Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > sci.electronics.design > #460119

Re: Microphone risk factors (OT -- long)

Show all headers | View raw

On 5/13/2017 12:26 PM, Terry Newton wrote:
> On Sat, 13 May 2017 11:57:17 -0700, Don Y wrote:
>
>>>> For example, when tablets are manufactured (by compressing "powdered
>>>> granulation"), a common problem is air becoming entrapped in the
>>>> "powder" as it is compressed.  When the pressure is released, the
>>>> compressed air causes the tablet to "explode" (the top pops off).
>
>>> Separate the sensor and electronics. Sensor can be a robust piece of
>>> metal foil etc.
>>
>> In the above example, a microphone is impractical -- due to other
>> constraints imposed by the application.  E.g., the tabletting area has
>> to be "cleanable" to ensure product from "tabletted compound X" doesn't
>> contaminate "tabletted compound Y" (you wouldn't want the BABY ASPIRIN
>> that you make in the afternoon to have traces of the VIAGRA that you
>> made in the morning!  :> )
> .....
>> You're surrounded by the noise of the mechanism (you're making 35-75
>> tablets concurrently) and trying to hear a little "pop" that is muffled
>> inside a steel box -- that is MOVING past your sensor.
>
> Isn't this a visual inspection issue? using a mic to hear a pop
> seems redundant when you can just look at it to see if it popped.

You're producing ~200 tablets a SECOND.  They are typically collected in
a *barrel* due to the quantity involved.  *One* of the discharge chutes
is visible in the left of this photo:
    <http://www.chrishillsethenterprises.com/images/P/400%20Manesty.jpg>
It's mate is hidden on the opposite side of the machine.  The large
black knobs control physical dimensions of the tablets plus where
(vertically) in the die it is formed.

There are camera inspection systems but they are expensive and have to
be tuned to the tablet geometry in question -- different tablets have different
shapes, some have holes in the center, etc.  And, the tablets are essentially
airborn when they exit the process -- you can't tell which side will be "up"
or "facing the camera(s)" (the bottom can pop off just like the top).

Lower right corner shows oblong tablets falling off a "deduster" into
a collection barrel; another deduster can be seen behind it on the far
side of the machine:
    <http://www.dmcindustries.com/ProductImages/112319/md_8ba5a4_1614b.jpg>

Additionally, you want to know about it in relatively real-time.  If you
have a production problem, you don't want to *keep* producing "defects".

A microphone is just a bad choice of sensor despite the apparent applicability
(imagine taking one 1/4inch "cell" from a piece of bubble wrap and driving a
diesel truck over it and trying to hear it pop -- given that it might not
even be present!)

Photo of a top view (the bottom view is the mirror image -- including the
large cylindrical LOWER punches that mirror the UPPER punches, shown here)
of a turret from a smaller machine to give you an idea of how the tablets
are formed (badly maintained, rusty, etc.):
   <https://www.fraingroup.com/wp-content/uploads/5D4776Y.jpg>

Note the oblong-shaped die exposed in the lower center.  If you could look
straight down into it, you would see the LOWER punch has been raised up
into it from below.  This gives you a cavity with an open top (before
the UPPER punch is driven into it).

The brass structure to the right (turret rotates clockwise so that represents
an earlier "place" in time for the die) is the feed frame.  Its full of
granulation ("pill powder") and falls into the open-topped die as they
pass beneath it.  The feed frame sits *on* the top surface of the turret/die
so it scrapes any excess off of each cavity as the cavity passes outside the
scope of the feed frame.  I.e., the height of the LOWER punch as it passes
under the feed frame determines the volume of the cavity -- the amount of
granulation that (theoretically) will be drawn into the cavity.

Once clear of the feedframe, the LOWER punch is pulled down to move the
granulation to a more opportune portion of the die (i.e., away from the top
edge).  Then, the UPPER punch is driven down into the cavity from above
while the LOWER punch holds its own vertical position.

Compression force is applied by the punches (upper and lower) being forced
between a pair of stationary rollers, one of which is visible in the extreme
upper left corner of the photo.  It's mate is not visible but acting on the
LOWER punch from below.

The distance between the upper and lower ROLLERS thus determines the
compressed thickness of the finished tablet; the volume of the cavity
as it passed through the feed frame determines the *weight* of the tablet;
the fixed dimensions of the die determine the outline of the tablet.
The material+dimensions+compressive force determine the hardness of the
tablet.  All of these are criteria that can be used to reject a
tablet as "defective".

Of course, punches and dies have tolerances on their physical dimensions
so you get a *range* of "nominal tablets" from the press;  trying to control
the process involves controlling a bunch of REALLY SIMILAR, but different,
processes running simulataneously yet sharing certain common aspects (e.g.,
the feedframe, rollers, etc.)

As the turret rotates, it may see 2 or 4 such "compressing events" on
each "station" (UPPER+LOWER+DIE) before the tablet is ejected by lifting
the UPPER punch off of the tablet WHILE raising the LOWER punch to force
the compressed tablet from the die.  Some machines have a two-stage compression
cycle -- a lighter force to "tamp" the granulation down followed by a
heavier compression force to *form* the tablet.

Some machines can make (or be reconfigured to make -- at a large fraction
of a megadollar you want to be able to apply the machine to a wide variety
of products!) "two layer" tablets by making the first layer on one
"side" of the machine and NOT ejecting it; rather, adding more (different)
granulation on top of it and forming the composite tablet on the second
"side" of the machine before ejecting a finished tablet.

[E.g., I think there was a popular antacid that was very obviously a
two-layered product -- with different colored layers to stress this.
More recently, two-layer tablets find use in osmotic pump delivery
systems for controlled rate of drug delivery:  the tablet is coated
with an insoluble film that acts as a semi-permeable membrane; the
membrane has a hole laser-drilled through it on the "active" (drug)
layer side of the tablet; the "far" layer/side of the tablet is
a hydrophilic substance that swells as moisture is absorbed through
the membrane thus *pushing* the active ingredient out through the
drilled orifice.  By tweeking the permeability of the coating layer
and swelling characteristics of the "pusher" layer, you can predict
the rate at which the active will be "dispensed"]

The "pop" occurs as the compressive pressure is released from the
tablet -- so, it occurs while the tablet is "buried" within the
upper+die+lower envelope.  *As* the tablet is batted off the turret,
you want to decide "Go-NoGo" and deflect defective tablets into a
scrap bin -- before the next tablet appears in 10ms and has to be
similarly dispatched.  Screw up and let a defective tablet fall
into the "good" bin and you have to inspect its entire contents
to find the offender(s).  This also can invalidate your process:
"How did a bad tablet NOT get detected and deflected?  What's
wrong with your *process* given that it was designed not to allow
this to happen?"

The "popped off" piece may be ejected with the tablet, or it may cling
to the face of the upper/lower punch and fall off at some later point.
Or, it may hang around and screw up the next tabletting formulation on
that station as it goes 'round again.

> But a laser-based sound sensor (if that's what you want) will be
> far more robust than anything with electronics where it needs
> to be cleaned. Just hose it off, no electronics to get wet.

I.e., use a technology that is easier to clean and more sensitive
in light of the dynamic range of the "signal"!  :>

Back to sci.electronics.design | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-10 21:27 -0700
  Re: Microphone risk factors Phil Allison <pallison49@gmail.com> - 2017-05-10 21:39 -0700
  Re: Microphone risk factors jrwalliker@gmail.com - 2017-05-11 07:01 -0700
    Re: Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-11 11:06 -0700
      Re: Microphone risk factors piglet <erichpwagner@hotmail.com> - 2017-05-12 09:02 +0100
      Re: Microphone risk factors "Michael Kellett" <nospam@invalid.com> - 2017-05-12 16:02 +0100
        Re: Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-12 09:37 -0700
          Re: Microphone risk factors "Michael Kellett" <nospam@invalid.com> - 2017-05-12 18:08 +0100
            Re: Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-12 11:02 -0700
              Re: Microphone risk factors Phil Allison <pallison49@gmail.com> - 2017-05-12 17:44 -0700
              Re: Microphone risk factors Terry Newton <nbill@bellsouth.net> - 2017-05-13 17:19 +0000
                Re: Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-13 11:57 -0700
                Re: Microphone risk factors Jan Panteltje <pNa0nStpealmtje@yahoo.com> - 2017-05-13 19:12 +0000
                Re: Microphone risk factors Jan Panteltje <pNa0nStpealmtje@yahoo.com> - 2017-05-14 08:40 +0000
                Re: Microphone risk factors Jan Panteltje <pNa0nStpealmtje@yahoo.com> - 2017-05-14 09:17 +0000
                Re: Microphone risk factors Jan Panteltje <pNa0nStpealmtje@yahoo.com> - 2017-05-14 10:49 +0000
                Re: Microphone risk factors Terry Newton <nbill@bellsouth.net> - 2017-05-13 19:26 +0000
                Re: Microphone risk factors (OT -- long) Don Y <blockedofcourse@foo.invalid> - 2017-05-13 13:37 -0700
            Re: Microphone risk factors "dcaster@krl.org" <dcaster@krl.org> - 2017-05-13 11:06 -0700
              Re: Microphone risk factors Don Y <blockedofcourse@foo.invalid> - 2017-05-13 12:11 -0700
              Re: Microphone risk factors krw@notreal.com - 2017-05-13 20:36 -0400

csiph-web