Author Topic: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network  (Read 15758 times)

0 Members and 1 Guest are viewing this topic.

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
(Inspired by coppercone2's thread, on various 3-D arrays, pls see '...Space-Time Simulation Network, (eevblog Technical Chat section).

   Some 'niche' applications benefit when alternate solutions are applied, as, for example: a data transfer network having sensors, and some digital 'logic', but also immersed in harsh conditions...like a volcano interior...
   At any rate, a fairly simple process, designing interconnecting path 'switches' allows for a series string network of (separate) stations, each individually accessable. The concept has each station in one of two modes: An active or controlled (rotary) output, or, a mode 'transparent' where some mechanical signal set is merely let 'through' to drive a same-format set (of rotary signal shafts).
   So, one of those 'rotary signals' is either used as 'data', of sorts, or can be used to do 'work' as CW and CCW rotations (CCW = counter clockwise).
   In this (current) plan, for the network construction project, there are 3 separate parallel signals, called simply 'A', 'B', and 'C'.
The plan theory of operation, resembles 'Leap Frog' actions where, for example, a first signal, 'SEL', or 'SELECT' enables a next use of the 'A' column, then a second signal, 'A' enables column 'B', and that X shaped 'leap frog' crosses and re-crosses, between the two columns, progressing (upwards direction on paper diagram). Meanwhile, signal column 'C' outputs get switched, along with 'B' column. This is why the passively switched signals in 'C' are termed as 'CLIENT' function signals. A Client signal, or rotary 'shaft' output would be doing any actual work, such as winding up (or down, release) of a string winding bobbin.
  PICT illustrates, the approximate type of 'Worm Gear' type useful as mechanical 'outputs', with position 'holding' power when any (rotary) forces have stopped.
   Plans include a series connected string of, perhaps 8 such stations, featuring 8 separate, individually accessable shaft outputs, each considered a digital 'zero or one's resolved number.

Pict showing small electric DC PM motor, driving a nylon worm gear. (Example from toy having reciprocal CCW-CW movements).
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Here is a second example PICT, included is the small PC board and custom micro-processor under a 'blob'.
  With larger mechanical networks, there might be several sub-system controller IC's. This picture also conveys approximate view and size, of ONE mechanical station, of 8, in one version. Then, of course, there can be 8 rows, for a 2-D network (8 by 8,)
and a possible 3-D configuration having (8 by 8 by 8), if the 2-D arrays are repeated 8 times...
  Now, so that is 512 individually accessable, (or addressable), but motor total count being 64...BUT, each string gets 4 motors, not one.
So, that's 256, actual, motors, when multiplied out.

You can appreciate, how such networks get impractical, past a point. A (5 by 5 by 5) network has much less shear bulk, but even a 5 X 5 X 5 network, gets big enough to demand special approaches. The concept here, is to use many, many plastic -printed parts, gears and rubber wheels, while reducing motor count requirements.


 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
So I'm treading on 'failure ground' a bit: but a lot accomplished, and better defined the problem, when 3-D ARRAYS component counts skyrocket...
   The best, is ability to skip some number of motors, then creating systems, mainly looks like strings of variable length, (station count) and requiring 4 motors, usually small DC PM toy size motors.
  Check out the waveform diagram, a processor output
causes rotary 'transfer' from (actual) motor, which can be either direction, CW or CCW.
   A 3.3 V system or 5 V system could do the switch move actions.
   Interesting, the 'A' signal shown, acts ahead, causing the stage '2' ahead to position that switch stage, actually. The set-up means that the 'A' signal travels into the following path switch, and emerges as a 'local' signal. That's what acts, then, to Select, the stage 2 ahead of present.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Please see picture, features one 'stage' as a 3 pole path switch, for rotation signalling.
By putting a brief CW or clockwise rotation of 200 milliSec. the path is changed, into 'local'.
That will mean, using a term 'JOB' for example in a stage 5, then an output can be labeled 'JOB5'.
That output, on column 'C', is grouped with col B as both are Col B type signals.
By omitting that motor, shown on left side, the rotary input that positions the SWITCH itself can be just another rotary input, from previous stage, usually.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Here is a fun application, those turn over tags, are similar to the old memory error latching flags.
Here, that's a 7 by 7 array, where you control each flag, along a horiz line row.
So, you have 4 motors  per row, that in itself would need 28 motors !
Shown is a crude attempt to draw a '2' on that 7 x 7 dot matrix.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   Presented as a series of 4 steps, the switch in this picture is shown 'SELECTED', then having 3 'local' access to outputs.
   First step, the 'C' colum Local Output is activated: THE JOB (rotary) output can be used - stopped - reversed - ....whatever, and for however long needed.
   Secondly then, job done, the 'A' colum is used. OR, 'A' is separate, allowing simultaneous activity, in this case.
Note that control action constantly does this alternating left-right-left ordering of signal paths. In that context, both 'B', and 'C' columns are 'right side'.

   This second step, 'A' rotation signal positions the next switch. It is called 'SEL' due to action, CW rotates
to SELECT the following stage. A '4th' BUS signal, this 'SEL' does NOT pass through, stage to stage but rather the signal is built, at each stage. The original signal is transferred by way of the BUS.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   Notice in this picture, that the next stage, Stage 6, is selected (into Local mode) and ready to go. The 'B' column signal is highlighted in green, showing a kind of 'un-select'; the 'A' layer being, also, split-ganged and can be reset, separate from 'B' or 'C' layer segments.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   Here is a view of a conceptual path switch: It has three separate levels on that toggle set.
Left side, for input is the dark blue wheel, up top.
Next, you can see two (yellow highlighted) wheels, as the other two inputs. So... all three levels gang together, at least when pushed clockwise, as a little tab on the 'A' wheel is what pushes the other two wheels, into 'Selected' position (CW rotation).
   Coming the other way, the 'A' layer switch gets pushed back (CCW) independently. (Just a simple tab that only pushes when moving 'A' toggle CW.)
Not shown is the toggle position setting, a small gear with post and toggle lever...simple really, nylon type small gears etc.
These multiple parallel path switches have potential, for chained contact, giving multiples, but not shown here. This 'toggle' platform is moved by signal from the 'A' column, usually. See other previous diagrams, 'A' column Local output...
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   Here is a picture, showing that '4th' step in the process, where this is the clean-up, or finish, by clearing the local mode. A semi-trick, a (rotary) tap is taken, off the stage 'CHAIN' output - this one is not used, until after stage has finished.
So, an extra gear take-off receives the 'A' signal rotation, that is always CW (clockwise), normally.
You just have invert that to CCW for pushing switch 5 closed. That is, the segments or layers for switching 'B' column, and 'C' column.
    Doing the stage switching this way, makes for nice speed and keeps the size vs time needed in a linear relation. Something like a 32 stage line-up is a goal, while still using 4 motors, at the line-up start.
   The 'A' column is responsible for starting, or SELECTING next stage. 'B' column is the STOP signal, causing transparent chain mode. 'C' column is yellow, as can be either state, '1' or '0'.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   The timing can be a little confusing, due to being pipelined signals. Interesting: the stage 5, as an example, will first SELECT stage 6, but no effect yet.
Next, both 'B' and 'C' can be active, because they are separate. Of course, ultimately those 2 rotary pulses came from the 'B' and the 'C' PM electric motor pulses, perhaps approx. 200 milli-seconds duration.
   But the waveforms show a quite simple and speedy access through the network, and LINEAR !!!
Looking like twice as many little stages would involve twice as much time...to access everything...
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
   Some practical observations:
   Accumulated friction probably will limit the number of series connected units: A string with 30 stages would be efficient, using 4 motors to turn the 30 (virtual rotating motor) addressable outputs.
   Of course, attempting to gain such efficiency does not work, for the other 2 axis, lacking yet another scheme, things (component count) blow up fast, for component counts.  The assumption being, that somewhat, a high component count is tolerable, when using re-cycled plastic 'line source' material, doing 3-D printing. It is the actual DC PM motors that have 'exotic' construction materials, ceramic magnets and other special metals used to make typical toy sized motors.
A 'few' actual motors is OK, but a motor count approaching 48, that's also going to involve driver electronics, say: 8 motor driver channels per driver PC board, that would require 6 PC boards, with processor control, and medium power output transistors (typical 800 mA motor current, each).
   The picture shows some options, for when outputs needed are moderate, such as a five or six output string.  For a 7 output string, only 2 stages are actually needed. PLUS, not discussed in depth, the first stage has its own dedicated motor, for switching all three ('A', 'B' and 'C' columns), and no 'gang-split' so all three rotary signals are simply switched together, as there is no conflict, of having a signal turn off its own path switch.
   After a system has a few stages, it starts to be operated more uniform, but the last stage has to be handled a bit different.
You can surmise: of the six signals being chained, down the serial lines (lines meaning connected rotary paths- analogous to electric wire lines),
Of the six BUS lines, the 'A' column, local, is for enable SELECT.
   The 'B' column, local, is for CANCEL local mode on 'A'.
   The 'C' column, local, is you active stage output, (or 'Client').
Then, also, are the 3 chain outputs, for maintaining the three transmission paths. In other words, the chain signals simply allow transfer stage to stage, (along non-selected stations, transparent or not affecting any components, in current stages as (rotary) signal progresses.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #11 on: November 02, 2021, 12:09:29 am »
   Things are starting to look good, at least the logic for switching down the line of stations, on the mechanical signal network (digital write-only BUS).
   As picture shows, the timing of 200 mSec. motor pulses (SEL1, 'A', 'B', and 'C' connection columns), is simple. Each movement impulse can be longer, and can even switch polarity, CCW rotation followed by 1 second CW rotation, for example.
   Highlighted is the Stage 4 (rotary) output, simply shown in this case, as a 200 mSec. CW or Clockwise movement, into the user's load or work to do.
   It's a challenge, working out the pipelined series of signals, as there are 3 separate stages being accessed, as including the current stage being output.
   Each new stage adds NO additional addressing requirement, a very helpful situation.
The logic utilizes an early enable, to next stage, tapped by a simple take-off wheel, to clear all switches in current stage.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #12 on: November 06, 2021, 09:49:26 pm »
   Having fleshed out a scheme for getting mechanical 'data' in rotary form, CCW or CW representing data '0' or data '1', here is shown how to read, remotely, from the 'mechanical' network:

   See in the picture, a signal to read a switch position is shown, in normally closed position, very similar to a single pole (electric) relay.  The yellow colored piece will toggle, causing selection of either input wheel.  Shown pushed to the right, this moving toggle piece contacts the (blue colored) wheel for CCW data result: that is a binary '0' representation.
The idea is to interrogate the 'toggle' position and send that back towards the home base. That contains the processor, for motor control, and the 4 small motors in the base unit. So, accessing the last in a string of, say 30 stations, along the series line, and allowing a local switch there to influence the rotary direction, a base unit can recieve the information, of which position the mechanical toggle is resting.
In position shown, a '1' sent to interrogate, will return a '0', implying the position of the sensor switch is 'N.C.' or 'normal closed or un-active'.
   As an alternate, suppose that (yellow colored) toggle is contacting the other wheel set:
Then, the rotation direction of the signal returned will be CW (clockwise).
   Each of how ever many remote switches are used will share that same 'rotary' return signal path, the current active stage being determined by which stage gets that interrogating 'clock', as it were.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #13 on: November 06, 2021, 09:59:39 pm »
(Sorry, this post has the close-up diagram, while last post has the network view).
   In the network, shown is stage 31, as an example.
As seen, an 'interrogation' pulse, (simple rotation, for 200 mSec.), where that pulse gets conducted down the network string of stations and the sensor switch converts that, according to actual switch position.

   As a, perhaps, bizzare example, this could be a mechanical network monitoring a hatch or other deep interior feature, in a nuclear core reactor. OR: perhaps an explosive hazard area, in a factory, where no electric devices can be used. This way, some simple, but crucial data points can be monitored, in a very computer-like manner.
  Not described, are how a base unit would convert such rotary signal, CCW or CW, back to conventional digital (electrical!), but somewhat easy to do that.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #14 on: November 17, 2021, 06:33:11 am »
   It's a TON of work, laying out parts, and designing.
But a quick couple minutes to assemble. In principal, two BUS signals, A and B, act on each other.
The 'B' signal column is 'GO to Ready',
while 'A' signals: 'STOP Ready state'.
  A general traffic on the BUS is usually A, B, or C and picture shows how each shaft is 'logically' a chain-in chain-out. A fourth signal, shaft, handles local control (select).
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #15 on: November 17, 2021, 06:39:30 am »
   The diagram shows a portion, in this case the yellow shaded wheels are pulleys, for bands. In the center, the B pulley is 1 1/2 inch DIA. and with belt drive from 3/4 inch pulley.
  If you care for the detail, the 'A' output is what will drive that large pulley, for the B switch toggle.
 

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 9166
  • Country: us
  • $
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #16 on: November 18, 2021, 04:06:03 am »
only thing I would add for construction is if you built it from a two part shell, planing the interface between the two parts flat might offer more repeatable and stable mating.. I mean on the bottom where the two boards meet, if its not done already. I think when its cut smooth it will work better then a saw cut surface.. from experience when I was joining multiple boards together to make cheap table covers I was doing this, and the side walls too, to ensure everything is super square
« Last Edit: November 18, 2021, 04:08:18 am by coppercone2 »
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #17 on: November 24, 2021, 12:43:59 am »
Yes, Coppertone2:
   I've just gotten enough done on the enclosure, that's always easier doing parallel (plates), as you said...

   Maybe doesn't look like much, but the 'brick /lunchbox' proto for a mechanical-addressed station helps deal with the relatively simple interference issues.
For example, each 'B' column receives a 'SELECT', which places that station into 'READY' mode (or SELECT mode).
For that SELECT signal, coming as an input, it is best handled, in or near the other input layer components. In this case, the receiver pulley is placed near to other inputs.
   For another case, taking the column 'A' signal, from the 'output' plane, dictates placement, of a receiver pulley, near other outputs.

   Notice the basic BUS seen, consists of 3 mechanical rotary signals, passing into and through each station (box). The fourth BUS signal, is used locally, for setting each station 'RELAY' into Ready status.
   Not obvious in picture, but this 'footprint' maintains position, while BUS passes from module to module without offsets.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #18 on: November 24, 2021, 01:50:46 am »
   It's a wacky solution, looking for a wacky problem, but...
Consider the HIGH VOLTAGE CAGE structure, needed to keep 'pesky humans' from trouble (accidentally contacting the 150 kV Southern Pike Line.)
This gajget made from wood or plastic could do a few ,(rotary) mechanical actuations, free of metal conductors.
But, of course, this 4-shaft BUS system might be better replaced with... simply 4 shafts. And for larger network (strings), some solutions gain too much simple friction, too fast:. But, it's a fun concept, an infinite network, using a 'NEXT ITEM' type access.
   A simple 8 way binary system would have 4 shafts, 3 for the binary address bits, plus a shaft acting as 'clock'. The actual decode done, classic style, with 3 shafts, bits 0, 1, and 2.
   Looking back at diagram, the station box pulls apart, back face (inputs) and front face (outputs), that way any shaft couplings etc can come free, as one plate is loosened.
   Even more detail: inside is half-way wall, for terminating shafts. Input shafts have terminating, as a bearing support, at the half-wall. Output shafts similar,
where half-wall supports shaft ends, while other ends exit the station housing.
   Assembly / troubleshooting is aided, by having easy to dis-assemble modules. Plus, metal screws or other metal fasteners should be avoided, for best design.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #19 on: November 25, 2021, 11:18:47 pm »
...Trying for a phrase: Best could come up with, regarding 'bloated designs' (and playing off the MEMS acronym for 'Micro - Electro - Mechanical' devices.)

   How about, for this 'Mechanical Logic Contraption':
...'GROSSLY Oversize Electro - Mechanical', or
   'GEMs'.
  Seriously, I discovered my (wood working) skills need updates: Much of that 'RELAY' box, pictured recently, is accurate, ONLY, to +/-  1/8 inch ...wow!
But that's boxwood, so it's easy to sand the soft wood.

   Came across an apt quote, via old LIFE Magazine article, on DISNEYLAND:
   "Intricate, hokey, hugely expensive assemblage..."
   "...embodied the businesslike use of fantasy, the
    no-nonsense approach to nonsense."

   That's gonna do, for now.

ENJOY YOUR HOLIDAY !
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #20 on: November 27, 2021, 03:21:56 am »
   I wanted to take time to describe my own take, on developing prototypes.

   Believe it, or not, that's my BABY, there, figuritively.
(See picture).  I'm thinking: "Any customer, of my product, is going to notice,  any dirty fingerprints, on that unfinished wood box".
   So that's one concern.  Inventers are, sometimes, a strange breed...You aren't 'working' for a living, no cash usually coming in today.  That can look a fool's pursuit.
REWARDING, heh yeah...maybe tomorrow.
You can't be expecting quick rewards, or you lose sight.  But it can be a very unpleasant journey.
   The problem area, with prototypes, is you can be faced with a wide open, BLANK page, and better get busy. Many folks balk at this point (ha! Wimps).

   But in reality, not so bleak as there are some set patterns of action that some creative people utilize.
Plus, pay attention to an opposite effect. Say, you've got a new dress design. Using existing designs, think of the burden, of review all those (thousands) of previous dress designs. In that dynamic, you are often better off ignoring everything else, and just get (your) design down, in paper, first.  THEN, you enter the 'prior art' review phase. (NO LEGAL ADVICE, just opinion).

   For dealing with novel design, my approach often starts with a goal, first prototype model, often precedes most functional goals. It's a packaging and 'fit and use' ergonomic consideration.
   Of course, this describes prototyping mechanics and with 'educational' market aspects, so the package appearance takes precedence.
The first (couple of) models, I let things get slightly 'wild', meaning that some bold or unusual feature might be emphasized. For example, after considering one particular model, requiring 4 foot tall cabinet, I went bold: exaggerating that cabinet, to 7 feet height.
Now, that's bold, but some restraint needed too!
   I often don't strictly require the whole 100 % of functions to be working, until 3rd prototype model is begun.
   My main point here is that it doesn't add that much to sort-of 'decorate' the dang thing: In this case, I wanted that clear wooden box to get a nice varnish, or oil rub finish. Can't hurt, and makes your new 'baby' purty to look at...
   Meanwhile, maybe you can't visualize the whole of it, mechanically especially. You give it your best shot, and add some cosmetically pleasing appearance. It can be an attitude changing approach.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #21 on: November 27, 2021, 03:56:40 am »
   Referring to previous photo, for sealing the wood case I like a product, WATCO DANISH OIL, a rub on oil finish.  I mean, why not get your prototype cabinet looking good, at a stage where things are starting to mostly fit together and work.
   "DON'T BE A 'WOOD BUTCHER'  as my 9th grade Wood Shop teacher would say. (lol).
By the third prototype box build I generally expect most of the nagging problems to be solved. For mechanical items it often comes down to a good 3-D perception, whether it's visual, or simply precedes any viable looking model. As you may purcieve, I put maybe more emphasis on 'looks' but that's an eye towards customer engagement. That's actually a HUGE vital aspect, if the effort is to create trade-worthy market products.
   Sometimes, a big / major error occurs. That's OK, part of this business cycle.  For example: suppose a co-worker points out: "...your signal output labels are totally backwards, on the bottom half..."
Now, perhaps a decision point: Remember the prototype isn't simply a 'wooden box': that's your investment, there, represents all those hours, time spent searching hardware store shelves, etc etc.
You have a management question: One approach is to be consistent and plan for at least part of (that 3rd proto model) to be functional and ready for some quality worthy testing, big time.
   Otherwise, you have to consider 'abandoning' that 3rd prototype model, and endure another round of build. Yes, I guess it's at this point that things can get serious. Every inventor / innovator has this kind of story...well almost all of us...

    "I don't like inventing, but the inventing likes me.."

(Borrowed from song lyrics: "I don't like THE DRUGS but the drugs like me".

   Inventor reading this?  Even if technically off-base, the information here is meant to encourage creative exploration, in engineering, and present some methods in approaching novelty in product development.
Thanks.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #22 on: November 29, 2021, 12:59:51 am »
By way of explanation, the attached picture shows, partially, the organization of 'internal compartments.
To simplify, there are two main rectangular compartments or volumes. That is, the interior division, depicted by the half-wall in the picture, is, first, to make the 'A' and the 'C' individual switch areas.
Then, interior to those, further divided (not shown here) are two volumes or 'compartments' where function is split: one compartment handles inputs, to the 3PDT switch, while the other is for Output components. It's all fairly simple, but with some demanding attention, for avoiding basic interferences, 3-D wise.
   Because the switch module involves this 4 layer stack of compartments, the depth available for components is fairly tight, (using 4 inch depth, in present prototype.)

   Perhaps too much detail info here, but the 'C' portion of the assembly sits behind the other ('A'), and literally, the shafts associated with 'A' had to 'poke through' the volume assigned to the 'C' toggle portion. That's OK, as clearances are carefully kept, between the two extra shafts, and the toggle lever (of 'C' ).
   It's messy, but easy to assemble.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #23 on: November 30, 2021, 01:14:36 am »
   Here is photo depicting an antenna modifying 'tuning' or impedance switching, as MEMS application. (That's Micro-Electro-Mechanical device).
   For special switching, roof-top or other settings, including in presence of high voltages, the Mechanical Network, discussed here, could be utilized, at the four switch points, shown in the example.
   Of course, would be easier, to just run 4 shafts with no complicated 'network node boxes'.  A better example might be, if there are several such antennas, needing switches, say, spread out over a roof-top.  Then, a network accessed mechanically, by four 'encoded' shafts, could replace a system, say four antennas, each needing 4 switch boxes. So that reduces the 'control' shafts from 16 original.
 

Offline RJSVTopic starter

  • Super Contributor
  • ***
  • Posts: 2099
  • Country: us
Re: Sharing some project planning phase: A (digital) ELECTRO-MECHANICAL Network
« Reply #24 on: November 30, 2021, 01:24:45 am »
   Picture shows prototype enclosure, the 3PDT mechanical switch box, in comparison with what half-sized unit would look like. (That's a standard wall box, for indoor light switch).
   By designing / building at that scale, volume goes down, by a factor of 8.  Much of the switch box could be done different: after all, it's really just a small rotary transmission.  It's the LOGICAL use here, that is central to operating as a networked string of devices.
   Current design uses rubber wheels, for rotary contact, but a gear to gear arrangement probably has lower total friction.  Friction is a critical factor,  when attempting to operate a string of, say, 40 devices (3PDT Switches).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf