 10
> > seconds then die. At 15V they die instantly. In both cases you
> > can see a bit of a black spot by the die. No fun
>
> How inconsiderate of them.......

Yes indeed. I think all electronic componants should fail spectacularly
with a loud bang, a shower of sparks and leaving highly visible
scorchmarks.

> If you have enough I/O you could drive the "fuses" by x/y, similar
> to the simplest keypad arrangement

IO is no big deal, I'll just use i2c IO expanders or whatever else it
takes. Density will really be the issue, because I want everything to
fit on the periphery of the grid, yet still have enough current capacity
to work. From prior experience with LED arrays, that can be rather
tough.

Yes, you are right about the x/y. For some reason I was thinking for a
bit that you needed diodes for the interconnections between the x and y
layers.  But no, I am completely and utterly wrong on that one. :)

> Could you perhaps use thermochromic paint in any way ? You can
> get both reversible and irreversible types. I know the irreversible
> stuff is used to record if eg refrigerated food has been exposed to
> too high a temperature. I think it goes blue or black, don't know
> what colour it starts at though

Thank you! That's a wonderfull idea. A quick google search for
"thermochromic irreversible" showed up http://www.chemsong.com/spec.htm
which states that they have inks that go from black to red, or black to
blue, at 120C. If that's true I could simply paint the whole backside of
the PCB with the ink. I'll have to ask the manufacturer about stability
though, don't want the inks to be the same as in thermal paper...

I just tried experimenting with regular paper too actually. It browns
very reliably at 500C (measured with a cheap thermocouple) and
predictably chars or starts to burn when exposed to a red-hot wire. If I
sandwitch a wire and the paper between two sheets of glass and apply
current I get a very nice scorch mark, yet no matter how much current I
apply I can't get it to actually catch fire, presumably because of how
little oxygen is present. If I can treat the paper somehow so it won't
support combustion I should be set, and safe.

Well... I think I have my irreversible thermochromic ink for my
prototype at least. I think a nice pristine white sheet of paper would
look very beautifull too, carefully sandwiched behind glass, over time
ending up as a white sheet of paper with 4000-odd burn marks on it...

--
RemoveMEpeteKILLspamspampetertodd.ca http://www.petertodd.ca

2006\01\23@030919 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 02:48:47AM -0500, Peter Todd wrote:
> Yes, you are right about the x/y. For some reason I was thinking for a
> bit that you needed diodes for the interconnections between the x and y
> layers.  But no, I am completely and utterly wrong on that one. :)

Er, no, I was right... 2am and all.

Yeah, they've gotta be diodes, just draw an x/y grid. Fuses pass current
in both directions, so your grid, is just one big interconnected mess.
x-y switches work, because normally none of the switches are connected,
by pressing the switch, you make one connection. They fail the second
you press more than one button at once.

So I need a diode in series with each fuse. And I'd much rather avoid
soldering thousands of diodes... I'd love to hear some better ideas!
Even if this is like asking for better ideas to what 1 + 1 should equal
too...

--
peteSTOPspamspamspam_OUTpetertodd.ca http://www.petertodd.ca

2006\01\23@031957 by Shawn Wilton

picon face
If you want a big bang, then I'm *really* surprised no one has mentioned
Tantalum capacitors yet...

On 1/22/06, Peter Todd <spamBeGonepeteSTOPspamspamEraseMEpetertodd.ca> wrote:
{Quote hidden}

> -

2006\01\23@032218 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 03:47:56PM +1300, Jinx wrote:
> > > www.maxim-ic.com/appnotes.cfm/appnote_number/1793
>
> > Yes, the above technote, and it's graphs going to 75 years, gave me
> > much hope. :)
>
> Ah, but.... notice the cut-off line at 10 years. You can calculate
> for many years past the physical life of the battery. eg throw in
> 10Ah D cells and you'd think, oooh, at 10nA Iq I could get more
> than 114,000 years out of those. Except the batteries will leak
> and rot before the end of Hilary's second term

Yeah, very true. Even pacemaker batteries get replaced every 5-10 years,
and that's an application where maximum lifetime is very, very much
desired. I suspect there are quite simply too many unknowns when it
comes to anything over 10 years, or just too little testing.

I haven't seen a single battery officially rated with more than a 10
year lifespan, except RTG's, and even NASA has a hard time getting the
plutonium for those... The resent pluto mission was nearly set back
because Los Alamos almost couldn't supply enough plutonium.

{Quote hidden}

Hmm... So an LM317, operating in current regulating mode, is not
"burning up" power, like a linear regulator is when operating in voltage
regulating mode? But I will have loses due to the quescent current and
the inherent voltage drop of it right?

> all of including some supplemetary supply ? Solar cell ? Except
> they have a use-by date too ;-((  Induced charge ?

Induced charge? What's that?

Part of the problem is I want this to run even when it's packed off in a
storage crate. So there aren't really any outside sources of power. See
I'm planning on including a little photocell in the device. If the peice
is on display, and has light falling on it, it will blow one set of
fuses. If it's hidden away in some crate, it'll blow another set. My
current idea for the title is "Do you love me?"


All that said. Maybe I should count on solar energy, and only mark time
when it's on display... But solar cells having a lifetime? Didn't know
that, I thought they lasted until someone broke them, I mean, solar
energy systems have cells that have been running for 20+ years now.

--
EraseMEpetespamEraseMEpetertodd.ca http://www.petertodd.ca

2006\01\23@034259 by William Chops Westfield

face picon face
On Jan 23, 2006, at 12:23 AM, Peter Todd wrote:

> I'd love to hear some better ideas!

How about using silkscreen or inkjet like processes to lay down
somewhat conductive ink and/or something thermally sensitive OVER
your PCB traces (and perhaps forming some of the components?  That
would get you solder-based wires (using solder paste and post-baking,
if you're lucky and it doesn't wick away) or hopefully something
more secure using (for instance) graphite based compositions.
Similarly, you might etch a pattern using assorted techniques out
of a conductive material OTHER than copper.  LOTS of metals are
available as "foil" and could presumably be laminated similar to
the Nth layer of a normal PCB...  Or you could look at ways of
"manually" making lots of connections to wire-like material in a
quicker way; pick&place jumpers made out of bismuth, or some sort
of sewing-machine-like thing.  People used to assemble core memory
with thousands of cores, and while it wasn't QUICK, it wasn't that
slow either.  Likewise, hand wirewrap proceeds at better than 30
wires per hour, so 1000-odd connections isn't unthinkable.

But I think your hopes of running 10+ years and blowing 1000s of
fuses (of whatever sort) off of a single set of batteries are pretty
optimistic.

BillW

2006\01\23@042923 by Russell McMahon

face
flavicon
face
> I wonder if it's possible to get circuit boards that have *just* tin
> as
> the conductive layer? Tin's fusing current is one-fifth that of
> copper.

To special order you can probably get people to plate PTH holes to a
very low thickness. I don't know how reliable this could be done. The
hole may act as a heat concentrator and may be able to be sealed at
each end to assist this. or may not :-).
YMWACV.

Long battery life is not necessarily essential unless this is a
specific selling feature. Every 5 years or so it could prompt you to
replace one of 2 batteries - or the only battery as long as you don't
take more than a few days to insert the new one :-).

You could try for alternative power sources, but guaranteeing those
long term may be as difficult as using standard chemical cells.
Lifetimes for eg solar, Peltier etc may be problematic. At low enough
levels using rectified "radio waves" may be viable. Nanowatts mean
suggests around eg 10 nano-Amp current which should be quite easily
doable by RF pickup in most environments. "Warning - please return
device to a RF rich environment within 3 months to ensure continued
operation ...". :-)




       RM

2006\01\23@052255 by Jinx

face picon face

> Er, no, I was right... 2am and all.
>
> Yeah, they've gotta be diodes, just draw an x/y grid

Yes, you are right. Now that I think of it, the x/y row/column
grid is the same as a simple k/b layout but the switches are all
normally-closed

2006\01\23@052840 by Jinx

face picon face
> Induced charge? What's that?

I was just thinking of the possibility of picking up a little power
parasitically from nearby mains wiring, but I don't know if it
would be useable because of the signal's high impedance


2006\01\23@053843 by Enrico Schuerrer

picon face

----- Original Message -----
From: "Russell McMahon" <@spam@apptech@spam@spamspam_OUTparadise.net.nz>
To: "Microcontroller discussion list - Public." <spamBeGonepiclistspamKILLspammit.edu>
Sent: Monday, January 23, 2006 10:26 AM
Subject: Re: [EE] PCB traces as fuses?



{Quote hidden}

As part of the art you can use a serial connection of natural elements - 2
apples or potatoes with magnesium- and silver-electrodes - will deliver
around 5 Volts... and you can burnish the electrodes to high gloss.

But don't forget - for 75 years lifetime you have to dimension the
Mg-electrode rather thick :-)

regards

Enrico


2006\01\23@065439 by olin piclist

face picon face
Peter Todd wrote:
> Good idea. So I would charge the cap up to maximum, disconect it from
> the charger, blow my fuse, and measure how much more current is needed
> to charge it again?

Or just measure the before and after voltage, compute the energy stored in
the cap at each, and the difference is the energy used to blow the fuse.

> Good idea. I'm not familiar with surface mount diodes, know any part
> numbers I should buy to try experimenting with?

Search the DigiKey web site.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\23@072847 by olin piclist

face picon face
Peter Todd wrote:
> Well I did some tests with 1N4148's... Tested about 50 of them at
> different voltages all supplied by a very beefy 25A power supply.

Your circuit won't have a beefy 25A power supply.  To make the test
meaningful you should blow them in a way your circuit could too.  I think
that means charging up a big cap to a certain voltage then discharging it
accross the fuse.  You can experiment with different capacitances and
voltages.  Low ESR caps will be better, which you can get by putting several
of them in parallel.  You will probably need to do that anyway to get
sufficient capacitance.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\23@123707 by Sergey Dryga

face picon face
Peter Todd <pete <at> petertodd.ca> writes:

> So I need a diode in series with each fuse. And I'd much rather avoid
> soldering thousands of diodes... I'd love to hear some better ideas!
> Even if this is like asking for better ideas to what 1 + 1 should equal
> too...
>
How about a x-y positioning table with a punch? You can slowly move punch using
2 motors into a position where to make a hole, then use solenoid to punch a
hole in paper or another "screen"?  Instead of punching a hole you can paint a
dot too.

Sergey



2006\01\23@132828 by Peter

picon face


On Sun, 22 Jan 2006, Peter Todd wrote:

{Quote hidden}

You don't *have* to use electrolytics. Solid electrolyte caps are
available. Many are available in milspec and I think that they boast 10+
year shelf life.

All smaller caps can be ceramic or tantalum and the energy storage cap
can be as above. Then there is nothing to vent.

Peter

2006\01\23@153126 by Morgan Olsson

flavicon
face
I did blow a couple LEDs apart with a nice bang once upon a time when i was a lad, trimming a cirquit using a long steel screwdriver and dit not look and I put it between 230VAC and panel connector...  -That device did not need anmore trimming....
/Morgan

Olin Lathrop 17:35 2006-01-22:
{Quote hidden}

>-

2006\01\23@154156 by Hector Martin

flavicon
face
Peter wrote:
>
> You don't *have* to use electrolytics. Solid electrolyte caps are
> available. Many are available in milspec and I think that they boast 10+
> year shelf life.

>>>'Caps' (the toy gun kind, not the capacitive kind) are cheap.

He's not talking that kind of caps :)

--
Hector Martin (.....hectorspam_OUTspammarcansoft.com)
Public Key: http://www.marcansoft.com/hector.asc

2006\01\23@154613 by Jinx

face picon face
> I'll be running my PIC chip at the absolute lowest possible
> speed. It doesn't take much processor power to count

BTW, I didn't mention what I thought might be the lowest Iq
option

The alive part of the circuit could be the guts from a digital
watch without the display. I've got a tiny one on a shelf in
the shed that goes for years off a button cell, and it would go
for even longer without the LCD. If you got one that has a
daily alarm to wake the PIC, that's very slow counting

2006\01\23@203533 by Russell McMahon

face
flavicon
face
> How about a x-y positioning table with a punch?


Or a magnifying glass :-)
Now THAT would really be art.
"Place in a bright sunlit position to ensure correct operation ..."


       RM

2006\01\23@213045 by Peter Todd

picon face
On Tue, Jan 24, 2006 at 01:31:41PM +1300, Russell McMahon wrote:
> > How about a x-y positioning table with a punch?
>
>
> Or a magnifying glass :-)
> Now THAT would really be art.
> "Place in a bright sunlit position to ensure correct operation ..."

Ha!

Actually I once saw a very interesting sundial based on that principle
at Sudbury, Ontario's Science North. A large, 6 inch, diameter glass
sphere was positioned in the sun and a band of paper was positioned
behind it such that the sun was magnified and burned the paper over the
course of the day. Of course, you had to replace the paper, a special
non-flame propegating type, each day.

A strip of caps would make for an interesting display too...

--
TakeThisOuTpete.....spamTakeThisOuTpetertodd.ca http://www.petertodd.ca

2006\01\23@221621 by Richard Prosser

picon face
On 24/01/06, Peter Todd <TakeThisOuTpeteKILLspamspamspampetertodd.ca> wrote:

>
> Actually I once saw a very interesting sundial based on that principle
> at Sudbury, Ontario's Science North. A large, 6 inch, diameter glass
> sphere was positioned in the sun and a band of paper was positioned
> behind it such that the sun was magnified and burned the paper over the
> course of the day. Of course, you had to replace the paper, a special
> non-flame propegating type, each day.
>

Isn't that just a standard (old style) sunshine recorder?

RP

> A strip of caps would make for an interesting display too...
>
> --
> .....petespamRemoveMEpetertodd.ca http://www.petertodd.ca
> -

2006\01\23@222631 by Peter Todd

picon face
On Tue, Jan 24, 2006 at 04:16:20PM +1300, Richard Prosser wrote:
> > Actually I once saw a very interesting sundial based on that principle
> > at Sudbury, Ontario's Science North. A large, 6 inch, diameter glass
> > sphere was positioned in the sun and a band of paper was positioned
> > behind it such that the sun was magnified and burned the paper over the
> > course of the day. Of course, you had to replace the paper, a special
> > non-flame propegating type, each day.
> >
>
> Isn't that just a standard (old style) sunshine recorder?

Quite likely, I've never heard of a sunshine recorder, but that device
would record sunshine just fine...

--
RemoveMEpetespamspamBeGonepetertodd.ca http://www.petertodd.ca

2006\01\23@223420 by David VanHorn

picon face
>
>
> Isn't that just a standard (old style) sunshine recorder?


I've seen that principle rediscovered on some shiny glass buildings.
Whoopsie.

2006\01\23@225026 by Paul Hutchinson

picon face
> -----Original Message-----
> From: spamBeGonepiclist-bounces@spam@spamspam_OUTmit.edu On Behalf Of Richard Prosser
> Sent: Monday, January 23, 2006 10:16 PM
>
> On 24/01/06, Peter Todd <TakeThisOuTpetespamspampetertodd.ca> wrote:
>
> > Actually I once saw a very interesting sundial based on that principle
> > at Sudbury, Ontario's Science North. A large, 6 inch, diameter glass
> > sphere was positioned in the sun and a band of paper was positioned
> > behind it such that the sun was magnified and burned the paper over the
> > course of the day. Of course, you had to replace the paper, a special
> > non-flame propegating type, each day.
>
> Isn't that just a standard (old style) sunshine recorder?

That does describe the classic Campbell-Stokes sunshine recorder.
The old one @ Blue Hill Observatory, in service from 1886 to 1993.
http://www.bluehill.org/instruments/bho_ins63.jpg

The new one currently in service:
http://www.bluehill.org/instruments/bho_ins51.jpg

Paul

2006\01\23@231233 by Denny Esterline

picon face
> Isn't that just a standard (old style) sunshine recorder?
>

No, I think that defines a tree. Records annual sunshine, rain, and assorted
other environmental conditions. Logs (pun intended) the data for several
hundred years without replacement batteries. Seems there really isn't
anything new under the sun. :)

-Denny

2006\01\24@061047 by Gerhard Fiedler

picon face
Paul Hutchinson wrote:

> The new one currently in service:
> http://www.bluehill.org/instruments/bho_ins51.jpg

That looks like they are still recording to paper. Is this correct? If so,
what's the advantage (especially on days with rain and sunshine or when the
paper change person is sick)? There must be easier ways to get to the
data...

Gerhard

2006\01\24@071728 by olin piclist

face picon face
Denny Esterline wrote:
> No, I think that defines a tree. Records annual sunshine, rain, and
> assorted other environmental conditions. Logs (pun intended) the data
> for several hundred years without replacement batteries.

Several thousand for some.  In most places rings from old wood have been
pieced together to form a much longer term climate record than the lifetime
of a single tree.  If I remember right the tree ring record for the the
southwestern US goes back about 5000 years, and I don't mean just from the
handful of trees that are actually that old.  Parts of northern Europe,
Ireland, and other places with bogs that preserve fallen wood also have long
tree ring records that have been pieced together.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\25@000852 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 06:56:07AM -0500, Olin Lathrop wrote:
> Peter Todd wrote:
> > Good idea. So I would charge the cap up to maximum, disconect it from
> > the charger, blow my fuse, and measure how much more current is needed
> > to charge it again?
>
> Or just measure the before and after voltage, compute the energy stored in
> the cap at each, and the difference is the energy used to blow the fuse.


> > Good idea. I'm not familiar with surface mount diodes, know any part
> > numbers I should buy to try experimenting with?
>
> Search the DigiKey web site.

Which brings me to my #1 pet peeve about digikey...

Why the !@#$ can't I do a search based on price? I've been searching for
SMD diodes, and I've found hundreds of different types in the
low-current range, but almost none of those are cheap enough, even in
bulk, to work for me. Most of them cost about 50cents or higher each and
that price doesn't go down all that quickly.

Multiply by 4096 and you have a major problem.

The only cheap ones I've found are 1N4148's and the like, at 8cents or
less each.

Interestingly, though I should have expected this really, the glass
cased diodes are aparently tougher than plastic. The glass can
widthstand higher temperatures and higher power dissipations. Then again
when glass does fail it seems to do so more visibly.

What I may do is just use diodes as my "fusing" elements, but don't push
them until they fuse. Instead push them hard enough to heat up to over
120C and have irreversable thermochromatic paint on them so it's
possible to see that they have been heated.

The soldering of so many will still suck, but that's where I grunt and
say I'm a man, or something.

--
peteEraseMEspampetertodd.ca http://www.petertodd.ca

2006\01\25@001542 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 11:39:41AM +0100, Enrico Schuerrer wrote:
> As part of the art you can use a serial connection of natural elements - 2
> apples or potatoes with magnesium- and silver-electrodes - will deliver
> around 5 Volts... and you can burnish the electrodes to high gloss.
>
> But don't forget - for 75 years lifetime you have to dimension the
> Mg-electrode rather thick :-)

Reminds me of the Rube Goldberg machine I helped my brother build for
his physics class awhile back... It could do anything, so long as it
demonstrated science, somehow. So we settled on making a song play using
a tape deck and a battery. The tape deck was pretty standard... as for
the battery... think 4 litres of hot kool-aid.

See kool-aid's main ingrediant, other than sugar, is citric acid. Add
copper and zinc for the electrods and you have a nice battery. To make
it work as well as possible we used a long trough heated by candles that
the kool-aid passed through after mixing from powdered form...

My brother told me that when he tried it, everything worked fine, so
well in fact they got bored of their little tape player and hooked it up
to a large boombox!

--
RemoveMEpeteEraseMEspamspam_OUTpetertodd.ca http://www.petertodd.ca

2006\01\25@002016 by Peter Todd

picon face
On Tue, Jan 24, 2006 at 09:45:08AM +1300, Jinx wrote:
> > I'll be running my PIC chip at the absolute lowest possible
> > speed. It doesn't take much processor power to count
>
> BTW, I didn't mention what I thought might be the lowest Iq
> option
>
> The alive part of the circuit could be the guts from a digital
> watch without the display. I've got a tiny one on a shelf in
> the shed that goes for years off a button cell, and it would go
> for even longer without the LCD. If you got one that has a
> daily alarm to wake the PIC, that's very slow counting

Absolutely. I've seen papers on the designs of those things... Just
insane. It really makes me want to try another posters suggestion of
running it off of RF, like an old crystal radio.

Perhaps I could use a pre-built RTC? I vaguely remember seeing some
reference timed alarms in their datasheets... Have the RTC turn the PIC
on every week or something.

My understanding is that much of the ability of watch circuits and what
not to operate at such low power is controlling leakage currents. It's
why you can never make a modern CPU operate at nanowatts, the process
technology just won't allow it. PICs are pretty good, but I can only
assume even better would be something designed to run at 32khz rather
than up to 20mhz...

--
@spam@peteRemoveMEspamEraseMEpetertodd.ca http://www.petertodd.ca

2006\01\25@005946 by Paul Hutchinson

picon face
> -----Original Message-----
> From: EraseMEpiclist-bouncesspam@spam@mit.edu On Behalf Of Gerhard Fiedler
> Sent: Tuesday, January 24, 2006 5:52 AM
>
> That looks like they are still recording to paper. Is this correct? If so,
> what's the advantage (especially on days with rain and sunshine
> or when the paper change person is sick)? There must be easier ways to get
> to the data...

It is a special paper that isn't damaged by rain and there are back up
observers to make the daily chart change. This gives a highly redundant and
resilient data collection system on a medium that will be readable for many
centuries, possibly millennia ;-) (ref. the last CD-R thread). They have all
the charts dating back to 1886 making it the longest running continuous
sunshine record in North America.

Actually, the Blue Hill Observatory uses the modern NWS automated equipment
as its primary data source. However, they maintain a large range of
redundant instruments (classic and modern) and use human observers to catch
and correct errors in the automated system and, prevent missed data due to
equipment failure.

Paul

2006\01\25@010415 by Jose Da Silva

flavicon
face
On January 24, 2006 09:23 pm, Peter Todd wrote:
> Which brings me to my #1 pet peeve about digikey...
>
> Why the !@#$ can't I do a search based on price? I've been searching
> for SMD diodes, and I've found hundreds of different types in the
> low-current range, but almost none of those are cheap enough, even in
> bulk, to work for me. Most of them cost about 50cents or higher each
> and that price doesn't go down all that quickly.
>
> Multiply by 4096 and you have a major problem.
>
> The only cheap ones I've found are 1N4148's and the like, at 8cents
> or less each.

--snip--

If all you are really looking for is to cook some paint, why don't you
simply use some small resistors instead of diodes.
A PIC will source/sink 25mA, so...
5V / 25mA = 200ohms.
5V x 5V / 200ohms = 1/8watt.

Just use 200ohm resistors smaller than 1/8watt, so now you have a penny
each for hot resistors.

or

If you want to have fun, you can silkscreen that resistive paint used to
repair the rear window defrosters on cars. Then silkscreen your
permachange paint on the silkscreened resistors.

Just use a bit of imagination and Have fun.

2006\01\25@011552 by Jinx

face picon face
> Multiply by 4096 and you have a major problem.
>
> The only cheap ones I've found are 1N4148's and the like, at
> 8cents or less each

You're probably shopping in the wrong place - I'm not familiar
with Digikey's pricing but suspect they wouldn't be close to a
proper wholesaler. For comparison, if I bought 1N4148 from
Radiospares, I'd be paying NZ$0.12 / 1000. I always buy in
bulk when I can from a trade outlet, eg Active Components /
Arrow / Hagemeyer etc. The last 1N4148s I got were, I think,
NZ$0.009 / 2000. I know it was near to a cent anyway. Although
only 1000-ish were required for the job in hand, for the extra
10 bucks I need never worry about running short of 1N4148s

2006\01\25@013311 by Jinx

face picon face

> Perhaps I could use a pre-built RTC? I vaguely remember seeing
> some reference timed alarms in their datasheets... Have the RTC
> turn the PIC on every week or something

As long-term accuracy isn't an issue for you, you might want to
consider the DS1284 / DS1286. Has its own internal 10-year
battery and settable alarm. Sounds ideal

http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2678



2006\01\25@013820 by Peter Todd

picon face
On Wed, Jan 25, 2006 at 07:14:25PM +1300, Jinx wrote:
> > Multiply by 4096 and you have a major problem.
> >
> > The only cheap ones I've found are 1N4148's and the like, at
> > 8cents or less each
>
> You're probably shopping in the wrong place - I'm not familiar
> with Digikey's pricing but suspect they wouldn't be close to a
> proper wholesaler. For comparison, if I bought 1N4148 from
> Radiospares, I'd be paying NZ$0.12 / 1000. I always buy in
> bulk when I can from a trade outlet, eg Active Components /
> Arrow / Hagemeyer etc. The last 1N4148s I got were, I think,
> NZ$0.009 / 2000. I know it was near to a cent anyway. Although
> only 1000-ish were required for the job in hand, for the extra
> 10 bucks I need never worry about running short of 1N4148s

Ahh, well I should clarify that the 8 cents was for 10 of them,
discounts after that are pretty steep, not quite 1 cent each, but pretty
close. Full rolls of 5000 were for less than 1 cent each.

I'm sure in quantities of 10 the main cost is the poor sod who had to
cut them...


Of course many of the diodes I found are just plain expensive. 0.80 for
10, and the best discounts were on the order of 0.30 each or seomthing.
Very annoying to have to wade through.

--
@spam@petespam_OUTspam.....petertodd.ca http://www.petertodd.ca

2006\01\25@014107 by Peter Todd

picon face
On Tue, Jan 24, 2006 at 10:09:04PM -0800, Jose Da Silva wrote:
> If all you are really looking for is to cook some paint, why don't you
> simply use some small resistors instead of diodes.
> A PIC will source/sink 25mA, so...
> 5V / 25mA = 200ohms.
> 5V x 5V / 200ohms = 1/8watt.
>
> Just use 200ohm resistors smaller than 1/8watt, so now you have a penny
> each for hot resistors.
>
> or
>
> If you want to have fun, you can silkscreen that resistive paint used to
> repair the rear window defrosters on cars. Then silkscreen your
> permachange paint on the silkscreened resistors.
>
> Just use a bit of imagination and Have fun.

Great ideas, but to address 4000-odd resistors you still need an x-y
grid. And x-y grids need diodes to work. So either way I need some
diodes, might as well make them do something usefull directly.

Hmm... Wonder if you can silkscreen diodes? Bloody hard I'm sure, but
maybe...

--
spamBeGonepeteEraseMEspampetertodd.ca http://www.petertodd.ca

2006\01\25@052446 by Gerhard Fiedler

picon face
Paul Hutchinson wrote:

> Actually, the Blue Hill Observatory uses the modern NWS automated
> equipment as its primary data source. However, they maintain a large
> range of redundant instruments (classic and modern) and use human
> observers to catch and correct errors in the automated system and,
> prevent missed data due to equipment failure.

Ah, this makes more sense :)

Thanks,
Gerhard

2006\01\25@065110 by Russell McMahon

face
flavicon
face
> The soldering of so many will still suck, but that's where I grunt
> and
> say I'm a man, or something.

If you are building more than 1 or 2 of these an auto insertion
machine would be nice.

Even as a one off surface mount components may make sense -
aesthetically as well as circuit wise.

Will your thermochromatic paint have faded too much after say 50 years
? :-)
Who can tell you?
Do you trust them?
Do you care? :-)

       RM


2006\01\25@080522 by olin piclist

face picon face
Peter Todd wrote:
> Why the !@#$ can't I do a search based on price? I've been searching for
> SMD diodes, and I've found hundreds of different types in the
> low-current range, but almost none of those are cheap enough, even in
> bulk, to work for me. Most of them cost about 50cents or higher each and
> that price doesn't go down all that quickly.

That is waaaay too high.

> The only cheap ones I've found are 1N4148's and the like, at 8cents or
> less each.

Then you haven't tried very hard.  I just went to DigiKey and the following
took well under 30 seconds:

1 - Typed 4148 in the search box.

2 - Selected diodes/rectifiers under Discrete Semiconductor Products.

3 - Selected LL-34 package, then APPLY FILTERS.

4 - The top entry was FDLL4148TR-ND which cost $.02 each for a reel of 2500.

There may well be other diodes and/or other packages that are even cheaper,
but that's your job to dig out.  I just went for an obvious cheap answer to
prove the point.

You can also get 0603 LEDs for about the same price.  These are probably a
little easier to pop since they have a higher forward voltage and therefore
higher dissipation at the same current.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2006\01\26@033034 by Peter Todd

picon face
On Wed, Jan 25, 2006 at 07:32:02PM +1300, Jinx wrote:
>
> > Perhaps I could use a pre-built RTC? I vaguely remember seeing
> > some reference timed alarms in their datasheets... Have the RTC
> > turn the PIC on every week or something
>
> As long-term accuracy isn't an issue for you, you might want to
> consider the DS1284 / DS1286. Has its own internal 10-year
> battery and settable alarm. Sounds ideal
>
> http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2678

I looked through the datasheet for that one, yes, it does sound ideal.

Thanks!

--
petespamBeGonespampetertodd.ca http://www.petertodd.ca

2006\01\28@020846 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 08:28:25PM +0200, Peter wrote:
> > I'm hoping to hermetrically seal the whole unit and have the insides
> > purged with, say, argon or something to ward off oxidisation and
> > condensation in the long run. So caps might mess up my pristene
> > atmosphere a bit.
>
> You don't *have* to use electrolytics. Solid electrolyte caps are
> available. Many are available in milspec and I think that they boast 10+
> year shelf life.

Thanks! That's very usefull info, I'll look into those.

> All smaller caps can be ceramic or tantalum and the energy storage cap
> can be as above. Then there is nothing to vent.

Exactly. Mind you, I thought tantalums had a finite service life? Or am
I misunderremembering?

An array of 4000 or so ceramic disc caps could look quite nice come to
think of it...

--
RemoveMEpete@spam@spamspamBeGonepetertodd.ca http://www.petertodd.ca

2006\01\28@022124 by Peter Todd

picon face
On Thu, Jan 26, 2006 at 12:48:22AM +1300, Russell McMahon wrote:
> > The soldering of so many will still suck, but that's where I grunt
> > and
> > say I'm a man, or something.
>
> If you are building more than 1 or 2 of these an auto insertion
> machine would be nice.

Or some service... http://www.protosmtassembly.com is one of the lower costing
ones I've seen, (not that I know that field too well) and for a 4000
part board it'd cost $800

Even more scary is at least with them there's no economy of scale,
ordering 100 boards reduces the price to $550 per board.

> Even as a one off surface mount components may make sense -
> aesthetically as well as circuit wise.

Actually I suspect raw physics will do the trick. It takes less energy
to heat a small component to 120C than a large one!

> Will your thermochromatic paint have faded too much after say 50 years
> ? :-)
> Who can tell you?
> Do you trust them?
> Do you care? :-)

Sadly yes, I won't even be on my deathbed in 50 years! Last thing I want
is irate owners asking my wizened 70 year old self to fix those
things... I'll probably have forgotten what a PIC was by then, depending
on how soon I switch to lead-free solder...

--
.....pete@spam@spamEraseMEpetertodd.ca http://www.petertodd.ca

2006\01\28@022745 by Peter Todd

picon face
On Mon, Jan 23, 2006 at 05:36:03PM +0000, Sergey Dryga wrote:
> Peter Todd <pete <at> petertodd.ca> writes:
>
> > So I need a diode in series with each fuse. And I'd much rather avoid
> > soldering thousands of diodes... I'd love to hear some better ideas!
> > Even if this is like asking for better ideas to what 1 + 1 should equal
> > too...
> >
> How about a x-y positioning table with a punch? You can slowly move punch using
> 2 motors into a position where to make a hole, then use solenoid to punch a
> hole in paper or another "screen"?  Instead of punching a hole you can paint a
> dot too.

That's a very interesting idea... One way to do it, would be to use
gravity as your source of power. Think a razor blade, slicing a ribbon
or something. Support the razor blade on some sort of "rocking walker"
that walks it's way down two groved bars side-by-side. Each tic is one
rock, triggered by some sort of low-power actuator. Design the mechanics
so it will take little force to release the stored energy of gravity.

One tricky part would be trying to make sure it'll run, in any
orientation. So put the whole thing in a gimbaled mount. Big, but
complex and pretty!

--
.....peteRemoveMEspampetertodd.ca http://www.petertodd.ca

2006\01\28@054211 by Shawn Wilton

picon face
>
> Sadly yes, I won't even be on my deathbed in 50 years! Last thing I want
> is irate owners asking my wizened 70 year old self to fix those
> things... I'll probably have forgotten what a PIC was by then, depending
> on how soon I switch to lead-free solder...


Why don't you just offer something along the lines of a warranty.
Guaranteed to explode caps for x years, after which you're SOL.  Or they can
opt to purchase a new piece of art at the end of the "warranty" period.

--
> .....peteSTOPspamspam@spam@petertodd.ca http://www.petertodd.ca
> -


'[EE] PCB traces as fuses?'
2006\02\04@111306 by Peter Todd
picon face
On Sun, Jan 22, 2006 at 09:18:38PM -0500, Peter Todd wrote:

Well, I got some boards back with my PCB fuses design etched into
them... Results are as follows:

http://petertodd.ca/persist/2006-02-04/pcb_fuses.jpg

The holes are standard 0.1 grid spacing. The fuse traces were 8mil wide.

Sure enough you guys were right, the shortest fuse, the extreme left
and right, didn't blow very well at all, it's heat getting conducted
away from the wire. The best results were obtained with the longest
fuses.

I didn't accurately measure current/energy used, but for what it's worth
I used 20V from a large power supply for each one. On almost every one
the current meter on the supply moved about half an amp, of course peak
currents were probably high all the same, but it's a good sign. The
longest traces blew with a bit of a spark, pretty much instantly.

Testing with 4mil traces could prove interesting... And I'll find myself
some low-ESR caps to properly measure the energy used sometime.

{Quote hidden}

> --

2006\02\05@034635 by William Chops Westfield

face picon face

On Feb 4, 2006, at 8:28 AM, Peter Todd wrote:
>
> Well, I got some boards back with my PCB fuses design etched into
> them... Results are as follows:
>
> http://petertodd.ca/persist/2006-02-04/pcb_fuses.jpg
>
Oh.  that looks pretty good.  I didn't expect the traces
to burn through nearly so completely as it looks like they
did.  With some backlighting and thin PCB material, it looks
like it should be pretty visible to the human eye.  Can
you post a picture of what it looks like with a few "blown"
and a few "intact" "fuses."

BillW

2006\02\05@044250 by Peter Todd

picon face
On Sun, Feb 05, 2006 at 12:46:33AM -0800, William Chops Westfield wrote:
>
> On Feb 4, 2006, at 8:28 AM, Peter Todd wrote:
> >
> > Well, I got some boards back with my PCB fuses design etched into
> > them... Results are as follows:
> >
> > petertodd.ca/persist/2006-02-04/pcb_fuses.jpg
> >
> Oh.  that looks pretty good.  I didn't expect the traces
> to burn through nearly so completely as it looks like they
> did.  With some backlighting and thin PCB material, it looks
> like it should be pretty visible to the human eye.  Can
> you post a picture of what it looks like with a few "blown"
> and a few "intact" "fuses."

Sure thing, I don't have an example with say, half-blown, half-not, but
here's the exact same set of traces as above from another copy of that
board:

http://petertodd.ca/persist/2006-02-05/pcb_fuses_intact.jpg

I think at worse a frensle lense would be just fine for making the whole
thing visible. Heck, put a magnifying glass on a string next to the
peice when it's on display.

Backlighting could be interesting... Make it so the peice's back is
fully transparent, so you could hang it by a window. I think even
standard PCB material is translucent enough to achieve that. Makes
finding a spot to put the batteries harder though.

--
RemoveMEpetespamspamBeGonepetertodd.ca http://www.petertodd.ca

More... (looser matching)
- Last day of these posts
- In 2006 , 2007 only
- Today
- New search...