How do library shapes affect pick and place data?

Warren Brayshaw schrieb:

Tilmann Reh wrote:

>> Warren Brayshaw schrieb:

>>

>>> Have a look at the forum post at eagle.userchat.eng

>>> Electronic Component Zero Orientation For CAD Library Construction

>>> 17 Nov 2012

>>>

>>> The information in the links illustrates the need to select one of

>>> the standards for all your librariey parts.

>>> I like IEC 61188-7 Level B

>>

>> The orientation declared in the document referenced in that thread is,

>> however, pretty uncommon.

>>

>> They define pin 1 of SOT/SOIC/QFP packages to be in the upper left

>> corner, while it is common practice to have pin 1 in the lower left

>> corner. A standard that is not commonly accepted is about useless.

 

Thanks for alerting others that there is information referrred to in that

post that is not a good choice.

When I referenced the post I felt readers would spot Olivers responce

contained a better later referrence.

Note I said "I like IEC 61188-7 Level B"

That revision made pin1 lower left

See Olivers final reference

http://blogs.mentor.com/tom-hausherr/blog/tag/ipc-standards/

Then scroll down to the referenced PDF - and that contains the level A

definitions again. What a mess.

It's really idiotic to have two contradictory "standards" under the same

document number. If someone refers to "IEC 61188", he'll probably not be

aware of the completely different "levels" and omit that very important

information.

So unfortunately, the only safe way is to stick to whatever definition

and then communicate it with your assembler.

(For us it's easy, since we produce 100% in-house. )

Tilmann

Tilmann Reh wrote:

Warren Brayshaw schrieb:

 

>> Have a look at the forum post at eagle.userchat.eng

>> Electronic Component Zero Orientation For CAD Library Construction

>> 17 Nov 2012

>>

>> The information in the links illustrates the need to select one of

>> the standards for all your librariey parts.

>> I like IEC 61188-7 Level B

 

The orientation declared in the document referenced in that thread is,

however, pretty uncommon.

 

They define pin 1 of SOT/SOIC/QFP packages to be in the upper left

corner, while it is common practice to have pin 1 in the lower left

corner. A standard that is not commonly accepted is about useless.

Thanks for alerting others that there is information referrred to in that

post that is not a good choice.

When I referenced the post I felt readers would spot Olivers responce

contained a better later referrence.

Note I said "I like IEC 61188-7 Level B"

That revision made pin1 lower left

See Olivers final reference

http://blogs.mentor.com/tom-hausherr/blog/tag/ipc-standards/

Warren

-

Viewed / responded via the newsgroup at

news.cadsoft.de

Warren Brayshaw schrieb:

Have a look at the forum post at eagle.userchat.eng

Electronic Component Zero Orientation For CAD Library Construction

17 Nov 2012

 

The information in the links illustrates the need to select one of the

standards for all your librariey parts.

I like IEC 61188-7 Level B

The orientation declared in the document referenced in that thread is,

however, pretty uncommon.

They define pin 1 of SOT/SOIC/QFP packages to be in the upper left

corner, while it is common practice to have pin 1 in the lower left

corner. A standard that is not commonly accepted is about useless.

By selecting a standard you can convey which one you have used to the

assembler. Doing this should ease and speed their work, all of which you pay

for.

I think the simple rule is: If you provide manufacturing data to an

assembler, you always need to provide information about your reference

orientations, too.

Tilmann

Tilmann Reh wrote:

Andrew Crosland schrieb:

 

>> For example is it important that the component footprint is exactly

>> centred on the origin when creating it in the library editor?

>>

>> Is this the same origin point referred to in the description of

>> mount-marks.ulp?

>>

>> How does the rotation value work? It guess it must depend on the

>> orientation of the library shape and how it is laid out on the board,

>> but how do I relate that to the way the component is oriented in the

>> feed to the pick and place machine, e.g. tape or waffle tray. Do I

>> have to be careful to draw the library shape with the correct initial

>> orientation. This may require different shapes depending if the

>> component is loaded on tape or in tubes or waffle trays.

 

It's pretty simple.

 

All angular values reference the orientation in the library.

 

At the p&p machine, you need to enter the part orientation in the

feeder anyway. So you simply also enter the rotation relative to your

library here.

 

Exactly the same way, different kinds or orientations of feeders are

also dealt with.

 

So basically, you just need to know how the parts are oriented in

your library. For this, a set of common rules is very useful.

Tilmann

Carrying on from Tilmann, regarding rotation.

Have a look at the forum post at eagle.userchat.eng

Electronic Component Zero Orientation For CAD Library Construction

17 Nov 2012

The information in the links illustrates the need to select one of the

standards for all your librariey parts.

I like IEC 61188-7 Level B

By selecting a standard you can convey which one you have used to the

assembler. Doing this should ease and speed their work, all of which you pay

for.

This is the best you can do to assist the assembler. They still have to

convert your zero rotation to the appropriate rotation required by their

machine and the way the parts are orientated on the tape etc.

The second best thing to assist the assembler is to include a part number on

each line of the file you provide. They should only have to do the work once

for each part number.

Now for the origin point placement.

For symmetrical parts like most devices the origin point should be in the

centre. The assembler is expecting that. the point is usually smooth for the

vacuum pickup point and rotation works for them easily.

For assymetric devices its up to you and is often on one central axis at a

nominal middle. This eases placement on  the board as you design it. The

assembler will pickup that part based on their needs. Vacuum picking needs a

smooth area and they may need to centre their 'pick' over the centre of

gravity (CofG) of  heavier parts.  In the set up for your board they will

apply offsets from your origin to that needed for their pick/place to

achieve the correct part placement. Datasheets may provide information on

CofG and make it the origin, basing dimensions on it.

If any body knows the source of a good explanation of this part of assemble

set-up could they please start a new thread in the UserChat forum.

HTH

Warren

--

Viewed / responded via the newsgroup at

news.cadsoft.de

Andrew Crosland schrieb:

For example is it important that the component footprint is exactly

centred on the origin when creating it in the library editor?

 

Is this the same origin point referred to in the description of

mount-marks.ulp?

 

How does the rotation value work? It guess it must depend on the

orientation of the library shape and how it is laid out on the board,

but how do I relate that to the way the component is oriented in the

feed to the pick and place machine, e.g. tape or waffle tray. Do I

have to be careful to draw the library shape with the correct initial

orientation. This may require different shapes depending if the

component is loaded on tape or in tubes or waffle trays.

It's pretty simple.

All angular values reference the orientation in the library.

At the p&p machine, you need to enter the part orientation in the feeder

anyway. So you simply also enter the rotation relative to your library here.

Exactly the same way, different kinds or orientations of feeders are

also dealt with.

So basically, you just need to know how the parts are oriented in your

library. For this, a set of common rules is very useful.

Tilmann