From Newsgroup: sci.electronics.design


I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On Wed, 07 Jan 2026 09:57:52 -0500, DJ Delorie <dj@delorie.com> wrote:

I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: >https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: >https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

Currents and voltages? kickback spikes? Magnetic dirt? ...

Joe
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Joe Gwinn <joegwinn@comcast.net> writes:

Currents and voltages? kickback spikes? Magnetic dirt? ...

Industrial lego - 24vdc, low power and sensing, twisted pair (cat5).
The machine mostly does wood, with occasional plastic or aluminum. No
steel. The machine itself is fairly precise and I have an idea for
holding the tool-plus-connector in the right spot for insertion, so the tolerance isn't terrible. It just isn't zero ;-)

The reason[*] for this project is I want a touch probe, but I don't want
to worry about batteries or forgetting to hook up the cable. There may
be other uses in the future but they'd be in the same category.

Here's the video that inspired me: https://youtu.be/YnWSN8lmbfw?si=XLYUFzsFlJ78PYGm&t=63

[*] other than "because I want to" of course ;-)
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/26 15:57, DJ Delorie wrote:

I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-loaded-connectors

https://www.hirose.com/en/product/series/HR34P#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Lasse Langwadt <llc@fonz.dk> writes:

https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-loaded-connectors

pass on pogo pins, and I don't see how those allow for misalignment

https://www.hirose.com/en/product/series/HR34P#

"Blind mating" was the term I needed. The hirose look overcomplicated
for what I want. I found some other blind mating connectors at digikey,
but they look like sawdust traps: https://www.digikey.com/en/products/detail/molex/0447690801/513218 https://www.digikey.com/en/products/detail/molex/0444320801/513232

The oscilloscope type at least had a wide open female layout I could
blast air through.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On Wed, 07 Jan 2026 15:54:47 -0500, DJ Delorie <dj@delorie.com> wrote:

Lasse Langwadt <llc@fonz.dk> writes:

https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-loaded-connectors

pass on pogo pins, and I don't see how those allow for misalignment

Typical pogo-pin interface is a pointed pin resting on a flat plate,
so misalignment is accommodated. All gold-plated machined brass.

Mil-Max stuff is usually very good. I'd dig deeper, maybe find a
sample to inspect.

https://www.hirose.com/en/product/series/HR34P#

"Blind mating" was the term I needed. The hirose look overcomplicated
for what I want.

Won't handle dirt and fragile too.

Joe

I found some other blind mating connectors at digikey,
but they look like sawdust traps: >https://www.digikey.com/en/products/detail/molex/0447690801/513218 >https://www.digikey.com/en/products/detail/molex/0444320801/513232

The oscilloscope type at least had a wide open female layout I could
blast air through.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/26 21:54, DJ Delorie wrote:

Lasse Langwadt <llc@fonz.dk> writes:

https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-loaded-connectors

pass on pogo pins, and I don't see how those allow for misalignment

https://www.hirose.com/en/product/series/HR34P#

"Blind mating" was the term I needed. The hirose look overcomplicated
for what I want. I found some other blind mating connectors at digikey,
but they look like sawdust traps: https://www.digikey.com/en/products/detail/molex/0447690801/513218 https://www.digikey.com/en/products/detail/molex/0444320801/513232

but you'd have to come up with some floating mount to account for misalignment, the hirose has it build in, specs say +/- 0.5mm

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 7:57 AM, DJ Delorie wrote:

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack:

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

A lifetime ago, we used a ZIF-ish connector pair to allow for a
"memory cartridge" to be inserted into an instrument "effortlessly"
to reduce the risk of requiring any sort of force to mate the
connectors DAMAGING the connectors and rendering the machine
inoperative.

Our implementation effectively ran the processor bus out to the
pins (buffered, of course) so it had loads of conductors (30?).

The (each!) pins, IIRC, were composed of groups of rigid strands
of gold plated conductors that "meshed" with an equivalent number
of strands in the opposing "pin(s)". Imagine two pieces of Litz
wire (for each conductor) "interpentrating" each other.

The connectors were costly but lasted forever -- mechanically and
electrically. (the device in question was several kilobucks in
1980 dollars)

My concern would be for the cleanliness of your environment;
I wouldn't know the types of particulate matter that could
become embedded (??) in those little meshes.

OTOH, our connector was up-facing so anything that was dropped
in its vacinity would find its way onto those "pins". It was
in a medical environment so liquid reagents were perpetually
present yet we didn't have problems, there.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 5:17 PM, Don Y wrote:

On 1/7/2026 7:57 AM, DJ Delorie wrote:

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack:

I've figured out mostly how to do it but I'm a bit stumped on the
connector.-a I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion.-a Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope:
https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

A lifetime ago, we used a ZIF-ish connector pair to allow for a
"memory cartridge" to be inserted into an instrument "effortlessly"
to reduce the risk of requiring any sort of force to mate the
connectors DAMAGING the connectors and rendering the machine
inoperative.

You might also consider mechanizing the mating of the connectors
as you likely have "programmed" the head to move to the "pick up
tool" location.

We had a device with a 600 (not a typo) coaxial pin connector that
used a motor driven screw mechanism to physically pull the two
connectors together (mere mortals wouldn't be able to do it,
let alone *align* the pins without damaging any of them as the
connector had a ~1 sq ft mating surface).

Note that the electrical connection might be able to just be electromagnetically "reinforced" (flat conductor faces)
given that the "tool head" has already been accommodated
in your design (?).

A lot, of course, depends on what is passing through those
contact pairs... current, voltage, "content". Can you reduce
the number of conductors (power, gnd, bidir data) my putting
an MCU on the toolhead?

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 5:17 PM, Don Y wrote:

OTOH, our connector was up-facing so anything that was dropped
in its vacinity would find its way onto those "pins".-a It was
in a medical environment so liquid reagents were perpetually
present yet we didn't have problems, there.

To be clear, in light of my followup on this, there was no
"mechanization" of the interface; you simply DROPPED the
cartridge onto the connector and the connection was made.
No "second step" to secure it as required in the case of ZIF
sockets for DIPS, flex cable, CPU sockets, etc.

As such, something else would have had to ensure the two devices
remained in contact (gravity, in our case; whatever mechanical
interlock holds your tool head in place in yours)
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Don Y <blockedofcourse@foo.invalid> writes:

You might also consider mechanizing the mating of the connectors
as you likely have "programmed" the head to move to the "pick up
tool" location.

My plan was to make a custom tool fork (it holds the tool holder) with a tapered slot in it for the "arm" of the connector. The arm would have a matching geometry, so when the machine moves Y- to insert the tool into
the fork, it also slots the arm into the tapered slots and mostly locks
it in place. At that point, the Z motion of the cnc is enough to
mate/unmate the connectors. Or destroy them.

The side of the connector that moves with the tool would have some
wiggle allowance which should be enough to allow the connectors to line
up, I just need the connectors to want to line up.

A lot, of course, depends on what is passing through those
contact pairs... current, voltage, "content". Can you reduce
the number of conductors (power, gnd, bidir data) my putting
an MCU on the toolhead?

I could reduce it to four, I suppose - power, ground, sense, and probe.
Nothing excessive. I chose eight because I put extra cat5's in the
cable run when I built it ;-)
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 5:42 PM, DJ Delorie wrote:

Don Y <blockedofcourse@foo.invalid> writes:

You might also consider mechanizing the mating of the connectors
as you likely have "programmed" the head to move to the "pick up
tool" location.

My plan was to make a custom tool fork (it holds the tool holder) with a tapered slot in it for the "arm" of the connector. The arm would have a matching geometry, so when the machine moves Y- to insert the tool into
the fork, it also slots the arm into the tapered slots and mostly locks
it in place. At that point, the Z motion of the cnc is enough to
mate/unmate the connectors. Or destroy them.

The side of the connector that moves with the tool would have some
wiggle allowance which should be enough to allow the connectors to line
up, I just need the connectors to want to line up.

To be ATTRACTED to each other! (see below)

A lot, of course, depends on what is passing through those
contact pairs... current, voltage, "content". Can you reduce
the number of conductors (power, gnd, bidir data) my putting
an MCU on the toolhead?

I could reduce it to four, I suppose - power, ground, sense, and probe. Nothing excessive. I chose eight because I put extra cat5's in the
cable run when I built it ;-)

What about repurposing a "magnetic PHONE connector" (no idea what they
are called -- USB connector that magnetically couples to its mate
when brought in close proximity.

[I looked into this when I wanted to make my "control panels" portable
(remove them from the wall and carry them around with you). At the time,
I was intending to run PoE to the panels so the connectors didn't fit
the bill. But, I have since changed the architecture so the smarts
are no longer in the panel -- it is just a "display" so 4 conductors
could work]

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Don Y <blockedofcourse@foo.invalid> writes:

What about repurposing a "magnetic PHONE connector" (no idea what they
are called -- USB connector that magnetically couples to its mate
when brought in close proximity.

Sawdust and vibration, that's why. I really wanted a wiping contact
along with the air jets to make sure it made a good reliable connection.
I've already broken enough things due to forgetting to connect probe
cabled, I don't need the machine doing the forgetting for me ;-)

The video I posted used pogo pins, which could be combined with magnets
and/or a printed funnel. The oscilloscope connector could hold itself
together with the force of its wiping contacts.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 5:47 PM, Don Y wrote:

On 1/7/2026 5:42 PM, DJ Delorie wrote:

To be ATTRACTED to each other! (see below)

A lot, of course, depends on what is passing through those
contact pairs... current, voltage, "content".-a Can you reduce
the number of conductors (power, gnd, bidir data) my putting
an MCU on the toolhead?

I could reduce it to four, I suppose - power, ground, sense, and probe.
Nothing excessive.-a I chose eight because I put extra cat5's in the
cable run when I built it ;-)

What about repurposing a "magnetic PHONE connector" (no idea what they
are called -- USB connector that magnetically couples to its mate
when brought in close proximity.

<https://www.amazon.com/Magnetic-Connector-JianHan-360%C2%B0Rotating-Adapter/dp/B085T7DLTF>

as one example.

(Be wary, some only support two conductors -- "charging cables")

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/7/2026 6:09 PM, DJ Delorie wrote:

Don Y <blockedofcourse@foo.invalid> writes:

What about repurposing a "magnetic PHONE connector" (no idea what they
are called -- USB connector that magnetically couples to its mate
when brought in close proximity.

Sawdust and vibration, that's why. I really wanted a wiping contact
along with the air jets to make sure it made a good reliable connection.
I've already broken enough things due to forgetting to connect probe
cabled, I don't need the machine doing the forgetting for me ;-)

But, why can't the machine VERIFY the connection (by "talking/listening"
to whatever is on the other side)?

This doesn't help with the vibration (but, the magnetic example I
posted may have enough attraction force to address that). Air jets
to clean it same as any other connector.

If you want a wiping action, then it will tend to require more
force (friction).

The video I posted used pogo pins, which could be combined with magnets and/or a printed funnel. The oscilloscope connector could hold itself together with the force of its wiping contacts.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

DJ Delorie <dj@delorie.com>wrote:

I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: >https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: >https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

Wireless power via induction or light (depends how much power you need)
and data optical or inductive or radio.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 7.1.2026 16.57, DJ Delorie wrote:

I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

The 'oscilloscope connector' seems to be an Amphenol Micro Ribbon.
There was a more robust similar family called Blue Ribbon.

I have no idea if they are still being available.
--

Tauno Voipio

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Don Y <blockedofcourse@foo.invalid> writes:

But, why can't the machine VERIFY the connection (by "talking/listening"
to whatever is on the other side)?

I'm assigning one pin to be "device connected" but the device isn't
smart. Basically, it's a switch, either on or off. The only way to
really *test* the device is to touch something, but if it doesn't
trigger, it's likely to just break off if the machine doesn't stop in
time.

If you want a wiping action, then it will tend to require more
force (friction).

I've got a big servo motor driving it down. I have enough force. I
probably have too much, hence wanting something that self-aligns the
last bit so the motor doesn't just destroy the connector.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Tauno Voipio <tauno.voipio@notused.fi.invalid> writes:

The 'oscilloscope connector' seems to be an Amphenol Micro Ribbon.
There was a more robust similar family called Blue Ribbon.

Bingo! I wonder if there's a modern equivalent...

(now I wonder if, with today's technology, I could just print one...
what are those metal bits made of again? :)

I have no idea if they are still being available.

I found a few on eBay that weren't stupidly expensive, and ordered them
before they vanished. Hopefully they work the way I want ;-)

Thanks!
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 2026-01-08 10:50, DJ Delorie wrote:

Tauno Voipio <tauno.voipio@notused.fi.invalid> writes:

The 'oscilloscope connector' seems to be an Amphenol Micro Ribbon.
There was a more robust similar family called Blue Ribbon.

Bingo! I wonder if there's a modern equivalent...

(now I wonder if, with today's technology, I could just print one...
what are those metal bits made of again? :)

I have no idea if they are still being available.

I found a few on eBay that weren't stupidly expensive, and ordered them before they vanished. Hopefully they work the way I want ;-)

Thanks!

Centronics connectors (used on old printers) work the same way, and may
be easier to find, especially if you have a bunch of old kit lying
around. ;) (GPIB is another approach.)

I pal around with folks who put optical instruments on train cars. The
only way to keep the window clean is to delete it and blow compressed
air continuously out of the hole. Some nice solenoid-valve version of
that might be the ticket, for a one-off.

Cheers

Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On Wed, 07 Jan 2026 15:54:47 -0500, DJ Delorie <dj@delorie.com> wrote:

Lasse Langwadt <llc@fonz.dk> writes:

https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-loaded-connectors

pass on pogo pins, and I don't see how those allow for misalignment

https://www.hirose.com/en/product/series/HR34P#

"Blind mating" was the term I needed. The hirose look overcomplicated
for what I want. I found some other blind mating connectors at digikey,
but they look like sawdust traps: >https://www.digikey.com/en/products/detail/molex/0447690801/513218 >https://www.digikey.com/en/products/detail/molex/0444320801/513232

The oscilloscope type at least had a wide open female layout I could
blast air through.

The microribbon conectors were usually accompanied by large guide
pins, even if the surrounding hardware was rail-mounted.

The connector should not be required to provide mounting support
for the 'parked' hardware.

You might look at battery pack connectors, but anything small
will collect dust. The simplest are spring-mounted pins aimed
at flat conductive targets.

Optical or magnetic coupling are probably the most R&D-worthy
modes of communication (and power) transfer between tools.

RL
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

legg <legg@nospam.magma.ca> wrote:

On Wed, 07 Jan 2026 15:54:47 -0500, DJ Delorie <dj@delorie.com> wrote:

Lasse Langwadt <llc@fonz.dk> writes: > >https://www.digikey.com/en/product-highlight/m/millmax/maxnetic-spring-lo >aded-connectors

pass on pogo pins, and I don't see how those allow for misalignment

https://www.hirose.com/en/product/series/HR34P#

"Blind mating" was the term I needed. The hirose look overcomplicated
for what I want. I found some other blind mating connectors at digikey, >but they look like sawdust traps: >https://www.digikey.com/en/products/detail/molex/0447690801/513218 >https://www.digikey.com/en/products/detail/molex/0444320801/513232

The oscilloscope type at least had a wide open female layout I could
blast air through.

The microribbon conectors were usually accompanied by large guide
pins, even if the surrounding hardware was rail-mounted.

The connector should not be required to provide mounting support
for the 'parked' hardware.

You might look at battery pack connectors, but anything small
will collect dust. The simplest are spring-mounted pins aimed
at flat conductive targets.

Look at the connectors on the sliding door of a van. They are exposed
to the weather and all kinds of abuse and they are probably available
quite cheaply as replacement spares.
--
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/8/2026 8:27 AM, DJ Delorie wrote:

Don Y <blockedofcourse@foo.invalid> writes:

But, why can't the machine VERIFY the connection (by "talking/listening"
to whatever is on the other side)?

I'm assigning one pin to be "device connected" but the device isn't
smart. Basically, it's a switch, either on or off. The only way to
really *test* the device is to touch something, but if it doesn't
trigger, it's likely to just break off if the machine doesn't stop in
time.

If you want a wiping action, then it will tend to require more
force (friction).

I've got a big servo motor driving it down. I have enough force. I
probably have too much, hence wanting something that self-aligns the
last bit so the motor doesn't just destroy the connector.

Ugh! Brain Fart! I've been reading "CNC" but *thinking* "3D printer",
hence looking for smaller, "daintier" connectors. I was wondering why "sawdust" was an issue! <:-(

When I was a kid, I would get my "electrical" parts from scrapped
"pin setters" (think: bowling). These typically had several 220V
fractional HP motors to drive the various mechanisms (bowling pins
are surprisingly heavy! moving ten of them at a time isn't something
you do casually).

Many of the subassemblies were on sleds (1/4 thick steel plate)
that would slide into the "chassis". As these are too heavy to
lift and carefully align, they would be coarsely aligned and then
"driven home" with a fair bit of blind force, hoping the connectors
would mate and not break.

The connectors floated with a pair of tapered steel pins to "home"
them relative to each other. The pins were flat copper strips
better than .125" wide (there's 14AWG wire carrying a couple of amps
attached to each so far from dainty -- but, no need to be as a pinsetter
is probably 2 dozen cubic feet in volume!). They could tolerate a great
deal of mechanical abuse as, invariably, these "sleds" wouldn't quite
line up as they were driven home and the connector bodies had to absorb
the abrupt impacts (including being mechanically "over driven")

We used a similar connector in the weight control module on tablet
presses. But, I can't seem to find photos of either application
by way of illustration.

Way back then, I didn't document things as that required real film
and a trip to the drug store (and an expense!).

I'll try some image searches and see if I can find anything
similar. But, you likely don't need as much current carrying
capacity, I assume...?

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 1/8/2026 2:58 PM, Don Y wrote:

I'll try some image searches and see if I can find anything
similar.-a But, you likely don't need as much current carrying
capacity, I assume...?

You can probably also try searching for "blind mating" connectors.
E.g., the battery packs in small (< 1KVA) UPSs are "driven home"
with no real chance to ensure the connectors will line-up when
they come into contact with each other. The connector *shells*
are tapered to self-align.

But, again, you're looking at larger currents though the connectors
are pretty small, considering.

[Also a ready source of surplus parts as so many UPSs are just discarded, intact]
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Don Y <blockedofcourse@foo.invalid> writes:

But, you likely don't need as much current carrying
capacity, I assume...?

Nope, it's just a touch probe. Tensl of milliamps at best, I would
assume.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 2026-01-07, DJ Delorie <dj@delorie.com> wrote:

Joe Gwinn <joegwinn@comcast.net> writes:

Currents and voltages? kickback spikes? Magnetic dirt? ...

Industrial lego - 24vdc, low power and sensing, twisted pair (cat5).
The machine mostly does wood, with occasional plastic or aluminum. No
steel. The machine itself is fairly precise and I have an idea for
holding the tool-plus-connector in the right spot for insertion, so the tolerance isn't terrible. It just isn't zero ;-)

The reason[*] for this project is I want a touch probe, but I don't want
to worry about batteries or forgetting to hook up the cable. There may
be other uses in the future but they'd be in the same category.

Here's the video that inspired me: https://youtu.be/YnWSN8lmbfw?si=XLYUFzsFlJ78PYGm&t=63

[*] other than "because I want to" of course ;-)

make it wireless and power it mechanically
--
Jasen.
Efc|Efca -i-+-#-#-# -u-|-C-#-u-+-u
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

Jasen Betts <usenet@revmaps.no-ip.org> writes:

make it wireless and power it mechanically

Wireless probes are common (and expensive), and if I was running the
machine daily I'd do that, but with my luck the battery would always be
dead when I needed it...

Also, I do not want to be in the room with anything I've designed that's powered off a 24,000RPM 3HP spindle :-)

(joking aside, it needs to be mounted in the spindle to be accurate, and
thus the spindle can't be turning when it's in use)
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On 2026-01-10, DJ Delorie <dj@delorie.com> wrote:

Jasen Betts <usenet@revmaps.no-ip.org> writes:

make it wireless and power it mechanically

Wireless probes are common (and expensive), and if I was running the
machine daily I'd do that, but with my luck the battery would always be
dead when I needed it...

Put a wireless charger in the tool rack?

Also, I do not want to be in the room with anything I've designed that's powered off a 24,000RPM 3HP spindle :-)

(joking aside, it needs to be mounted in the spindle to be accurate, and
thus the spindle can't be turning when it's in use)

Ok the spidle is out. but is there a dust vacuum you can use as a
power source? (I'm guessing you're not using coolant when machining wood)

Wireless capable microcontroller modules are a few bucks. eg from NRF
or ESP.
--
Jasen.
Efc|Efca -i-+-#-#-# -u-|-C-#-u-+-u
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: sci.electronics.design

On Wed, 07 Jan 2026 09:57:52 -0500, DJ Delorie <dj@delorie.com> wrote:

I had a hard time searching for this because I can't think of how to
describe it that doesn't mean "machines making connectors" or "build
your own robot".

I'm thinking of a new feature for my cnc machine that lets it
automatically make electrical connections to a tool the spindle loads.

It's an automatic tool changer with a tool rack: >https://www.delorie.com/photos/cnc2/PXL_20240324_210753133.html

I've figured out mostly how to do it but I'm a bit stumped on the
connector. I've seen this done with pogo pins but I'd like something
more reliable, even though I plan on using air jets to clean things
before insertion. Ideally, such a connector would be self-centering,
wiping, self-gripping (vs held in place by the spindle), and support 8
or so conductors.

I was thinking of something like this, from my childhood oscilloscope: >https://www.delorie.com/photos/cnc2/PXL_20260107_144338434.html

Ideas?

Better to keep the intelligence in the rack, rather than the tool head
itself.

A mechanical profile on the head can press buttons/switches in the
rack for identity. Collar ring diameter, width, position, quantity
etc.

RL
--- Synchronet 3.21a-Linux NewsLink 1.2