Prev: Solution MANUAL
Next: Please tell me what is the cheapest ARM with my request ip(100K annual)
From: Enrico on 9 Oct 2009 09:14
Hi,

I'm routing a board that will be tested by an Automatic Test Equipment
line.

I purchased and read the "SMTA/TMAG Testability Guidelines TP-101D"
from smta.org:
http://www.smta.org/store/book_detail.cfm?book_id=176

That document states that I have to make room on the TOP LAYER for 10
plastic PUSH FINGERS (radius 1.6 inches). The purpose of the PUSH
FINGERS is to press the board down to the bed of nail.

That seems a lot of room. Does anybody know how many PUSH FINGERS are
typically needed?

Thanks,
Enrico
From: Stef on 9 Oct 2009 11:13
In comp.arch.embedded,
Enrico <zigbee(a)libero.it> wrote:
> Hi,
>
> I'm routing a board that will be tested by an Automatic Test Equipment
> line.
>
> I purchased and read the "SMTA/TMAG Testability Guidelines TP-101D"
> from smta.org:
> http://www.smta.org/store/book_detail.cfm?book_id=176
>
> That document states that I have to make room on the TOP LAYER for 10
> plastic PUSH FINGERS (radius 1.6 inches). The purpose of the PUSH
> FINGERS is to press the board down to the bed of nail.

Isn't that the radius of the pushing surface of the 'finger', so it's not
a flat end, but a slight curve? the diameter of the 'finger' should be a
lot less. 3.2" diameter sounds more like an elephants foot that a finger.

> That seems a lot of room. Does anybody know how many PUSH FINGERS are
> typically needed?

The document says 10 always? I would say it depends on the number of nails,
the size of the PCB and so on. The function is to avoid excessive bending.



--
Stef (remove caps, dashes and .invalid from e-mail address to reply by mail)

"The urge to destroy is also a creative urge."
-- Bakunin
[ed. note - I would say: The urge to destroy may sometimes be a creative urge.]
From: larwe on 9 Oct 2009 12:05
On Oct 9, 9:14 am, Enrico <zig...(a)libero.it> wrote:

> I'm routing a board that will be tested by an Automatic Test Equipment
> line.

ATE typically tests the board before depanelization. Some if not all
of those fingers will press down on the panel in the waste area around
the edge. Some will press at the point where four boards meet,
typically covering a mounting hole at each corner, which is already
partially waste space.
From: Enrico on 9 Oct 2009 12:07
Hi,

> Isn't that the radius of the pushing surface of the 'finger', so it's not
> a flat end, but a slight curve? the diameter of the 'finger' should be a
> lot less. 3.2" diameter sounds more like an elephants foot that a finger.

> > That seems a lot of room. Does anybody know how many PUSH FINGERS are
> > typically needed?

> The document says 10 always? I would say it depends on the number of nails,
> the size of the PCB and so on. The function is to avoid excessive bending.

The GUIDELINE P-23 (excerpt): says:

To minimize board flex and provide even pressure against test probes,
push fingers are required on the top plate of the fixture.
These fingers will contact the surface of the PCB and ensure the PCB
will remaing flat during testing.
Maintain the following keep out and spacing criteria:

- A 0.160in (4.1 mm) diameter keep out area free of components should
be made available in a grid-like pattern every 1.50 in to 2.00 in (3.8
mm to 5.1 mm),
across the entire surface of the board. A rule of thumb is to provide
one pusher for every 10 probes.

- Additional sites may be required in areas of high probe density such
as around ball grid arrays (BGAs) or near connectors.
In certain cases, it is acceptable to push directly on the BGA or
connector.

-The height of Push Fingers should be within 0.15


If I understand correctly, the keep out area of the push finger should
be 4.1 mm diameter.

I don't quite understand "The height of Push Fingers should be within
0.15"... What does that mean?

thanks,
Enrico
From: Tilmann Reh on 9 Oct 2009 12:18
Enrico schrieb:

> Maintain the following keep out and spacing criteria:
> - A 0.160in (4.1 mm) diameter keep out area free of components should

....so this is not *1.6in radius*, but *0.16in diameter*.
That's correctly 4.1mm, which seems pretty reasonable.

> be made available in a grid-like pattern every 1.50 in to 2.00 in (3.8
> mm to 5.1 mm),

Error on unit conversion: 1.5in to 2.0in is 3.8 to 5.1 *cm*, not mm.
This again appears reasonable.

> across the entire surface of the board. A rule of thumb is to provide
> one pusher for every 10 probes.

So it's not always 10 pushers, but one pusher for every 10 contact
needles - to handle the forces of their springs.

> - Additional sites may be required in areas of high probe density such
> as around ball grid arrays (BGAs) or near connectors.
> In certain cases, it is acceptable to push directly on the BGA or
> connector.

Logical, isn't it? If you have an area with very many contact needles,
you will apply much force locally, bending the board - unless you also
provide more pushers at the component side.

> -The height of Push Fingers should be within 0.15

This sounds like the height tolerance (between the many fingers).

The person who gave you that specification should be able to explain you
what they really want. Why don't you simply ask?

Tilmann
 |  Next  |  Last
Pages: 1 2
Prev: Solution MANUAL
Next: Please tell me what is the cheapest ARM with my request ip(100K annual)