• Hardware is HARD

    From The Natural Philosopher@3:770/3 to All on Wed Oct 25 13:57:20 2023
    And to think I used to do this for a living...
    My Pi ZERO W heating controller is alive, but it took some forceps...

    First Problem:

    Not all transistors are created equal. I used obsolete transistors that
    I used to use years ago - BC182... but they didn't work. Turns out
    today's BC182 has a different pinout from what I was used to - *BC182L*
    was the solution. That worked.

    Second Problem:

    Pi connectors. I bought kit to make up Pi female connectors to go from
    the pins on one board to pins on another. I wasted a whole day and a
    half and was completely unable to make a satisfactory lead up , either
    by crimping or by soldering.
    Gave up and bought jumper leads.

    Third problem:

    All earths are not created equal. I wanted to use a 5 way connector to
    pins 31,33,35,37 with pin 39 as power and signal earth on the Pi Zero W.

    I don't know what pin 39 is connected to, but it sure ain't earth. maybe
    via a high resistance, but not enough to power a Pi. I used pin 6, and
    that worked...

    There are still some software issues to resolve, but the unit now has a
    nice 3D printed case and the relays all click and the LEDs all light...

    *phew*.

    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0027.JPG http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

    --
    It is the folly of too many to mistake the echo of a London coffee-house
    for the voice of the kingdom.

    Jonathan Swift

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Andy Burns@3:770/3 to The Natural Philosopher on Wed Oct 25 14:49:12 2023
    The Natural Philosopher wrote:

    the unit now has a nice 3D printed case http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

    If your friend can enable monotonic infill, it would avoid those
    diagonal stripes where it navigates around the LED holes ...

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From The Natural Philosopher@3:770/3 to Andy Burns on Wed Oct 25 15:51:12 2023
    On 25/10/2023 14:49, Andy Burns wrote:
    The Natural Philosopher wrote:

    the unit now has a nice 3D printed case
    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

    If your friend can enable monotonic infill, it would avoid those
    diagonal stripes where it navigates around the LED holes ...

    I will shortly be getting my own printer. I would appreciate any tips
    like these you can give me!

    --
    "Socialist governments traditionally do make a financial mess. They
    always run out of other people's money. It's quite a characteristic of them"

    Margaret Thatcher

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Computer Nerd Kev@3:770/3 to The Natural Philosopher on Thu Oct 26 07:23:12 2023
    The Natural Philosopher <tnp@invalid.invalid> wrote:

    Pi connectors. I bought kit to make up Pi female connectors to go from
    the pins on one board to pins on another. I wasted a whole day and a
    half and was completely unable to make a satisfactory lead up , either
    by crimping or by soldering.
    Gave up and bought jumper leads.

    If they were ribbon cable connectors then I can understand that -
    you really want the proper tool for pushing those together when
    they're that wide.

    I've got a bag full of old PC IDE cables, so it's easy just to
    attack one of those instead because they're also 40 pin.

    All earths are not created equal. I wanted to use a 5 way connector to
    pins 31,33,35,37 with pin 39 as power and signal earth on the Pi Zero W.

    I don't know what pin 39 is connected to, but it sure ain't earth. maybe
    via a high resistance, but not enough to power a Pi. I used pin 6, and
    that worked...

    Pin 6 might just be connected by a tiny little track on the Pi
    Zero's PCB because it's only supposed to be a signal ground. I go
    to the trouble of grounding all the ground pins on the GPIO. But
    just for driving relays I guess you can cut some corners, just
    not too many apparantly.

    --
    __ __
    #_ < |\| |< _# | Note: I won't see posts made from Google Groups |

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Andy Burns@3:770/3 to The Natural Philosopher on Thu Oct 26 08:30:04 2023
    The Natural Philosopher wrote:

    I bought kit to make up Pi female connectors to go from the pins on one
    board to pins on another. I wasted a whole day and a half  and was completely unable to make a satisfactory lead up ,  either by crimping
    or by soldering.
    Gave up and bought jumper leads.

    Don't think I'd trust a bunch of DuPont wires long term ... at least hot
    snot them in place.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From The Natural Philosopher@3:770/3 to All on Thu Oct 26 09:12:00 2023
    On 26/10/2023 05:27, 56d.1152 wrote:
    On 10/25/23 8:57 AM, The Natural Philosopher wrote:
    And to think I used to do this for a living...
    My Pi ZERO W heating controller is alive, but it took some forceps...

    First Problem:

    Not all transistors are created equal. I used obsolete transistors
    that I used to use years ago - BC182... but they didn't work.  Turns
    out today's BC182 has a different pinout from what I was used to -
    *BC182L* was the solution. That worked.

      IF possible, stick to super-popular "legacy" transistors.
      2n2222 are very versatile, 3809s are ok if you don't need
      to handle much power.

      There are some 'logic level' MOSFET-family out there that
      are very good for higher power levels. Again, stick to the
      classics if possible.


    Second Problem:

    Pi connectors. I bought kit to make up Pi female connectors to go from
    the pins on one board to pins on another. I wasted a whole day and a
    half  and was completely unable to make a satisfactory lead up ,
    either by crimping or by soldering.
    Gave up and bought jumper leads.


      'Connections' ALWAYS seem to be an unnecessary bitch.


    Third problem:

    All earths are not created equal. I wanted to use a 5 way connector to
    pins 31,33,35,37 with pin 39 as power and signal  earth on the Pi Zero W.

      Heh, heh ... yep, 'earths' are NOT all created equal.
      Look into RS-4xx data connections for fun ... some
      'earths' can cause HUGE DEADLY current flow  :-)u

      Clue, two "ground" connections even 500-1000 feet
      apart are NOT gonna be at the same potential. Oft
      wondered if you can get actual usable power from
      that ...... can 0.25 volts dif, at a potential
      100-1000 amps, be converted to higher voltage/
      lower amps these days ?


    I don't know what pin 39 is connected to, but it sure ain't earth.
    maybe via a high resistance, but not enough to power a Pi. I used pin
    6, and that worked...

      The 7th pin, on the top edge, is a fairly reliable ground.
      Be cautious otherwise unless it's something stupid like
      powering a fan.

    There are still some software issues to resolve, but the unit now has
    a nice 3D printed case and the relays all click and the LEDs all light...

    *phew*.

    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0027.JPG
    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

      I'm slowly putting together a home-security system
      with all the goodies - including opto-isolated
      inputs. It is gonna run on a Pi4 or now maybe on
      a Pi5, sporting a usable web interface. Switches,
      detectors, video ... all in one.


    Good project.

    I have all the wiring here for a home security system but I never
    installed one because I retired at that point and the house was never empty

    --
    “Puritanism: The haunting fear that someone, somewhere, may be happy.”

    H.L. Mencken, A Mencken Chrestomathy

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From The Natural Philosopher@3:770/3 to Andy Burns on Thu Oct 26 09:16:56 2023
    On 26/10/2023 08:30, Andy Burns wrote:
    The Natural Philosopher wrote:

    I bought kit to make up Pi female connectors to go from the pins on
    one board to pins on another. I wasted a whole day and a half  and was
    completely unable to make a satisfactory lead up ,  either by crimping
    or by soldering.
    Gave up and bought jumper leads.

    Don't think I'd trust a bunch of DuPont wires long term ... at least hot
    snot them in place.

    Used 100% to connect model plane servos. They are pretty good. That
    units going to be screwed onto a wall and never moved again.

    I might add a zip tie or two.

    I wouldn't trust it in a car though - there I would follow your advice


    --
    "And if the blind lead the blind, both shall fall into the ditch".

    Gospel of St. Mathew 15:14

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From David Higton@3:770/3 to Computer Nerd Kev on Thu Oct 26 16:16:32 2023
    In message <65398740@news.ausics.net>
    not@telling.you.invalid (Computer Nerd Kev) wrote:

    The Natural Philosopher <tnp@invalid.invalid> wrote:

    Pi connectors. I bought kit to make up Pi female connectors to go from
    the pins on one board to pins on another. I wasted a whole day and a
    half and was completely unable to make a satisfactory lead up , either
    by crimping or by soldering. Gave up and bought jumper leads.

    If they were ribbon cable connectors then I can understand that - you
    really want the proper tool for pushing those together when they're that wide.

    I've found it not too difficult to do ribbon insulation displacement
    cables if you have a vice, but the caveats I can think of are:

    Use a VERY sharp pair of scissors to cut the cable. If it isn't cut
    cleanly, there may be little bits of copper core hanging out, just
    waiting to short to a neighbour. (Been there.)

    Connectors need to be crunched on at very close to 90 degrees to the
    cable. It's surprising how little margin for error there is, or else
    there may be shorts between adjacent cores.

    The vice's jaws need to cover the whole connector and to close parallel.
    If the jaw surfaces are serrated, some dense cardboard may help, but you
    will need them held in place simply to avoid having to keep so many bits
    all in the right place as you tighten the vice.

    David

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Martin Gregorie@3:770/3 to David Higton on Thu Oct 26 16:03:52 2023
    On Thu, 26 Oct 2023 16:16:33 +0100, David Higton wrote:

    In message <65398740@news.ausics.net>
    not@telling.you.invalid (Computer Nerd Kev) wrote:

    The Natural Philosopher <tnp@invalid.invalid> wrote:

    Pi connectors. I bought kit to make up Pi female connectors to go
    from the pins on one board to pins on another. I wasted a whole day
    and a half and was completely unable to make a satisfactory lead up
    , either by crimping or by soldering. Gave up and bought jumper
    leads.

    If they were ribbon cable connectors then I can understand that - you
    really want the proper tool for pushing those together when they're
    that wide.

    I've found it not too difficult to do ribbon insulation displacement
    cables if you have a vice, but the caveats I can think of are:

    Use a VERY sharp pair of scissors to cut the cable. If it isn't cut
    cleanly, there may be little bits of copper core hanging out, just
    waiting to short to a neighbour. (Been there.)

    Connectors need to be crunched on at very close to 90 degrees to the
    cable. It's surprising how little margin for error there is, or else
    there may be shorts between adjacent cores.

    The vice's jaws need to cover the whole connector and to close parallel.
    If the jaw surfaces are serrated, some dense cardboard may help, but you
    will need them held in place simply to avoid having to keep so many bits
    all in the right place as you tighten the vice.

    Curiosity: would it help to use one of the spray-on contact cements or one
    of the cyanoacrylates (e.g. Zap) to hold things in place while they're in
    the vice?

    --

    Martin | martin at
    Gregorie | gregorie dot org

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Ahem A Rivet's Shot@3:770/3 to The Natural Philosopher on Thu Oct 26 17:38:18 2023
    On Wed, 25 Oct 2023 15:51:12 +0100
    The Natural Philosopher <tnp@invalid.invalid> wrote:

    I will shortly be getting my own printer. I would appreciate any tips
    like these you can give me!

    I have a few :) You may know some of this.

    This one is really important - when you open a reel of filament you will find that the end is clipped in place or fed through holes in the
    sides. Once you release it DO NOT LET GO of it, not even for an instant,
    until it is securely fixed in the printer's feed mechanism. If the filament ever springs back onto the roll it WILL tangle (Sod's law applies) and then
    it will jam during a print ruining the print - usually after hours of
    printing.

    All printing filaments are hygroscopic and need to be protected
    from humidity. Vacuum bags are essential and a drying box is useful.

    When choosing a printer to buy make sure that you understand how
    the bed levelling process works and that you are satisfied that it is
    effective and reliable - bed levelling is the single most important thing
    to get right before printing.

    FWIW I currently have a FlashForge Creator Pro and am reasonably
    happy with it (it certainly isn't perfect and there are many equally good
    and better options - oh and worse too) - next time I feel like spending
    money on a printer it will be one of these (no connection I just like the design and reviews):

    https://ratrig.com/3d-printers/rat-rig-kits.html.

    --
    Steve O'Hara-Smith
    Odds and Ends at http://www.sohara.org/
    Host: Beautiful Theory meet Inconvenient Fact
    Obit: Beautiful Theory died today of factual inconsistency

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Chris Elvidge@3:770/3 to The Natural Philosopher on Thu Oct 26 18:30:16 2023
    On 25/10/2023 15:51, The Natural Philosopher wrote:
    On 25/10/2023 14:49, Andy Burns wrote:
    The Natural Philosopher wrote:

    the unit now has a nice 3D printed case
    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

    If your friend can enable monotonic infill, it would avoid those
    diagonal stripes where it navigates around the LED holes ...

    I will shortly be getting my own printer. I would appreciate any tips
    like these you can give me!


    Look on Tom's Hardware. https://www.tomshardware.com/best-picks/best-3d-printers


    --
    Chris Elvidge, England
    I WILL NOT YELL "SHE'S DEAD" DURING ROLL CALL

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to The Natural Philosopher on Thu Oct 26 23:42:00 2023
    On 10/26/23 4:12 AM, The Natural Philosopher wrote:
    On 26/10/2023 05:27, 56d.1152 wrote:
    On 10/25/23 8:57 AM, The Natural Philosopher wrote:
    And to think I used to do this for a living...
    My Pi ZERO W heating controller is alive, but it took some forceps...

    First Problem:

    Not all transistors are created equal. I used obsolete transistors
    that I used to use years ago - BC182... but they didn't work.  Turns
    out today's BC182 has a different pinout from what I was used to -
    *BC182L* was the solution. That worked.

       IF possible, stick to super-popular "legacy" transistors.
       2n2222 are very versatile, 3809s are ok if you don't need
       to handle much power.

       There are some 'logic level' MOSFET-family out there that
       are very good for higher power levels. Again, stick to the
       classics if possible.


    Second Problem:

    Pi connectors. I bought kit to make up Pi female connectors to go
    from the pins on one board to pins on another. I wasted a whole day
    and a half  and was completely unable to make a satisfactory lead up
    , either by crimping or by soldering.
    Gave up and bought jumper leads.


       'Connections' ALWAYS seem to be an unnecessary bitch.


    Third problem:

    All earths are not created equal. I wanted to use a 5 way connector
    to pins 31,33,35,37 with pin 39 as power and signal  earth on the Pi
    Zero W.

       Heh, heh ... yep, 'earths' are NOT all created equal.
       Look into RS-4xx data connections for fun ... some
       'earths' can cause HUGE DEADLY current flow  :-)u

       Clue, two "ground" connections even 500-1000 feet
       apart are NOT gonna be at the same potential. Oft
       wondered if you can get actual usable power from
       that ...... can 0.25 volts dif, at a potential
       100-1000 amps, be converted to higher voltage/
       lower amps these days ?


    I don't know what pin 39 is connected to, but it sure ain't earth.
    maybe via a high resistance, but not enough to power a Pi. I used pin
    6, and that worked...

       The 7th pin, on the top edge, is a fairly reliable ground.
       Be cautious otherwise unless it's something stupid like
       powering a fan.

    There are still some software issues to resolve, but the unit now has
    a nice 3D printed case and the relays all click and the LEDs all
    light...

    *phew*.

    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0027.JPG
    http://vps.templar.co.uk/Odds%20and%20Ends/BCD_0026.JPG

       I'm slowly putting together a home-security system
       with all the goodies - including opto-isolated
       inputs. It is gonna run on a Pi4 or now maybe on
       a Pi5, sporting a usable web interface. Switches,
       detectors, video ... all in one.


    Good project.

    It's fun - AND very useful. I've been making progress,
    but I'll have more time when fully retired. Recently
    picked up a multi-relay board meant for a Beaglebone
    Black - but adaptable. Perfect for alarm sounders
    and even turning on room lights. Sometimes you need
    AMPERAGE.

    Retiring, I need to know what's going bump in
    the night AND be able to see/know what's going
    on when I'm not home. Do NOT trust commercial
    providers one bit. So ...

    Look into the popular commercial alarm systems.
    I think the Honeywell/Ademco ones are the most
    popular. They're incredibly versatile, but they
    employ HORRIBLE programming approaches oriented
    around terminals with like 25x2 or 25x4 LED
    displays. It's clunkier than a FORTH pgm, and
    I speak from experience. A proper, if simple,
    web interface is THE way to go. Been there, done
    those, so ....

    Let my tombstone read "I did it MY way" :-)

    Ok, what's it cost these days to be deep-frozen,
    $300k ??? Kinda like the lottery, if you DON'T
    play you CAN'T win ..........

    I have all the wiring here for a home security system but I never
    installed one because I retired at that point and the house was never empty

    Yea, but SOMETIMES you have to go away, even if
    to a stupid dental appt. It'd be nice to have a
    system where you COULD be informed of odd goings-on.
    Easy these days to send a text over the web.

    Alas, actual "human-form detection" worth a crap is
    still kinda the realm of boards much more powerful
    than any Pi. Yea, yea, some packages, even cams,
    CLAIM to offer something like that, but read the
    fine print. 95+ percent of your 'positives' are gonna
    be FALSE. A bush in the wind WILL kinda emulate a
    human-form pattern a dozen times a day. Seen it.

    The only thing worse than a false negative are lots
    of false positives. Hey, WAY back in the day, I
    remember when the public schools decided to have
    the cops bring sniffer dogs to find lockers that
    smelled like reefer. So, the potheads made a point
    to put a little reefer dust and/or roaches in the
    lockers of the Honor Roll students. The program
    was terminated immediately :-)

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to The Natural Philosopher on Thu Oct 26 23:44:48 2023
    On 10/26/23 4:16 AM, The Natural Philosopher wrote:
    On 26/10/2023 08:30, Andy Burns wrote:
    The Natural Philosopher wrote:

    I bought kit to make up Pi female connectors to go from the pins on
    one board to pins on another. I wasted a whole day and a half  and
    was completely unable to make a satisfactory lead up ,  either by
    crimping or by soldering.
    Gave up and bought jumper leads.

    Don't think I'd trust a bunch of DuPont wires long term ... at least
    hot snot them in place.

    Used 100% to connect model plane servos. They are pretty good. That
    units going to be screwed onto a wall and never moved again.

    I might add a zip tie or two.

    I wouldn't trust it in  a car though - there I would follow your advice


    Solder IS the best.

    But, remember "wire wrap" ? Still have the tools
    for doing that. The more-extensive wrapping CAN
    be as reliable as soldered connections.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From The Natural Philosopher@3:770/3 to All on Fri Oct 27 05:32:34 2023
    On 27/10/2023 04:44, 56d.1152 wrote:


      Solder IS the best.

    No, actually it often isn't...

    The problem with solder on *stranded* wire is that it wicks up the wire
    and reduces it's flexibility by binding the strands together: That
    creates a stress concentration where it stops and a likely cause of
    fracture failure under vibration. That is why crimps are the choice for
    cheaper connections in high stress environments (e.g. automotive) or if
    you must solder you absolutely need to mechanically support the wire
    upstream of the solder joint.

      But, remember "wire wrap" ?

    I was trained, as an apprentice, on military spec hardware construction:
    Wire wrap was one system that was used extensively in early digital
    hardware backplane prototyping.

    Once the design was stabilised they tended to replace all that with a 2
    layer PCB.

    Still have the tools
      for doing that. The more-extensive wrapping CAN
      be as reliable as soldered connections.

    Yes. especially in high vibration environments where components are not mechanically connected except by their leads or pins.



    --
    "Anyone who believes that the laws of physics are mere social
    conventions is invited to try transgressing those conventions from the
    windows of my apartment. (I live on the twenty-first floor.) "

    Alan Sokal

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Ahem A Rivet's Shot@3:770/3 to 56d.1152@ztq9.net on Fri Oct 27 07:42:46 2023
    On Thu, 26 Oct 2023 23:44:48 -0400
    "56d.1152" <56d.1152@ztq9.net> wrote:

    But, remember "wire wrap" ? Still have the tools
    for doing that. The more-extensive wrapping CAN
    be as reliable as soldered connections.

    A good wire wrap connection (done with a power tool - we always
    used Gardner-Denver) is gas tight and extremely reliable as well as being faster and easier to modify than soldering.

    --
    Steve O'Hara-Smith
    Odds and Ends at http://www.sohara.org/
    Host: Beautiful Theory meet Inconvenient Fact
    Obit: Beautiful Theory died today of factual inconsistency

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Martin Gregorie@3:770/3 to TimS on Fri Oct 27 09:58:08 2023
    On 27 Oct 2023 09:29:03 GMT, TimS wrote:

    I was told, back in the day, that it's more reliable. Many mainframes
    had wire-wrapped backplanes. I recall seeing engineers come in during maintenance time to replace wires due to upgrades or bug fixes; their
    being wire-wrapped made it rather easier.

    If you can find a copy, Tracey Kidder's book "The Soul of a New Machine"
    is a nice read. He is a professional writer who was embedded in the team
    who designed and built the prototype Data General 'Eagle's. This was one
    of the first 32 bit minicomputer systems. Both prototypes were constructed
    as wire-wrapped frames.




    --

    Martin | martin at
    Gregorie | gregorie dot org

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From TimS@3:770/3 to tnp@invalid.invalid on Fri Oct 27 09:29:02 2023
    On 27 Oct 2023 at 05:32:34 BST, "The Natural Philosopher"
    <tnp@invalid.invalid> wrote:

    But, remember "wire wrap" ?

    I was trained, as an apprentice, on military spec hardware construction:
    Wire wrap was one system that was used extensively in early digital
    hardware backplane prototyping.

    Once the design was stabilised they tended to replace all that with a 2
    layer PCB.

    Still have the tools for doing that. The more-extensive wrapping CAN
    be as reliable as soldered connections.

    Yes. especially in high vibration environments where components are not mechanically connected except by their leads or pins.

    I was told, back in the day, that it's more reliable. Many mainframes had wire-wrapped backplanes. I recall seeing engineers come in during maintenance time to replace wires due to upgrades or bug fixes; their being wire-wrapped made it rather easier.

    --
    Tim

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From David Higton@3:770/3 to Martin Gregorie on Fri Oct 27 15:36:58 2023
    In message <uhe2l8$1mm2q$1@dont-email.me>
    Martin Gregorie <martin@mydomain.invalid> wrote:

    On Thu, 26 Oct 2023 16:16:33 +0100, David Higton wrote:

    In message <65398740@news.ausics.net>
    not@telling.you.invalid (Computer Nerd Kev) wrote:

    The Natural Philosopher <tnp@invalid.invalid> wrote:

    Pi connectors. I bought kit to make up Pi female connectors to go
    from the pins on one board to pins on another. I wasted a whole day
    and a half and was completely unable to make a satisfactory lead up
    , either by crimping or by soldering. Gave up and bought jumper
    leads.

    If they were ribbon cable connectors then I can understand that - you really want the proper tool for pushing those together when they're
    that wide.

    I've found it not too difficult to do ribbon insulation displacement
    cables if you have a vice, but the caveats I can think of are:

    Use a VERY sharp pair of scissors to cut the cable. If it isn't cut cleanly, there may be little bits of copper core hanging out, just
    waiting to short to a neighbour. (Been there.)

    Connectors need to be crunched on at very close to 90 degrees to the
    cable. It's surprising how little margin for error there is, or else
    there may be shorts between adjacent cores.

    The vice's jaws need to cover the whole connector and to close parallel.
    If the jaw surfaces are serrated, some dense cardboard may help, but you will need them held in place simply to avoid having to keep so many bits all in the right place as you tighten the vice.

    Curiosity: would it help to use one of the spray-on contact cements or one
    of the cyanoacrylates (e.g. Zap) to hold things in place while they're in
    the vice?

    Dunno. I was thinking more of double sided tape.

    David

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Another Dave@3:770/3 to The Natural Philosopher on Fri Oct 27 17:01:54 2023
    On 25/10/2023 3:51 pm, The Natural Philosopher wrote:

    I will shortly be getting my own printer. I would appreciate any tips
    like these you can give me!

    I use octoprint (octoprint.org) on a Pi 2 I had lying about.

    To design items I use tinkercad.com. If you already know about CAD use
    whatever you're familiar with. I was 77 when I started and this was
    adequate.

    Pass the designed object (.stl file) to prusa slicer:

    https://github.com/prusa3d/PrusaSlicer/releases

    and then to octopi.

    Adhesion of the object to the print bed is THE major problem, too lttle adhesion and the print WILL at some stage float off the bed, too much
    adhesion and it will be impossible to remove from the surface. I use a
    magnetic layer, you remove it from the bed and bend it to remove the print.

    The instructions will say use a sheet of paper to set the nozzle height.
    Use a 0.30mm feeler gauge.

    If you intend to print gears or pulleys come back to me.

    If you're prone to frustration take a tranquiliser before you start.

    Good luck

    Another Dave

    --
    Change nospam to techie

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From The Natural Philosopher@3:770/3 to Another Dave on Fri Oct 27 19:06:42 2023
    On 27/10/2023 17:01, Another Dave wrote:
    On 25/10/2023 3:51 pm, The Natural Philosopher wrote:

    I will shortly be getting my own printer. I would appreciate any tips
    like these you can give me!

    I use octoprint (octoprint.org) on a Pi 2 I had lying about.

    To design items I use tinkercad.com. If you already know about CAD use whatever you're familiar with. I was 77 when I started and this was
    adequate.

    Pass the designed object (.stl file) to prusa slicer:

    https://github.com/prusa3d/PrusaSlicer/releases

    and then to octopi.

    Adhesion of the object to the print bed is THE major problem, too lttle adhesion and the print WILL at some stage float off the bed, too much adhesion and it will be impossible to remove from the surface. I use a magnetic layer, you remove it from the bed and bend it to remove the print.

    The instructions will say use a sheet of paper to set the nozzle height.
    Use a 0.30mm feeler gauge.

    If you intend to print gears or pulleys come back to me.

    If you're prone to frustration take a tranquiliser before you start.

    Good luck

    Another Dave

    All advice noted!

    --
    “It is hard to imagine a more stupid decision or more dangerous way of
    making decisions than by putting those decisions in the hands of people
    who pay no price for being wrong.”

    Thomas Sowell

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to The Natural Philosopher on Fri Oct 27 21:57:06 2023
    On 10/27/23 12:32 AM, The Natural Philosopher wrote:
    On 27/10/2023 04:44, 56d.1152 wrote:


       Solder IS the best.

    No, actually it often isn't...

    The problem with solder on *stranded* wire is that it wicks up the wire
    and reduces it's flexibility by binding the strands together: That
    creates a stress concentration where it stops and a likely cause of
    fracture failure under vibration. That is why crimps are the choice for cheaper connections in high stress environments (e.g. automotive) or if
    you must solder you absolutely need to mechanically support the wire
    upstream of the solder joint.

    I've had very bad luck with 'crimps' - esp when they are
    exposed to the weather, but even in higher-vibration
    environments. Wasted lots of time trying to track down
    problems related to crimps. The worst case was a relative
    who was into boats ... the connections WOULD show voltage
    or a ground with today's high-impedence meters. Ultimately
    had to build a LOW-impedence meter, with a big resistor
    inside, that pulled a LOAD of about two amps. THEN you
    could find the faults.

    Solder correctly and put heat-shrink over it. You'll
    be fine.


       But, remember "wire wrap" ?

    I was trained, as an apprentice, on military spec hardware construction:
    Wire wrap was one system that was used extensively in early digital
    hardware backplane prototyping.

    Yep, it was VERY popular for prototypes back in
    the day - and for good reason.


    Once the design was stabilised they tended to replace all that with a 2
    layer PCB.

    I've also just seen plenty of cases where the successful
    proto just got all its wraps soldered for posterity. Had
    my hands on a whole LSI-11 kit computer done like that.

       Still have the tools
       for doing that. The more-extensive wrapping CAN
       be as reliable as soldered connections.

    Yes. especially in high vibration environments where components are not mechanically connected except by their leads or pins.

    Vibration is the 'hidden threat' ... not always
    an obvious issue to eye or mind, but it'll GET you.
    Even pure copper wire will eventually develop
    stress cracks due to vibration. I did a lot of
    automotive/truck projects back in the day and we
    learned to strap-down *everything* to keep it from
    buzzing around.

    Most 'computer' people these days are really most
    entirely 'software'. They can't see mechanical
    or noise or RF issues that crop up in the Real
    World. The 'techs' who DO may not be as good
    with software. It's a problem. There's a window
    of maybe 1960 to 1985 where 'programming' and
    'design/construction' kinda went together. Not
    nearly as many now who cover one end to the other
    except maybe the "BattleBots" crowd.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Martin Gregorie@3:770/3 to All on Sat Oct 28 10:42:30 2023
    On Fri, 27 Oct 2023 21:57:07 -0400, 56d.1152 wrote:

    Most 'computer' people these days are really most entirely
    'software'. They can't see mechanical or noise or RF issues that crop
    up in the Real World. The 'techs' who DO may not be as good with
    software. It's a problem. There's a window of maybe 1960 to 1985
    where 'programming' and 'design/construction' kinda went together.
    Not nearly as many now who cover one end to the other except maybe
    the "BattleBots" crowd.

    No argument here (my background is systems design and programing on
    Mainframes, minicomputers and fault-tolerant systems.

    The other gang that understand interactions between electronics and
    vibration are model flyers: radio control models, drones and, the most vibration of all, IC powered free flight competition models: a current F1C class model has a 2.5cc engine putting out around 1.3 HP at 30,000 rpm and carrying electronic flight timers and a GPS-based radio beacon as a
    retrieval aid. Free flight models regularly travel 2km or so in the course
    of a 3 minute flight, especially if the contest is being run on
    Sculthorpe, an ex-RAF/USAF base in Norfolk,UK. This area is known for its
    fresh sea breezes.


    --

    Martin | martin at
    Gregorie | gregorie dot org

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From David Higton@3:770/3 to 56d.1152@ztq9.net on Sat Oct 28 20:32:48 2023
    In message <BuudnQ1ot_sf96H4nZ2dnZfqn_GdnZ2d@earthlink.com>
    "56d.1152" <56d.1152@ztq9.net> wrote:

    On 10/27/23 12:32 AM, The Natural Philosopher wrote:
    On 27/10/2023 04:44, 56d.1152 wrote:


       Solder IS the best.

    No, actually it often isn't...

    The problem with solder on *stranded* wire is that it wicks up the wire
    and reduces it's flexibility by binding the strands together: That
    creates a stress concentration where it stops and a likely cause of fracture failure under vibration. That is why crimps are the choice for cheaper connections in high stress environments (e.g. automotive) or if
    you must solder you absolutely need to mechanically support the wire upstream of the solder joint.

    I've had very bad luck with 'crimps' - esp when they are
    exposed to the weather, but even in higher-vibration
    environments. Wasted lots of time trying to track down
    problems related to crimps.

    I still remember a very interesting one in the 1980s with a joystick
    for a Sinclair Spectrum. The box proudly proclaimed the use of gold
    plating in the internal connections. They used fully (i.e. not
    selectively) gold plated crimps onto stranded tinned copper wire.
    One of the joints, although completely mechanically sound, was not
    conducting electricity. Yes, that really was the interface between
    the crimp and the wire.

    I ran solder in and it was fine.

    There is a known, but unfortunately not widely known, problem in the
    interface between gold and tin.

    The lesson boils down to:

    1) Never mix gold plated and tin plated connectors.

    2) If you're using gold plated connectors, use selectively gold plated
    crimps in them.

    David

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to Martin Gregorie on Mon Oct 30 00:51:14 2023
    On 10/28/23 6:42 AM, Martin Gregorie wrote:
    On Fri, 27 Oct 2023 21:57:07 -0400, 56d.1152 wrote:

    Most 'computer' people these days are really most entirely
    'software'. They can't see mechanical or noise or RF issues that crop
    up in the Real World. The 'techs' who DO may not be as good with
    software. It's a problem. There's a window of maybe 1960 to 1985
    where 'programming' and 'design/construction' kinda went together.
    Not nearly as many now who cover one end to the other except maybe
    the "BattleBots" crowd.

    No argument here (my background is systems design and programing on Mainframes, minicomputers and fault-tolerant systems.

    I did a lot of microcontroller-based projects back in
    the day. This was "end to end" design. You had to not
    only program, but implement the hardware - with
    attention to all the real-world annoyances, and there
    might be software tweaks required to HELP with those
    Real World issues. Actual electronics/soldering was
    involved.

    Since that time, software/hardware have kind of become
    distinct specialties ... but that middle ... where it
    all has to come together ... seems to have been lost.

    The other gang that understand interactions between electronics and
    vibration are model flyers: radio control models, drones and, the most vibration of all, IC powered free flight competition models: a current F1C class model has a 2.5cc engine putting out around 1.3 HP at 30,000 rpm and carrying electronic flight timers and a GPS-based radio beacon as a
    retrieval aid. Free flight models regularly travel 2km or so in the course
    of a 3 minute flight, especially if the contest is being run on
    Sculthorpe, an ex-RAF/USAF base in Norfolk,UK. This area is known for its fresh sea breezes.

    An actual hydrocarbon-fuel engine WILL create a lot
    of very buzzy vibration - which WILL take its toll
    on every connection. It's amazing how quickly some
    connections will fail - chips can even work their
    way out of conventional sockets.

    And then, if an ignition system is involved,
    transient electrical noise ! :-)

    (which is usually high-voltage/micro-amperage, which
    makes opto's esp useful)

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to Ahem A Rivet's Shot on Mon Oct 30 01:39:52 2023
    On 10/27/23 2:42 AM, Ahem A Rivet's Shot wrote:
    On Thu, 26 Oct 2023 23:44:48 -0400
    "56d.1152" <56d.1152@ztq9.net> wrote:

    But, remember "wire wrap" ? Still have the tools
    for doing that. The more-extensive wrapping CAN
    be as reliable as soldered connections.

    A good wire wrap connection (done with a power tool - we always
    used Gardner-Denver) is gas tight and extremely reliable as well as being faster and easier to modify than soldering.

    Yep. Some of those "old world" solutions can
    STILL be good solutions !

    And, when everything's 100%, you can always
    solder the wire-wrap.

    As said somewhere, one employer had a *kit* LSI-11
    box. It had BASIC, but everything of use was writ
    in FORTRAN. They'd worked out all the wiring issues
    and soldered the wire-wrap. From there on it was
    just a software box. It's STILL around in a
    storage spot somewhere, I'll bet on it. And it'll
    still WORK - (assuming the 8" floppies can still
    be read). Oh, dual 8" Shugart floppy unit ...
    weighed about 75 POUNDS, BIG transformer :-)

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to David Higton on Mon Oct 30 01:23:06 2023
    On 10/28/23 3:32 PM, David Higton wrote:
    In message <BuudnQ1ot_sf96H4nZ2dnZfqn_GdnZ2d@earthlink.com>
    "56d.1152" <56d.1152@ztq9.net> wrote:

    On 10/27/23 12:32 AM, The Natural Philosopher wrote:
    On 27/10/2023 04:44, 56d.1152 wrote:


       Solder IS the best.

    No, actually it often isn't...

    The problem with solder on *stranded* wire is that it wicks up the wire
    and reduces it's flexibility by binding the strands together: That
    creates a stress concentration where it stops and a likely cause of
    fracture failure under vibration. That is why crimps are the choice for
    cheaper connections in high stress environments (e.g. automotive) or if
    you must solder you absolutely need to mechanically support the wire
    upstream of the solder joint.

    I've had very bad luck with 'crimps' - esp when they are
    exposed to the weather, but even in higher-vibration
    environments. Wasted lots of time trying to track down
    problems related to crimps.

    I still remember a very interesting one in the 1980s with a joystick
    for a Sinclair Spectrum.


    DID have one ! :-)


    The box proudly proclaimed the use of gold
    plating in the internal connections. They used fully (i.e. not
    selectively) gold plated crimps onto stranded tinned copper wire.
    One of the joints, although completely mechanically sound, was not
    conducting electricity. Yes, that really was the interface between
    the crimp and the wire.

    I ran solder in and it was fine.


    Yep ... crimps SOUND great - until you hit
    the Real World.

    Crimps are a cheap quick fix.


    There is a known, but unfortunately not widely known, problem in the interface between gold and tin.

    The lesson boils down to:

    1) Never mix gold plated and tin plated connectors.

    2) If you're using gold plated connectors, use selectively gold plated
    crimps in them.

    Material compatibility IS always important.

    Gold MOSTLY works with everything but, by
    experience, not ALWAYS

    Soldering IS a bigger pain in the ass, no question.
    But it's MUCH more sure. It'll last 20-50 YEARS under
    really crappy conditions.

    My old old HOUSE is mostly SOLDERED electrical
    connections - not even sure how they DID that
    conveniently back in the day. However I'm NOT
    worried about any of those connections - mostly
    in the attic/ceiling - overheating like modern
    screw-terminal connections that oxidize or
    where the copper compresses and becomes loose.

    A couple extra hours in 1950 meant 100+
    years of reliability. When (likely soon) the
    house is destructed the ELECTRIC will still
    be 100%

    Anyway ... BAD experiences with 'crimps' under
    a variety of conditions. Soldered is much more
    sure. Solder, add heat-shrink to spread out
    the mechanical stress and weatherproof ...
    electronic nirvana ! :-)

    Sorry, I'm very old-school in this respect.
    I really design for 100 years safe service.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Martin Gregorie@3:770/3 to All on Mon Oct 30 13:05:26 2023
    On Mon, 30 Oct 2023 00:51:14 -0400, 56d.1152 wrote:

    The other gang that understand interactions between electronics and
    vibration are model flyers: radio control models, drones and, the most
    vibration of all, IC powered free flight competition models: a current
    F1C class model has a 2.5cc engine putting out around 1.3 HP at 30,000
    rpm and carrying electronic flight timers and a GPS-based radio beacon
    as a retrieval aid. Free flight models regularly travel 2km or so in
    the course of a 3 minute flight, especially if the contest is being run
    on Sculthorpe, an ex-RAF/USAF base in Norfolk,UK. This area is known
    for its fresh sea breezes.

    An actual hydrocarbon-fuel engine WILL create a lot of very buzzy
    vibration - which WILL take its toll on every connection. It's
    amazing how quickly some connections will fail - chips can even work
    their way out of conventional sockets.

    And then, if an ignition system is involved, transient electrical
    noise ! :-)

    High performance model engines have no ignition: they're either diesels (COMPRESSION IGNITION burning an oil/kerosene/ether mix) or gloplug
    ignition (battery heated plug with platinum coil to start, when running combustion keeps the plug hot and burning methanol/oil mix, sometimes with added nitromethane for more power.

    You quickly learn how to protect in-model electronics and associated
    wiring from the engine's vibration - if you don't you'll have lots of more
    or less spectacular crashes. An F1C class model can climb vertically to
    around 120m with a 3 second engine run.


    --

    Martin | martin at
    Gregorie | gregorie dot org

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to Martin Gregorie on Sun Nov 5 00:33:12 2023
    On 10/30/23 9:05 AM, Martin Gregorie wrote:
    On Mon, 30 Oct 2023 00:51:14 -0400, 56d.1152 wrote:

    The other gang that understand interactions between electronics and
    vibration are model flyers: radio control models, drones and, the most
    vibration of all, IC powered free flight competition models: a current
    F1C class model has a 2.5cc engine putting out around 1.3 HP at 30,000
    rpm and carrying electronic flight timers and a GPS-based radio beacon
    as a retrieval aid. Free flight models regularly travel 2km or so in
    the course of a 3 minute flight, especially if the contest is being run
    on Sculthorpe, an ex-RAF/USAF base in Norfolk,UK. This area is known
    for its fresh sea breezes.

    An actual hydrocarbon-fuel engine WILL create a lot of very buzzy
    vibration - which WILL take its toll on every connection. It's
    amazing how quickly some connections will fail - chips can even work
    their way out of conventional sockets.

    And then, if an ignition system is involved, transient electrical
    noise ! :-)

    High performance model engines have no ignition: they're either diesels (COMPRESSION IGNITION burning an oil/kerosene/ether mix) or gloplug
    ignition (battery heated plug with platinum coil to start, when running combustion keeps the plug hot and burning methanol/oil mix, sometimes with added nitromethane for more power.

    Yea - but those are MODELS. Compression/glow + fuel still
    means a lot of vibration. It's just not so important. Now
    in actual 18-wheelers, 737s, M1 tanks - it IS very important.

    You quickly learn how to protect in-model electronics and associated
    wiring from the engine's vibration - if you don't you'll have lots of more
    or less spectacular crashes. An F1C class model can climb vertically to around 120m with a 3 second engine run.

    Soldered is THE standard. Kinda mush the strands together and
    solder and make sure you have some heat-shrink about 2X the
    width of the exposed copper for stress relief. If the env is
    "buzzy" then you have to strap down the wiring too.

    I live in an old house. All the wiring was initially SOLDERED.
    Not sure HOW they used to do that efficiently. The connections
    where then either put in early twist cons and/or covered with
    the olde-tyme rubbery "friction tape" (which they STILL sell).
    The big wires are all industrial-qual early ROMEX with the
    fiberglas shield inside and mostly in metal conduit.

    It's a house with a lot of WOOD - no drywall crap. I am NOT
    worried about what's in the attic or walls because the cons
    are SOLDERED. They'll be good for 200 years. No oxidation,
    no loosening probs associated with 'compression' connectors.

    Anyway, as they say, you get what you pay for. Modern cons
    are "adequate" - but not for 25 or 50, much less 200, years.
    The wiring WILL oxidize or loosen - then get HOT. For
    real-world equipment soldered IS still best as well for
    reliability.

    Recently replaced a plug socket. Put SOLDER over the
    ends of the wires and used the SCREW DOWN terminals
    rather than the "push in" ones. A little extra effort
    PAYS OFF longer term.

    Just sayin' ... there's "to code" and then there's GOOD.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From mm0fmf@3:770/3 to All on Sun Nov 5 10:09:30 2023
    On 05/11/2023 04:33, 56d.1152 wrote:
    Recently replaced a plug socket. Put SOLDER over the
      ends of the wires and used the SCREW DOWN terminals
      rather than the "push in" ones. A little extra effort
      PAYS OFF longer term.

    But best practice documents say using soldered/tinned wire in screw
    connectors is bad due to differing expansion rates of the brass fitting
    and solder. With enough sufficient thermal cycles the connection loosens causing bad connections.

    Obviously the experts should have consulted with you.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to All on Sun Nov 5 22:44:50 2023
    On 11/5/23 5:09 AM, mm0fmf wrote:
    On 05/11/2023 04:33, 56d.1152 wrote:
    Recently replaced a plug socket. Put SOLDER over the
       ends of the wires and used the SCREW DOWN terminals
       rather than the "push in" ones. A little extra effort
       PAYS OFF longer term.

    But best practice documents say using soldered/tinned wire in screw connectors is bad due to differing expansion rates of the brass fitting
    and solder. With enough sufficient thermal cycles the connection loosens causing bad connections.

    Don't expect any problems with a microns-thin
    solder coating. The main goal there is to shield
    the copper from oxidation/corrosion. My environ
    IS a bit corrosive. House fires due to older
    connections is NOT unheard of. Breakers protect
    from overload, but cannot detect a degrading
    connection.

    Now if you just glob LOTS of solder on there ...

    There ARE some kinds of 'goop' that claim to
    protect the copper, but I'll stick with what
    I know is good. Some of that goop is hydrocarbon
    based - basically gasoline if it gets hot. The
    silicone-based ... I'd have to see some good
    studies as to HOW much that restricts current flow
    and whether it can degrade PVC insulation long-term.

    Obviously the experts should have consulted with you.

    They should have. I have lots of real-world experience
    with marine applications.

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From David Higton@3:770/3 to scott@alfter.diespammersdie.us on Mon Nov 6 21:56:30 2023
    In message <UQ92N.38940$sqIa.2684@fx07.iad>
    scott@alfter.diespammersdie.us wrote:

    56d.1152 <56d.1152@ztq9.net> wrote:
    Recently replaced a plug socket. Put SOLDER over the
    ends of the wires and used the SCREW DOWN terminals
    rather than the "push in" ones. A little extra effort
    PAYS OFF longer term.

    Sounds like a recipe for a fire. I've seen screw terminals on 3D-printer motherboards that were scorched after tinned wires had been put in them,
    and they only carried maybe 100-200W at 12-24V.

    Never ever use screw terminals onto stranded wire that you've tinned.
    Solder suffers from creep, so the pressure slowly relaxes until the
    joint is loose. As stated above, a recipe for a fire.

    David

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From 56d.1152@3:770/3 to David Higton on Tue Nov 7 23:35:38 2023
    On 11/6/23 4:56 PM, David Higton wrote:
    In message <UQ92N.38940$sqIa.2684@fx07.iad>
    scott@alfter.diespammersdie.us wrote:

    56d.1152 <56d.1152@ztq9.net> wrote:
    Recently replaced a plug socket. Put SOLDER over the
    ends of the wires and used the SCREW DOWN terminals
    rather than the "push in" ones. A little extra effort
    PAYS OFF longer term.

    Sounds like a recipe for a fire. I've seen screw terminals on 3D-printer
    motherboards that were scorched after tinned wires had been put in them,
    and they only carried maybe 100-200W at 12-24V.

    Never ever use screw terminals onto stranded wire that you've tinned.
    Solder suffers from creep, so the pressure slowly relaxes until the
    joint is loose. As stated above, a recipe for a fire.

    True !

    But be aware that electrical-quality copper ALSO 'creeps'.

    Had a mystery power prob at my home a few years ago - it
    turned out to be that the old screw-down connections in
    the MAIN BREAKER had 'crept', loosened, over the years.
    The electrician I'd hired put on rubber gloves and
    re-tightened everything HOT. Good for another 20 years
    I suppose.

    For awhile I was responsible for control panels in
    20-50 horsepower industrial devices. Once most were
    10+ years old I decided to test the old power
    connections. They were ALL getting loose - the copper
    had 'compressed'.

    DO sometimes run into home sockets and such where the
    screw-down has a sort of bronze/brass PLATE under it.
    Put the wire under the plate. It creates spring-tension
    that'll greatly reduce the 'compression' issue.

    Note that 'tinning' the stranded wires DOES seem to
    slow down the 'creep', if you 'tin' correctly
    (meaning JUST enough solder to solidify/bond, not
    big blobs). IMHO, house-wiring should be done almost
    entirely with SOLID wire though. Alas CORROSION like
    you'd get on sea-coasts can STILL ruin things. For
    that a VERY thin coat of solder serves well - wipe
    on a little flux and LIGHTLY tin and wipe. Just a
    few microns.

    The most horrible mistake was back when they first
    introduced aluminum wiring for homes. The metal in
    the sockets/switches/etc were NOT super-compatible.
    Esp in 'corrosive' environs the connections would
    fail quite early - and start fires.

    Remember - breakers protect against SHORTS, but
    can NOT detect bad connections - which get hotter
    and hotter as the resistance increases.

    You get what you pay for. The 'discount' stuff from
    Home Despot or whatever will be 'functional', but
    for HOW LONG ? Alas, if/when it fails, it really
    might burn your house down.

    Anyway, enough of Practical Power Electrics in
    a Pi group :-)

    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)