Maker Pro
Maker Pro

PADS, ASCII output question: Scale factor?

J

Joerg

Hello Folks,

We weren't able to sort this out. The layouter said this is the usual
file that SMT assemblers use to program their pick&place machines but
the assembler said they can't use it since the scale factor or data
format for the coordinates is unknown. PADS was set to the American
dimensioning of "mils" (1/1000th of an inch). Contacting Mentor resulted
in no help. Here is one example from the ASCII file:

R75.2 R66.2
-190690500 289179000 1 381000 1024 TEARDROP N 90 90
-190690500 287274000 31 381000 512 TEARDROP P 90 90

What does "-190690500 289179000" mean in terms of mils, or millimeter?
 
C

Christopher Ott

Joerg said:
Hello Folks,

We weren't able to sort this out. The layouter said this is the usual file
that SMT assemblers use to program their pick&place machines but the
assembler said they can't use it since the scale factor or data format for
the coordinates is unknown. PADS was set to the American dimensioning of
"mils" (1/1000th of an inch). Contacting Mentor resulted in no help. Here
is one example from the ASCII file:

R75.2 R66.2
-190690500 289179000 1 381000 1024 TEARDROP N 90 90
-190690500 287274000 31 381000 512 TEARDROP P 90 90

What does "-190690500 289179000" mean in terms of mils, or millimeter?

That actually looks like gerber info to me.

I just generated a pick and place file out of Protel99SE, here's what part
of it looks like:

Designator Footprint Mid X Mid Y Ref X Ref
Y Pad X Pad Y TB Rotation Comment
U33 SO-8 1790mil 310mil 1915mil
385mil 1915mil 385mil B 180.00 LM2672M-5.0
U34 SO-8 840mil 310mil 965mil
385mil 965mil 385mil B 180.00 LM2672M-5.0
R304 0805_-_100MIL 4275mil 4605mil 4320mil
4605mil 4320mil 4605mil B 180.00 6.04k
R302 0805_-_100MIL 3555mil 4335mil 3600mil
4335mil 3600mil 4335mil B 180.00 10.0k
R301 0805_-_100MIL 3555mil 4235mil 3510mil
4235mil 3510mil 4235mil B 360.00 10.0k
U54 SO-18_WIDE 3380mil 11330mil 3170mil
11130mil 3170mil 11130mil B 360.00 ULN2803AFW
U53 SO-18_WIDE 3380mil 9730mil 3170mil
9530mil 3170mil 9530mil B 360.00 ULN2803AFW

They don't really need the P&P file. It just makes it a little faster to
program the machine.

Chris
 
J

Joerg

Christopher said:
That actually looks like gerber info to me.

I just generated a pick and place file out of Protel99SE, here's what part
of it looks like:

Designator Footprint Mid X Mid Y Ref X Ref
Y Pad X Pad Y TB Rotation Comment
U33 SO-8 1790mil 310mil 1915mil
385mil 1915mil 385mil B 180.00 LM2672M-5.0
U34 SO-8 840mil 310mil 965mil
385mil 965mil 385mil B 180.00 LM2672M-5.0
R304 0805_-_100MIL 4275mil 4605mil 4320mil
4605mil 4320mil 4605mil B 180.00 6.04k
R302 0805_-_100MIL 3555mil 4335mil 3600mil
4335mil 3600mil 4335mil B 180.00 10.0k
R301 0805_-_100MIL 3555mil 4235mil 3510mil
4235mil 3510mil 4235mil B 360.00 10.0k
U54 SO-18_WIDE 3380mil 11330mil 3170mil
11130mil 3170mil 11130mil B 360.00 ULN2803AFW
U53 SO-18_WIDE 3380mil 9730mil 3170mil
9530mil 3170mil 9530mil B 360.00 ULN2803AFW

They don't really need the P&P file. It just makes it a little faster to
program the machine.

Thanks, Chris. Maybe the layouter generated the wrong file. Anyhow,
since he's traveling now they went ahead and started programming the
machine by hand. Over 500 parts...
 
H

Hal Murray

They don't really need the P&P file. It just makes it a little faster to
Thanks, Chris. Maybe the layouter generated the wrong file. Anyhow,
since he's traveling now they went ahead and started programming the
machine by hand. Over 500 parts...

What's the error rate on "hand" programming?

Reducing sources of errors seems like a good idea to me.
 
J

Joerg

Hal said:
What's the error rate on "hand" programming?

Reducing sources of errors seems like a good idea to me.

It's surprisingly low. In the late 80's it was all hand-programming and
the errors were usually zero. But we do not have a choice here because
the layouter can't be reached right now and the stuffed boards must be
at the client Monday morning.

It's 80% analog and I'll find a stuffing mistake pretty quickly. The
only drag would be that we'd then have to correct it about 40 times.
 
C

Christopher Ott

Hal Murray said:
What's the error rate on "hand" programming?

Reducing sources of errors seems like a good idea to me.


My machine uses mechanical centering (vs laser centering) and tends to be
off a little from the values the P&P file generates anyway. Manual
programming is actually done with the camera, so placement errors are not
really an issue. Most boards are easy to program manually, so not having the
P&P file is not really a big deal.

Chris
 
J

Joerg

Christopher said:
My machine uses mechanical centering (vs laser centering) and tends to be
off a little from the values the P&P file generates anyway. Manual
programming is actually done with the camera, so placement errors are not
really an issue. Most boards are easy to program manually, so not having the
P&P file is not really a big deal.

Which brings up a question: Do you do contract assembly? Looked on your
web site but it's under construction.
 
B

Brad Velander

Joerg,
The ASCII output is in PADs "basic units". If you have a PADs manual
there is usually a section in the rear of the manual (at least the older
manuals) that explains the basic ACSII format output and I believe it may
also explain the basic units measure. Today it is maybe in a PDF or text
file in one of your install directories or just left uninstalled on your
original install CD.

From memory (I haven't used PADs consistently for more than 8 years) the
basic units are something like 254 microinches (or was it 25.4 microinches).
It is the base unit of measure that the PADs database uses and allows them
to easily convert metric to inches and back again without round-off errors.
You should be able to look at a component or two, do some math and confirm
the basic unit measure knowing the components real location from the
database zero-zero point.

The line that you quoted would seem to be a trace between R75 Pin 2 and
R66 Pin 2, with teardropping. The first line will be the location of the one
end start point and the second line will be the location of the end point.
The "-190690500 289179000" are the base unit measures, the other data
relates to both trace width and the teardrop sizing and orientation. That is
my best recollection anyway.
 
J

Joerg

Brad said:
Joerg,
The ASCII output is in PADs "basic units". If you have a PADs manual
there is usually a section in the rear of the manual (at least the older
manuals) that explains the basic ACSII format output and I believe it may
also explain the basic units measure. Today it is maybe in a PDF or text
file in one of your install directories or just left uninstalled on your
original install CD.

From memory (I haven't used PADs consistently for more than 8 years) the
basic units are something like 254 microinches (or was it 25.4 microinches).
It is the base unit of measure that the PADs database uses and allows them
to easily convert metric to inches and back again without round-off errors.
You should be able to look at a component or two, do some math and confirm
the basic unit measure knowing the components real location from the
database zero-zero point.

The line that you quoted would seem to be a trace between R75 Pin 2 and
R66 Pin 2, with teardropping. The first line will be the location of the one
end start point and the second line will be the location of the end point.
The "-190690500 289179000" are the base unit measures, the other data
relates to both trace width and the teardrop sizing and orientation. That is
my best recollection anyway.


Thanks, Brad. Now I know where to look for the next round. Or actually
where our layouter could look because I don't have PADS.

I just wish Mentor had told our assembler that (our layouter was
traveling so he couldn't call them). But they more or less brushed him
off. Anyhow, now they have programmed the machine by hand and the boards
are coming off the reflow this afternoon.
 
C

Christopher Ott

Joerg said:
Which brings up a question: Do you do contract assembly? Looked on your
web site but it's under construction.


I do small volumes of control electronics, but so far it's just been my own
stuff. I can certainly look at your boards to see if I could run them
though. You can email me direct for more info.

Chris
 
J

Joerg

Christopher said:
I do small volumes of control electronics, but so far it's just been my own
stuff. I can certainly look at your boards to see if I could run them
though. You can email me direct for more info.

Chris

I'll do that when the next round comes. It's mostly fine-pitch SMT
stuff, up to 6U VME sizes (but not full length), some TSSOP and 0603.
Occasionally I have to drop down to 0402 sizes.
 
C

Christopher Ott

Joerg said:
I'll do that when the next round comes. It's mostly fine-pitch SMT stuff,
up to 6U VME sizes (but not full length), some TSSOP and 0603.
Occasionally I have to drop down to 0402 sizes.

My machine can handle boards up to 12"x10", and passives down to 0603.
0402's take a special feeders which I don't have. I've never tried TSSOP's
on it, but I can place 50mil pitch IC's no problem. No BGA's or anything
requiring x-ray inspection though.

Incidentally, if you're ever shopping for a P&P machine, laser centering is
a must. Mine uses the older mechanical jaws to square the parts on the head
before placement. It's effective, but it's also the primary limitation when
placing smaller than 25mil pitch parts.

Chris
 
J

Joerg

Christopher said:
The people I traditionally sell to aren't real interested in websites.
Somehow I manage to stay busy...

Same here. The only thing they sometimes want to look at is what
services can be provided.
 
J

Joerg

Christopher said:
My machine can handle boards up to 12"x10", and passives down to 0603.
0402's take a special feeders which I don't have. I've never tried TSSOP's
on it, but I can place 50mil pitch IC's no problem. No BGA's or anything
requiring x-ray inspection though.

Incidentally, if you're ever shopping for a P&P machine, laser centering is
a must. Mine uses the older mechanical jaws to square the parts on the head
before placement. It's effective, but it's also the primary limitation when
placing smaller than 25mil pitch parts.

My boards are rarely larger than that. But TSSOP would be 25mil pitch.
And I stay away from BGAs whenever possible because they are all but
non-inspectable.

But I won't shop for P&P machines, I'll leave that work to the experts
like you. Or, as the poker players say, you've got to know when to
fold'em ;-)
 
H

Hawker

Joerg,
There is a way to interpret that, but I use an unsupported script under
VB scripts in Layout to make my X-Y sheet. Some 3rd party programs that
some assemblers use can directly read a PADS ASCII file, that is why
they asked for that, but apparently yours cannot so use the built in
script under VB scripts.

BTW I thought you were on the PADS ListServer as well. Probably get
better answers there than here.

Hawker
 
Top