Homebrew PCBs

How are those tiny metal cilinder named?i`m talking about the one 
used to connect top layer to bottom layer.

On Mon, 16 Aug 2004 09:17:59 -0000, preda mihai <myke_oltcit@...>  
wrote:

> How are those tiny metal cilinder named?i`m talking about the one
> used to connect top layer to bottom layer.

they are called plated (through-)hole, or via (but i think via only if  
there is no
component wire soldered through them).


ST

----- Original Message ----- 

From: "preda mihai" <myke_oltcit@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Monday, August 16, 2004 10:17 AM
Subject: [Homebrew_PCBs] double layer question


> How are those tiny metal cilinder named?i`m talking about the one 
> used to connect top layer to bottom layer.

Eyelets?

Leon

At 03:17 AM 8/16/2004, preda mihai wrote:
>How are those tiny metal cilinder named?i`m talking about the one
>used to connect top layer to bottom layer.

The term is "via".

If you are looking for the name of the parts used to make vias on homemade 
PCBs, I can think of two: Harwin "track pins" and Multicore 
"copperset".  Track pins are a tapered pin that fits tightly into a #66 
drill hole - you solder both top and bottom sides to make the 
connection.  Copperset vias are used to make real vias on a PCB - you wind 
up with a plated through hole.  But they are expensive and time consuming 
to use.

dwayne

-- 
Dwayne Reid   <dwayner@...>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 20 years of Engineering Innovation (1984 - 2004)
  .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
     `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

In a message dated 8/16/2004 4:19:02 AM Central Standard Time, 
myke_oltcit@... writes:
How are those tiny metal cilinder named?
"Eyelets"


[Non-text portions of this message have been removed]

--- In Homebrew_PCBs@yahoogroups.com, "Leon Heller" 
<leon_heller@h...> wrote:
> ----- Original Message ----- 
> From: "preda mihai" <myke_oltcit@y...>
> To: <Homebrew_PCBs@yahoogroups.com>
> Sent: Monday, August 16, 2004 10:17 AM
> Subject: [Homebrew_PCBs] double layer question
> 
> 
> > How are those tiny metal cilinder named?i`m talking about the one 
> > used to connect top layer to bottom layer.
> 
> Eyelets?

Also called rivets:

http://www.megauk.com/through_hole_rivets.php

These seem to go down to 0.7mm hold diameter, 2.2mm length. £14:50 
doesn't seem particularly cheap to me though.

Mike

In a message dated 8/16/2004 10:35:49 PM Central Standard Time, 
yahoo@digital-wizardry.co.uk writes:
£14:50 doesn't seem particularly cheap to me though.<<
Nope!  Hmmm . . .   One wonders what kind of firm or operation would be their 
primary customers for this kind of pricing??? 


[Non-text portions of this message have been removed]

Hi,


has one of you ever tried making a multi-position switch directly on the
PCB like most handheld meters have?
I can't gold plate it (yet).
I think tinning would do for now.

I think of using a M6 screw as axis, just a washer on each side to make a  
bearing
of a hole in the PCB. Then take a piece of FR4 and make the "wiper" with  
contacts
of thin sheetmetal.
The M6 screw would fit into a knob on the front panel.
The "snap" action could be achieved by drilling holes in the PCB and  
letting
some spring catch them.

What do you think of that? stupid idea?

It seems i can't get rotrary switches with enough positions.
Would be nice to be able to just make them how i want.
After all it works in meters...

tell me what you thik about that.

ST

On Monday 30 August 2004 03:12 pm, Stefan Trethan wrote:

> has one of you ever tried making a multi-position switch directly on the
> PCB like most handheld meters have?

Not me!

> I can't gold plate it (yet).
> I think tinning would do for now.
>
> I think of using a M6 screw as axis, just a washer on each side to make a
> bearing of a hole in the PCB. Then take a piece of FR4 and make the "wiper"
> with contacts of thin sheetmetal. The M6 screw would fit into a knob on the
> front panel. The "snap" action could be achieved by drilling holes in the
> PCB and letting some spring catch them.
>
> What do you think of that? stupid idea?

I wouldn't go that far.  :-)  But I wouldn't do that,  either.

> It seems i can't get rotrary switches with enough positions.

How many positions are you looking for?  I've seen up to 11,  maximum,  with 
very few exceptions -- one time I heard of 22,  but that would be a 
custom-made job like what you're talking about here.

Tinned contact on the board,  and metal?  Whether that will work out or not 
depends on what you plan to switch with this.

> Would be nice to be able to just make them how i want.
> After all it works in meters...

Both of my meters here use rows of buttons.  :-)

My old (as old as I am!) Simpson 260 is the only one with a rotary switch.

> tell me what you thik about that.

What are you switching?  How many positions do you really _need_?  Have you 
noticed that rotary switches are in large part going away,  or at least they 
have been since the fifties?  There's a reason for that.  Feel free to take 
this over to Electronics101 if we get too far off the topic for in here.

>
> What are you switching?  How many positions do you really _need_?  Have  
> you
> noticed that rotary switches are in large part going away,  or at least  
> they
> have been since the fifties?  There's a reason for that.  Feel free to  
> take
> this over to Electronics101 if we get too far off the topic for in here.
>

I need 2*10 positions, and can only get 2*6 or 1*12.
For selecting cell count on a battery charger (no current at all).
Almost all handheld meters i've ever seen use rotary switches, and all i  
took
apart had "custom" printed contacts. Oh, yes, there was one rotten piece  
of bad engineering
that had four buttons instead. The damn thing had dozens of design flaws,  
like blowing up
each time you pulled the current plug....

I could try to pull some mechanical stunt and stack 2 1*12 switches.
I don't know if that would be possibe with the switches i can get.
But then, one switch for 2.50eur is not exactly cheap, not speaking of 2...

ST

On Monday 30 August 2004 06:27 pm, Stefan Trethan wrote:
> > What are you switching?  How many positions do you really _need_?  Have
> > you noticed that rotary switches are in large part going away,  or at
> > least they have been since the fifties?  There's a reason for that.  Feel
> > free to take this over to Electronics101 if we get too far off the topic
> > for in here.

> I need 2*10 positions, and can only get 2*6 or 1*12.

Sounds about right.  The way they used to do that was to stack two switch 
wafers, not a big deal.  I'm sure that such a part is out there and available 
somewhere,  though whether it's available at a reasonable cost to you with 
where you're located is another issue that I can't really address.

> For selecting cell count on a battery charger (no current at all).
> Almost all handheld meters i've ever seen use rotary switches, and all i
> took apart had "custom" printed contacts. Oh, yes, there was one rotten
> piece of bad engineering that had four buttons instead. The damn thing had
> dozens of design flaws, like blowing up each time you pulled the current
> plug....
>
> I could try to pull some mechanical stunt and stack 2 1*12 switches.
> I don't know if that would be possibe with the switches i can get.
> But then, one switch for 2.50eur is not exactly cheap, not speaking of 2...

That *IS* cheap!  What does it look like?  The typical rotary switch I know of 
has a metal front with mounting bushing that the shaft comes through,  with 
two "legs" that are crimped beyond the point where they go through the wafer. 
The heavier-duty ones,  and the ones that are designed for more than one 
wafer,  are done with longish bolts that go through the metal part and 
through the wafers with spacers in between each wafer and between the last 
one and the front metal,  to space them out.

Wafers seem to be phenolic,  typically,  though I've also seen some that were 
done in ceramic as well.

Contacts can be shorting or non-shorting,  or all sorts of strange 
configurations in some applications.

Maybe if you can find an old tv set that's good for scrap you can do something 
with the tuner?  It'll require some mechanical work,  but it may give you a 
usable multi-level rotary switch with (perhaps) enough positoins.  Typical 
here is channels 2-13 and a "UHF" position that selects a different 
tuner,what's typical there?

> I need 2*10 positions, and can only get 2*6 or 1*12.
> For selecting cell count on a battery charger (no current at all).
> Almost all handheld meters i've ever seen use rotary switches, and
all i  
> took
> apart had "custom" printed contacts. Oh, yes, there was one rotten
piece  
> of bad engineering
> that had four buttons instead. The damn thing had dozens of design
flaws,  
> like blowing up
> each time you pulled the current plug....
> 
> I could try to pull some mechanical stunt and stack 2 1*12 switches.
> I don't know if that would be possibe with the switches i can get.
> But then, one switch for 2.50eur is not exactly cheap, not speaking
of 2...


Would be nice to have a couple more details on the circuit this is
going to be connected to.

Are you changing a resistance? Could you use a potentiometer and then
work out some kind of mechanical detent? Why do you need a dual-pole
switch? Have you looked at analog multiplexer chips? Digital
potentiometer chips? Don't reject them without looking at them, they
are not all expensive. A 16:1 analog mux/demux CD74HCT4067 costs 90
cents in DIP, 48 cents in SOIC. Two of those give you your dual-pole
circuit if you need it. They accept a simple 4-bit input to select
which output goes where. You could provide that input with a
16-position hex switch, or a 4-bit counter (rollover at 12) operated
by pushbuttons. Another 50-cent 74HC154 4:16 decoder could give you an
LED feedback showing which output is selected.

It would be neat to have 12 LEDs surrounding a single central
pushbutton which increments the LED position.

Or maybe you could just go for a rotary switch with fewer positions,
like 6. Then you use a dual-pole switch to choose between high range
and low range. Though even 6-position dual-pole rotaries are all very
expensive unless you get lucky in surplus stores.

As for rolling your own? It'll just have to be trial and error.

--- Stefan Trethan <stefan_trethan@...> wrote:

> Hi,
> 
> 
> has one of you ever tried making a multi-position
> switch directly on the
> PCB like most handheld meters have?
Yes, but not at home with limited tools.... Its great
in volume production as its super cost effective, but
we would take care of the mechanical issues in the
molds.


> I can't gold plate it (yet).
> I think tinning would do for now.
Its a function of life span and environment. Gold is
pretty soft, but at least it doesn't corrode.

> 
> I think of using a M6 screw as axis, just a washer
> on each side to make a  
> bearing
> of a hole in the PCB. Then take a piece of FR4 and
> make the "wiper" with  
> contacts
> of thin sheetmetal.
I think it would be trouble. We always used berylium
copper, and had them formed off site, even for
proto's. I don't think I would want to mess with it at
home. Its nasty stuff.


> The M6 screw would fit into a knob on the front
> panel.
> The "snap" action could be achieved by drilling
> holes in the PCB and  
> letting
> some spring catch them.
FR4 can be pretty abrasive to use as a detent. (we
tired), then again we wanted 20,000 cycles.

> 
> What do you think of that? stupid idea?
Could you use a 360 degree pot, and then detents with
a cam and spring arrangement. It would make the
electrical wear and tear a lot less of a headache.

One could also use optocouplers and a mylar disk...
bit it gets spendy.

I've also seen a magnet, with a sphereical ball
actuator that handled the contact switching as it
rolled along.


> 
> It seems i can't get rotrary switches with enough
> positions.
> Would be nice to be able to just make them how i
> want.
> After all it works in meters...
And other place, a rotary switch is generally not cost
effective, but it is the most intuitive, at least for
those of us in the older generation. The  youngsters
may have grown up with bottom remotes etc, but I still
have a few tv sets with rotary volume and channel
select. EVen Fluke tried to go away from the rotary
switch in the late 80's.... the product failed. They
tried again with a lower priced model in the late
nineties. I don't think its doing to well sales wise.


> 
> tell me what you thik about that.
DIY without extansive cad and proto facilities would
be a challenge, then agan the end result would be
pretty cool. 
Ron
> 
> ST
> 
> 
> 
> ------------------------ Yahoo! Groups Sponsor
> --------------------~--> 
> $9.95 domain names from Yahoo!. Register anything.
>
http://us.click.yahoo.com/J8kdrA/y20IAA/yQLSAA/bGYolB/TM
>
--------------------------------------------------------------------~->
> 
> 
> Be sure to visit the group home and check for new
> Bookmarks and files:
> http://groups.yahoo.com/group/Homebrew_PCBs 
> Yahoo! Groups Links
> 
> 
>     Homebrew_PCBs-unsubscribe@yahoogroups.com
> 
>  
> 
> 



		
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail

On Monday 30 August 2004 07:29 pm, ron amundson wrote:

> One could also use optocouplers and a mylar disk...
> bit it gets spendy.

Not necessarily.  I have a couple of motors equipped with such encoders.  And 
you have two of them in any given mouse that's out there -- it shouldn't be 
too hard to adapt that technology to be used with a knob and a shaft,  though 
a good bit of weight behind the panel gives it a better feel.

This sort of thing was actually used,  in a z80-based system to control most 
all functions on the Moog Source -- you had what was basically a membrane 
panel,  where you selected what function you wanted to mess with,  you had a 
couple or three digits of LED display to tell you what the current value was,  
and you had one big knob to fiddle with it.

Not bad...

----- Original Message ----- 

From: "Roy J. Tellason" <rtellason@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Tuesday, August 31, 2004 12:43 AM
Subject: Re: [Homebrew_PCBs] homebrew rotary multi-position switch


> On Monday 30 August 2004 07:29 pm, ron amundson wrote:
>
> > One could also use optocouplers and a mylar disk...
> > bit it gets spendy.
>
> Not necessarily.  I have a couple of motors equipped with such encoders.
And
> you have two of them in any given mouse that's out there -- it shouldn't
be
> too hard to adapt that technology to be used with a knob and a shaft,
though
> a good bit of weight behind the panel gives it a better feel.
>
> This sort of thing was actually used,  in a z80-based system to control
most
> all functions on the Moog Source -- you had what was basically a membrane
> panel,  where you selected what function you wanted to mess with,  you had
a
> couple or three digits of LED display to tell you what the current value
was,
> and you had one big knob to fiddle with it.

Electro-mechanical encoders are quite cheap, they are made by ALPS and
Bourne. One of those with a small MCU and a display will do what you want.

Leon

--- Leon Heller <leon_heller@...> wrote:

> ----- Original Message ----- 
> From: "Roy J. Tellason" <rtellason@...>
> To: <Homebrew_PCBs@yahoogroups.com>
> Sent: Tuesday, August 31, 2004 12:43 AM
> Subject: Re: [Homebrew_PCBs] homebrew rotary
> multi-position switch
> 
> 
> > On Monday 30 August 2004 07:29 pm, ron amundson
> wrote:
> >
> > > One could also use optocouplers and a mylar
> disk...
> > > bit it gets spendy.
> >
> > Not necessarily.  I have a couple of motors
> equipped with such encoders.
> And
> > you have two of them in any given mouse that's out
> there -- it shouldn't
> be
> > too hard to adapt that technology to be used with
> a knob and a shaft,
> though
> > a good bit of weight behind the panel gives it a
> better feel.
> >
> > This sort of thing was actually used,  in a
> z80-based system to control
> most
> > all functions on the Moog Source -- you had what
> was basically a membrane
> > panel,  where you selected what function you
> wanted to mess with,  you had
> a
> > couple or three digits of LED display to tell you
> what the current value
> was,
> > and you had one big knob to fiddle with it.
> 
> Electro-mechanical encoders are quite cheap, they
> are made by ALPS and
> Bourne. One of those with a small MCU and a display
> will do what you want.
> 
> Leon
> 

So what I gather, is a guy could put a good sized
weight on the shaft to create inertia and give a
perception of quality, then use optical interrupters
to capture disc position?

I've looked at the mechanical rotary encoders in the
past, but even in volume they were a couple bucks. The
idea of using oprtical is quite interesting. I'd
completely forgotten that is the basic mouse concept.
All one would need to do, is add some type of position
encoding to the mouse wheel, and one would be all set.

Speaking of which, it would seem practical to use one
of those LED's that illuminates through a hole in the
board for the coupler section. Has any one had any
issues with pad sizing... it seems like it might be
somewhat dependant on laminate material, as you have
this big hole right under the part. Its so rare that
one can actually use the manufacturers recommended pad
size without tweaking. I'd like to know if anyone has
rules of thumb. 

Ron

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

On Monday 30 August 2004 10:41 pm, ron amundson wrote:

> So what I gather, is a guy could put a good sized weight on the shaft to
> create inertia and give a perception of quality, then use optical
> interrupters to capture disc position?

Capturing the _change_ in position is more like it,  depending on where the 
software that was dealing with it assumed was the position to start with.

> I've looked at the mechanical rotary encoders in the past, but even in
> volume they were a couple bucks. The idea of using oprtical is quite
> interesting. I'd completely forgotten that is the basic mouse concept.

I have a couple of dead ones here that I'm thinking about playing with at some 
point...

It's interesting to note that the LED/Detector pair on these is discrete 
components.  I also have some optointerrupters I'm planning on trying to do 
something with that I salvaged out of dead floppy drives.  Also some similar 
parts which apparently were designed to use a reflective principle.

> All one would need to do, is add some type of position encoding to the mouse
> wheel, and one would be all set.

Nah,  that'll all be handled in the software.  As would the resolution,  for 
that matter.

Could you use multiplexers? That would be more reliable and could 
have as many switches as you want. Mike


-- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@g...> wrote:
> Hi,
> 
> 
> has one of you ever tried making a multi-position switch directly 
on the
> PCB like most handheld meters have?
> I can't gold plate it (yet).
> I think tinning would do for now.
> 
> I think of using a M6 screw as axis, just a washer on each side to 
make a  
> bearing
> of a hole in the PCB. Then take a piece of FR4 and make the "wiper" 
with  
> contacts
> of thin sheetmetal.
> The M6 screw would fit into a knob on the front panel.
> The "snap" action could be achieved by drilling holes in the PCB 
and  

> letting
> some spring catch them.
> 
> What do you think of that? stupid idea?
> 
> It seems i can't get rotrary switches with enough positions.
> Would be nice to be able to just make them how i want.
> After all it works in meters...
> 
> tell me what you thik about that.
> 
> ST

> > All one would need to do, is add some type of
> position encoding to the mouse
> > wheel, and one would be all set.
> 
> Nah,  that'll all be handled in the software.  As
> would the resolution,  for 
> that matter.

Could you elaborate as to how..... Especially in a
power down condition where upon one could loose track.
I take it, there is some trick, as many devices, use
relative rather than absolute positioning. WIthout a
home position to index off of, it would seem one could
loose the position of the pointer if it were turned
and power was off.... Or is it interrupt driven such
that the micro stays alive and wakes up should the
shaft be turned???

Ron


		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail

> 
> Could you elaborate as to how..... Especially in a
> power down condition where upon one could loose track.
> I take it, there is some trick, as many devices, use
> relative rather than absolute positioning. WIthout a
> home position to index off of, it would seem one could
> loose the position of the pointer if it were turned
> and power was off.... Or is it interrupt driven such
> that the micro stays alive and wakes up should the
> shaft be turned???
> 
> Ron

Maybe do like was already stated earlier and just use LEDs around the 
dial to indicate position.  Then don't put any position indicators on 
the dial itself.  Then when you turn the unit on, you could have it 
start in the "home" position no matter how much the dial itself has 
moved while off.


Or you could use a bunch of reed switches mounted in a circle and a 
magnet mounted to a disc on the shaft, but I'm sure that would be 
pretty pricy.

I've heard that green LEDs can act as photodetectors.  They are cheap 
if you could get them to work.  Then just put a bunch in a circle a 
detectors and one on a disc on the shaft as an emitter.  Anybody have 
any luck using LEDs as detectors?

Ben Englund

> Electro-mechanical encoders are quite cheap, they are made by ALPS and
> Bourne. One of those with a small MCU and a display will do what you  
> want.
>
> Leon

Thanks, but they are pretty useless for me.
I can't change what  i need, the IC expects a 2*10 switch.
It is a charger, and meant to be "jumpered" for cell count, but i want
a universal charger.

ST

>
> So what I gather, is a guy could put a good sized
> weight on the shaft to create inertia and give a
> perception of quality, then use optical interrupters
> to capture disc position?

Weight makes it continue to spin too long.
Better a good fit of a shaft and a bushing with oil between, like
a 10-turn pot knob. This gives you resistance for fine setting, and
you can spin it relatively fast still.

ST

On Mon, 30 Aug 2004 21:23:23 -0700 (PDT), ron amundson  
<mnphysicist@...> wrote:

>> > All one would need to do, is add some type of
>> position encoding to the mouse
>> > wheel, and one would be all set.
>>
>> Nah,  that'll all be handled in the software.  As
>> would the resolution,  for
>> that matter.
>
> Could you elaborate as to how..... Especially in a
> power down condition where upon one could loose track.
> I take it, there is some trick, as many devices, use
> relative rather than absolute positioning. WIthout a
> home position to index off of, it would seem one could
> loose the position of the pointer if it were turned
> and power was off.... Or is it interrupt driven such
> that the micro stays alive and wakes up should the
> shaft be turned???
>
> Ron
>

Usually "thumbwheel" encoders are only used incremental.
like up/down buttons. the absolute positions are stored in memory if needed
and often the thumbwheel is used to adjust many values, selected by  
buttons.

ST

On Tue, 31 Aug 2004 03:52:44 -0000, mikezcnc <eemikez@...> wrote:

> Could you use multiplexers? That would be more reliable and could
> have as many switches as you want. Mike
>

I could, but i'd rather not if i can avoid it.

ST

Stefan,

I think your idea will work, but have concern about using "sheet 
metal" for the wiper...

Is this for a production item, or just a few?

Because I'd sure use the phosphor bronze/beryllium copper/whatever of 
a "real" wiper(s) from the switch you CAN get. Then use your pcb to 
make the other half.

I believe you will have MUCH better success this way.

I'd also use an IGUS iglide circlip bushing for the switch "axle". 
These are made for thin sheet stock, and again will improve the 
smoothness and lifetime of your product/project. They're cheap and 
available worldwide as well.

Hope this helps,

Ballendo

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@g...> wrote:
> 
> >
> > What are you switching?  How many positions do you really 
_need_?  Have  
> > you
> > noticed that rotary switches are in large part going away,  or at 
least  
> > they
> > have been since the fifties?  There's a reason for that.  Feel 
free to  
> > take
> > this over to Electronics101 if we get too far off the topic for 
in here.
> >
> 
> I need 2*10 positions, and can only get 2*6 or 1*12.
> For selecting cell count on a battery charger (no current at all).
> Almost all handheld meters i've ever seen use rotary switches, and 
all i  
> took
> apart had "custom" printed contacts. Oh, yes, there was one rotten 
piece  
> of bad engineering
> that had four buttons instead. The damn thing had dozens of design 
flaws,  
> like blowing up
> each time you pulled the current plug....
> 
> I could try to pull some mechanical stunt and stack 2 1*12 switches.
> I don't know if that would be possibe with the switches i can get.
> But then, one switch for 2.50eur is not exactly cheap, not speaking 
of 2...
> 
> ST

Garrett,

Good tack.

You could also use an encoder style "pot" and a PIC...

These encoders are getting pretty inexpensive, and the PIC can take 
the input and "route" it as necessary.

Another option is the 0-9 thumbwheel switches. I see these in ALL the 
surplus places these days, and they ARE a rotary switch, just a 90 
degree orientation change. Most are stack-able, so the 2x10 need is 
easily met...

They are available both with 11 pins, and more commonly with 4/5 pins 
giving BCD output--which would work with your suggested parts.

Hope this helps,

Ballendo 

>"cybermace5" <cybermace5@y...> wrote:
> Would be nice to have a couple more details on the circuit this is
> going to be connected to.
> 
> Are you changing a resistance? Could you use a potentiometer and 
then
> work out some kind of mechanical detent? Why do you need a dual-pole
> switch? Have you looked at analog multiplexer chips? Digital
> potentiometer chips? Don't reject them without looking at them, they
> are not all expensive. A 16:1 analog mux/demux CD74HCT4067 costs 90
> cents in DIP, 48 cents in SOIC. Two of those give you your dual-pole
> circuit if you need it. They accept a simple 4-bit input to select
> which output goes where. You could provide that input with a
> 16-position hex switch, or a 4-bit counter (rollover at 12) operated
> by pushbuttons. Another 50-cent 74HC154 4:16 decoder could give you 
an
> LED feedback showing which output is selected.
> 
> It would be neat to have 12 LEDs surrounding a single central
> pushbutton which increments the LED position.
> 
> Or maybe you could just go for a rotary switch with fewer positions,
> like 6. Then you use a dual-pole switch to choose between high range
> and low range. Though even 6-position dual-pole rotaries are all 
very

> expensive unless you get lucky in surplus stores.
> 
> As for rolling your own? It'll just have to be trial and error.

In a message dated 8/30/2004 11:01:28 PM Pacific Standard Time,  
ben_englund@... writes:

I've  heard that green LEDs can act as photodetectors.  They are cheap 
if  you could get them to work.  Then just put a bunch in a circle a  
detectors and one on a disc on the shaft as an emitter.  Anybody have  
any luck using LEDs as detectors?

Ben  Englund


----------------------------------------
 
Yes,
 
I have used "plain jane" red leds as photodetectors. A voltages is produced  
across the leads in response to direct sunlight. It appears to be fairly  
sensitive.
 
A different voltage (lower) was read with the LED exposed to light, but in  
the shade opposed to direct sun light.
 
Ken


[Non-text portions of this message have been removed]

On Tuesday 31 August 2004 12:23 am, ron amundson wrote:
> > > All one would need to do, is add some type of position encoding to the
> > > mouse wheel, and one would be all set.

> > Nah,  that'll all be handled in the software.  As would the resolution, 
> > for that matter.

> Could you elaborate as to how..... Especially in a power down condition
> where upon one could loose track.

To what,  keep track of the "switch position"?  Store a value in a ram chip 
somewhere.  If it needs to stick around store it in a battery-backed ram.  
You can stuff an awful lot of info into a pretty small (2048 byte) part...

> I take it, there is some trick, as many devices, use relative rather than
> absolute positioning.

It's just a matter of keeping track of "where you were" and using the 
direction of rotation to figure out "where you're going" and how fast. What 
happens when you fire up software that uses a mouse?  I see a lot of stuff 
assuming that the pointer should be in the center of the screen...

> WIthout a home position to index off of, it would seem one could loose the
> position of the pointer if it were turned and power was off.... Or is it
> interrupt driven such that the micro stays alive and wakes up should the
> shaft be turned???

I think this is getting into system design now.  What you'd do depends on what 
you were trying to use this idea for...

On Tuesday 31 August 2004 01:39 am, ben_englund wrote:
> > Could you elaborate as to how..... Especially in a
> > power down condition where upon one could loose track.
> > I take it, there is some trick, as many devices, use
> > relative rather than absolute positioning. WIthout a
> > home position to index off of, it would seem one could
> > loose the position of the pointer if it were turned
> > and power was off.... Or is it interrupt driven such
> > that the micro stays alive and wakes up should the
> > shaft be turned???
> >
> > Ron
>
> Maybe do like was already stated earlier and just use LEDs around the
> dial to indicate position.  Then don't put any position indicators on
> the dial itself.  Then when you turn the unit on, you could have it
> start in the "home" position no matter how much the dial itself has
> moved while off.

Just like a mouse pointer starts in the middle of a screen.

> Or you could use a bunch of reed switches mounted in a circle and a
> magnet mounted to a disc on the shaft, but I'm sure that would be
> pretty pricy.

Radio Shack used to sell a bundle of those for a fairly low price,  as I 
recall.

> I've heard that green LEDs can act as photodetectors.  They are cheap
> if you could get them to work.  Then just put a bunch in a circle a
> detectors and one on a disc on the shaft as an emitter.  Anybody have
> any luck using LEDs as detectors?

I've never tried this,  never saw the need.  If you want some cheap (free?) 
photodetctors and emitters,  take any trashed floppy drive (that people would 
normally just throw away) and grab the sensors out of it.  Some will use a 
mechanical switch for the ones along the edge of the disk package, some will 
use optical parts,  but almost all that I can figure will use an optical pair 
for the index sensor,  so you get at least one.  There can be as many as four 
pairs in there (disk type, write protect, index, and home position).  Or take 
apart a dead mouse,  there will be at least four pairs of sensors in there...

On Tuesday 31 August 2004 05:28 am, ballendo wrote:

> Another option is the 0-9 thumbwheel switches. I see these in ALL the
> surplus places these days, and they ARE a rotary switch, just a 90
> degree orientation change. Most are stack-able, so the 2x10 need is
> easily met...
>
> They are available both with 11 pins, and more commonly with 4/5 pins
> giving BCD output--which would work with your suggested parts.

I have some of these,  and have yet to come up with something I want to use 
them in...

--- In Homebrew_PCBs@yahoogroups.com, "Roy J. Tellason"
<rtellason@b...> wrote:
> On Tuesday 31 August 2004 05:28 am, ballendo wrote:
> 
> > Another option is the 0-9 thumbwheel switches. I see these in ALL the
> > surplus places these days, and they ARE a rotary switch, just a 90
> > degree orientation change. Most are stack-able, so the 2x10 need is
> > easily met...
> >
> > They are available both with 11 pins, and more commonly with 4/5 pins
> > giving BCD output--which would work with your suggested parts.
> 
> I have some of these,  and have yet to come up with something I want
to use 
> them in...


Handy for setting a device ID. Good when you have several devices on a
bus, or wireless, that you want to identify.

On Tuesday 31 August 2004 03:11 am, Stefan Trethan wrote:
> > Electro-mechanical encoders are quite cheap, they are made by ALPS and
> > Bourne. One of those with a small MCU and a display will do what you
> > want.
> >
> > Leon
>
> Thanks, but they are pretty useless for me.
> I can't change what  i need, the IC expects a 2*10 switch.
> It is a charger, and meant to be "jumpered" for cell count, but i want
> a universal charger.

What IC is this?  Do you have a link for a data sheet?

On Tuesday 31 August 2004 09:55 am, cybermace5 wrote:
> --- In Homebrew_PCBs@yahoogroups.com, "Roy J. Tellason"
>
> <rtellason@b...> wrote:
> > On Tuesday 31 August 2004 05:28 am, ballendo wrote:
> > > Another option is the 0-9 thumbwheel switches. I see these in ALL the
> > > surplus places these days, and they ARE a rotary switch, just a 90
> > > degree orientation change. Most are stack-able, so the 2x10 need is
> > > easily met...
> > >
> > > They are available both with 11 pins, and more commonly with 4/5 pins
> > > giving BCD output--which would work with your suggested parts.
> >
> > I have some of these,  and have yet to come up with something I want
> > to use them in...
>
> Handy for setting a device ID. Good when you have several devices on a
> bus, or wireless, that you want to identify.

Yeah,  but that's a lot of bulk and panel space (and rectangular holes needed 
which I can't cut just yet :-) for something you'd set once and forget about, 
which is why those functions normally seem to use jumper blocks...

At 04:27 PM 8/30/2004, Stefan Trethan wrote:

>I need 2*10 positions, and can only get 2*6 or 1*12.
>For selecting cell count on a battery charger (no current at all).

<http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836083832&rd=1>
<http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836173808&rd=1>

There are others as well . . .
<http://search.ebay.com/rotary-switch_W0QQcatrefZC6QQfromZR7QQnojsprZyQQpfidZ0QQsacategoryZQ2d1QQsacqtyopZgeQQsacurrencyZ0QQsaslopZ1QQsatitleZrotaryQ20switchQ2aQQsocolumnlayoutZ3QQsofocusZbsQQsopostalZT5SQ201K8QQsorecordsperpageZ50QQsosortorderZ1QQsosortpropertyZ1QQsspagenameZhQ3ahQ3afitemQ3aUS>

dwayne

-- 
Dwayne Reid   <dwayner@...>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 20 years of Engineering Innovation (1984 - 2004)
  .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
     `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

> What IC is this?  Do you have a link for a data sheet?
>

<http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1666>

i found another solution.
I will hardwire the chip for one cell, and then use a voltage divider for
the cell number.
I could use a 1*12 pos. switch with a resistor chain, or i could use a  
10-turn
pot (helipot), which is what i will do because i have one already.

The result it not only a solved problem, but i can also use this voltage
for a treshold voltage for a discharge circuit, which i couldn't if i use
the chargers internal divider.

I thought about the homebrew switch, and even made a layout.
one could wire the switch as needed from only one component in the library.
The bad thing is it is relatively big (around 6cm diameter).

Assembly would be as follows:
take a M6 screw.
Make the "wiper" from a piece of pcb, solder on suitable spring contacts  
(like
the fingers from a GND shield spring - sheetmetal) in the right places.
drill a 6mm hole in the center and solder to the screw (all the way up  
against
the screwhead. Put a washer on the screw, then stick it through the PCB.
put another washer on this side, a nut. now take another strip of FR4,  
solder
a "ball" type item on the end, and affix it with another nut to the screw  
(which
also counters the first nut).

the PCB would have two concentric rings of contacts, which are connected by
your contact, and outside of them a ring of
holes which suit the "ball" as a snap-in mechanism. To set endstops you'd  
simply
put a M3 screw in the appropriate holes.
If you couldn't picture the instructions above note that the contact wiper
is on one side of the PCB, and the "snap in" mechanism slides on the other  
side.
this is required if you don't want your snap in to slide over connecting  
tracks.

Not too hard to manufacture, and i bet it would work, but as said i will  
probably
not need it for this project any more.

ST

On Tue, 31 Aug 2004 08:40:44 -0600, Dwayne Reid <dwayner@...>  
wrote:

> At 04:27 PM 8/30/2004, Stefan Trethan wrote:
>
>> I need 2*10 positions, and can only get 2*6 or 1*12.
>> For selecting cell count on a battery charger (no current at all).
>
> <http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836083832&rd=1>
> <http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836173808&rd=1>
>
> There are others as well . . .
> <http://search.ebay.com/rotary-switch_W0QQcatrefZC6QQfromZR7QQnojsprZyQQpfidZ0QQsacategoryZQ2d1QQsacqtyopZgeQQsacurrencyZ0QQsaslopZ1QQsatitleZrotaryQ20switchQ2aQQsocolumnlayoutZ3QQsofocusZbsQQsopostalZT5SQ201K8QQsorecordsperpageZ50QQsosortorderZ1QQsosortpropertyZ1QQsspagenameZhQ3ahQ3afitemQ3aUS>
>
> dwayne
>


Thanks, but ebay was the first i looked.
I assume you either don't know i'm not in the US or you don't know what  
i'd have
to pay for shipping and taxes...
Nothing useful at ebay in a country where i'm willing to buy.

ST

At 11:39 PM 8/30/2004, ben_englund wrote:

>Maybe do like was already stated earlier and just use LEDs around the
>dial to indicate position.  Then don't put any position indicators on
>the dial itself.  Then when you turn the unit on, you could have it
>start in the "home" position no matter how much the dial itself has
>moved while off.

Modern digital audio consoles do this - the Yamaha PM-1D uses 32 LEDs per 
rotary encoder to indicate 64 discrete positions (all off, 1 on, 1+2 on, 2 
on, etc).  Its darned cool to punch up a different preset scene and watch 
all the indicators snap to their new positions.

dwayne

-- 
Dwayne Reid   <dwayner@...>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 20 years of Engineering Innovation (1984 - 2004)
  .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
     `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

Stefan,

ALPHA makes a 2P12T rotary switch that is $1.57 for quantity 1, at www.mouser.com, with Mouser part number 10WR212.

That one is make-before-break ("shorting"), with solder lugs. I don't see any 2P12T (or 10T) Alpha-made (i.e. low-cost)  switches at mouser.com that are non-shorting, or that are pcb mount.

For a lot more $, both Electroswitch and NKK have many models that would probably meet your need. They have switches listed at mouser.com that go up to (just to name some of the extremes) 2P23T, 4P17T, 6P12T, 9P4T, 8P6T, 12P4T, and even have a model that will handle 30 Amps that goes up to 5P11T (for only $114.00!).

I am sure that you could eventually get your homebrew pcb-based rotary switch idea implemented. But the Alpha switch would be much cheaper/easier/faster, assuming that it meets your specs.

Tom Gootee

http://www.fullnet.com/u/tomg

-----------------------------------



----------

From: 	Homebrew_PCBs@yahoogroups.com[SMTP:Homebrew_PCBs@yahoogroups.com]
Sent: 	Tuesday, August 31, 2004 8:55 AM
To: 	Homebrew_PCBs@yahoogroups.com
Subject: 	[Homebrew_PCBs] Digest Number 801

   Date: Mon, 30 Aug 2004 21:12:52 +0200
   From: "Stefan Trethan" <stefan_trethan@...>
Subject:  homebrew rotary multi-position switch

Hi,


has one of you ever tried making a multi-position switch directly on the
PCB like most handheld meters have?
I can't gold plate it (yet).
I think tinning would do for now.

I think of using a M6 screw as axis, just a washer on each side to make a  
bearing
of a hole in the PCB. Then take a piece of FR4 and make the "wiper" with  
contacts
of thin sheetmetal.
The M6 screw would fit into a knob on the front panel.
The "snap" action could be achieved by drilling holes in the PCB and  
letting
some spring catch them.

What do you think of that? stupid idea?

It seems i can't get rotrary switches with enough positions.
Would be nice to be able to just make them how i want.
After all it works in meters...

tell me what you thik about that.

ST


[Non-text portions of this message have been removed]

> > 
> > It seems i can't get rotrary switches with enough
> > positions.
> > Would be nice to be able to just make them how i
> > want.
> > After all it works in meters...
> And other place, a rotary switch is generally not cost
> effective, but it is the most intuitive, at least for
> those of us in the older generation. The  youngsters
> may have grown up with bottom remotes etc, but I still
> have a few tv sets with rotary volume and channel
> select. EVen Fluke tried to go away from the rotary
> switch in the late 80's.... the product failed. They
> tried again with a lower priced model in the late
> nineties. I don't think its doing to well sales wise.

The fluke meters were just too feature poor.  I don't need much for 
field work, voltage, continunity and current readings from contorl 
signals.  But they didn't offer 4-20mA as an input when they came out.

The scaled down units had less features than the radioshack models at 
the time.

I remeber looking for an AMP unit at about the same time.  It had a 
smaller case and digital display, but no one carried it.

Dave

> I need 2*10 positions, and can only get 2*6 or 1*12.

Electroswitch has 12-position switches that take a stop to reduce the
available positions -- you can use a 2x12 as a 2x10 by moving the stop a
couple of positions.  You might be able to get a free sample from your local
distributor.

-Alex

At 08:45 AM 8/31/2004, Stefan Trethan wrote:
> >
> > 
> <http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836083832&rd=1>
> > 
> <http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26211&item=3836173808&rd=1>
>
>Thanks, but ebay was the first i looked.
>I assume you either don't know i'm not in the US or you don't know what
>i'd have to pay for shipping and taxes...

I had assumed that you were not from North America from your email 
address.  That's why I included links only to sellers who ship 
worldwide.  In particular, the 2nd link above is selling from Lithuania.

dwayne

-- 
Dwayne Reid   <dwayner@...>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 20 years of Engineering Innovation (1984 - 2004)
  .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
     `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

> I had assumed that you were not from North America from your email
> address.  That's why I included links only to sellers who ship
> worldwide.  In particular, the 2nd link above is selling from Lithuania.
>
> dwayne
>

i see.. thanks.

ST

> i see.. thanks.
> 
> ST


A big lesson: there is no such thing as "I already checked eBay." It's
like watching a road for a certain car to go by, you can't look once
and then say you've already checked.

On Thu, 02 Sep 2004 14:17:43 -0000, cybermace5 <cybermace5@...>  
wrote:

>> i see.. thanks.
>>
>> ST
>
>
> A big lesson: there is no such thing as "I already checked eBay." It's
> like watching a road for a certain car to go by, you can't look once
> and then say you've already checked.
>

I can stop looking.
also, when i know the car takes 4 to 10 days to pass, from the moment
i can first see it, and i need the switch *now* i can say with one look
it isn't going to come around the next 4 to 10 days.
I already said in another post the problem is solved in another way
and the switch is no longer required.

I constantly monitor ebay austria and ebay germany for things i might like,
and quite regularly buy.

you surely understand my reluctance to import a simple switch from abroad.
Again, the problem is solved without the switch. thanks.

ST

Hi,

I tried to peel off the copper at one side of a glassfiber pcb.
I used a knife to separate the first layer of glass (with the copper)
 from the rest, at the edge, and then peeled.

It is a bit tedious but it is possible to do it.
You must be careful not to let the edges break.
I think it might help to bash the edges with a hammer, to break up the
bond, and later cut off 5mm, but i haven't tried.

You end up with a (slightly thinner) single sided pcb and a very thin, very
nice, flexible pcb.

The surface has the "glassfiber matt" structure of course, which looks
quite nice but is not useable for TT of component legend, which is the
reason why i won't use that method. The grain is so course you can't
really get the transfer stick, and the toner is squeezed into the
low areas, which makes it unreadable.

Peeling the copper alone doesn't work well, it just breaks into smaller
flakes which can't be removed at all.

The only other method to use i can think of is etching the copper.
I would have liked to find another way because it uses a lot of etchant,  
and
a lot of time, covering the one side and all.

Until now i just made the transfer on one side, and had both sides etched  
at once,
the unwanted one simply completely. the downside of that is i can't  
transfer
the component legend before etching. It takes longer, and is much harder  
to align.
I would like to do the "simple" PCBs (one layer, and component legend on  
top) with
one transfer run.

Well, let me know if you have found a better way.

ST

----- Original Message ----- 

From: "Stefan Trethan" <stefan_trethan@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Wednesday, September 08, 2004 10:32 AM
Subject: [Homebrew_PCBs] Removing copper at one side


> Hi,
>
> I tried to peel off the copper at one side of a glassfiber pcb.
> I used a knife to separate the first layer of glass (with the copper)
>  from the rest, at the edge, and then peeled.

I've removed small areas of copper by heating it with a large bit on my
soldering iron, after scoring it with a Stanley knife. It peeled off very
easily.

Leon

> I've removed small areas of copper by heating it with a large bit on my
> soldering iron, after scoring it with a Stanley knife. It peeled off very
> easily.
>
> Leon

That could just be a little bit too tedious for big areas.
I tried heating with the heat gun and wasn't really convinced by the  
results...

I guess i will etch the boards, further mechanical attempts were not  
leading
to good results.
I'll see if  sticky tape is tight enough where it overlaps or if i need to  
use paint...

ST

----- Original Message ----- 

From: "Stefan Trethan" <stefan_trethan@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Wednesday, September 08, 2004 4:17 PM
Subject: Re: [Homebrew_PCBs] Removing copper at one side


>
> > I've removed small areas of copper by heating it with a large bit on my
> > soldering iron, after scoring it with a Stanley knife. It peeled off
very
> > easily.
> >
> > Leon
>
> That could just be a little bit too tedious for big areas.
> I tried heating with the heat gun and wasn't really convinced by the
> results...
>
> I guess i will etch the boards, further mechanical attempts were not
> leading
> to good results.
> I'll see if  sticky tape is tight enough where it overlaps or if i need to
> use paint...

I've used both paint and masking tape. The latter allows some leakage, so a
waterproof tape
would be better. I've also used double-sided material with the unexposed
side as a ground plane. That
works very well.

Leon

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@g...> wrote:
> 
> > I've removed small areas of copper by heating it with a large bit 
on my
> > soldering iron, after scoring it with a Stanley knife. It peeled 
off very
> > easily.
> >
> > Leon
> 
> That could just be a little bit too tedious for big areas.
> I tried heating with the heat gun and wasn't really convinced by 
the  
> results...
> 
> I guess i will etch the boards, further mechanical attempts were 
not  
> leading
> to good results.
> I'll see if  sticky tape is tight enough where it overlaps or if i 
need to  
> use paint...
> 
> ST

I found that FR-4 has an extreemly tight bond with the copper and 
some phenolics will allow you to peal away huge sheets.

With my etcher I route the traces isolation and then peal away huge 
areas of copper.  Time consuming to say the least.

My very white phenolic allowed me to peal the copper with minimal 
effort and the sufrace is smooth and looks untouched.

IF I had to do a large board of FR-4 I'd considder spray paint or 
floor wax.  both would be easy to put on and remove enough for future 
etching.

Dave