PCB Libraries Forum Homepage
Forum Home Forum Home > Libraries > Footprints / Land Patterns > Altium
  New Posts New Posts RSS Feed - Altium 0402 Resistor Footprint
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Altium 0402 Resistor Footprint

 Post Reply Post Reply Page  12>
Author
Message
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Topic: Altium 0402 Resistor Footprint
    Posted: 13 Jan 2020 at 10:25am
I just started using Altium having previously used PADS.

When using the IPC Compliant footprint wizard (medium density) using the dimension from Vishay CRCW data sheet I get the following footprint dimensions:

a: 0.55     b: 0.65    I: 0.2

The recommended dimensions from the Vishay datasheet are:

a: 0.40     b: 0.60    I: 0.5

The dimension from our PADS library are:

a: 0.57     b: 0.62    I: 0.39

It seems that the "I" dimension (spacing between pads) is  a lot smaller using the Altium wizard compared to the Vishay datasheet and current PADS library.

I asked our contract assembler and they said to use the Altium footprint as it will result in less tombstoning.

Would appreciate any opinions on this.

Thanks

Back to Top
Back to Top
Tom H View Drop Down
Admin Group
Admin Group
Avatar

Joined: 05 Jan 2012
Location: San Diego, CA
Status: Offline
Points: 5718
Post Options Post Options   Thanks (0) Thanks(0)   Quote Tom H Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 12:05pm
When I use the Free Library Expert Pro "Nominal Density Level" I get these dimensions: 


The PADS library parts seems to be the closest match. However, the PADS b: dimension is slightly large because there is no metal on the side of a resistor terminal and the package Width Tolerance is only +/- 0.05 mm. 

The part is 1.00 mm long with a 0.50 mm gap (between two 0.25 mm lead widths). 

The Altium I: dimension of 0.20 mm is too close and the heel joint would be 0.15 mm. Way out of line and I don't know why the assembly shop would recommend that one. 

The Library Expert Pro heel joint is 0.05 mm to cover the terminal tolerance. 

Download the Free (no license required) Library Expert Pro and use the Altium interface to produce an accurate footprint and High Quality 3D STEP model - 

Library Expert Pro actually shows you the package and terminal outlines on the pad so you can visually see where the solder is going to flow. And the body and terminal are translated to Altium on a mechanical layer. 

Stay connected - follow us! X - LinkedIn
Back to Top
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 3:43pm
Tom,

Thanks for the great advice.

It is kind of disappointing that our assembler does not know what a proper footprint is. 

The wizard in Altium has many options, maybe if I select the right ones I would get a footprint that makes more sense. 

I will give Library Expert Pro a try.

Back to Top
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 4:19pm
When I run it with the numbers from the datasheet I get slightly different results:

a: 0.58     b: 0.60    I: 0.35

Not sure why it is different from your results but not a long way off.
Also are the rounded corners (set to 25% of pad width) preferred to square corners?

Thanks
Back to Top
Tom H View Drop Down
Admin Group
Admin Group
Avatar

Joined: 05 Jan 2012
Location: San Diego, CA
Status: Offline
Points: 5718
Post Options Post Options   Thanks (0) Thanks(0)   Quote Tom H Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 4:40pm
Are these the same package dimensions and tolerances? 

The tolerances are important. 




I get these results. If your results are different, we need to talk.  



Stay connected - follow us! X - LinkedIn
Back to Top
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 5:22pm
Tom,

Using the numbers from the datasheet you posted the link to I do indeed get the same numbers as you.
Vishay has a lot of versions of the CRCW datasheet.
Here is the one I am using that matches the P/N I am using: http://www.vishay.com/docs/28773/crcwce3.pdf

The tolerances are different and it is dated 02-Nov-17.
The one you posted the link to is dated 18-Nov-10

Also we have Stackpole as a second source, fortunately their dimensions match Vishay except the tolerance on the termination width is 0.15mm instead of 0.1mm for Vishay.
Back to Top
Tom H View Drop Down
Admin Group
Admin Group
Avatar

Joined: 05 Jan 2012
Location: San Diego, CA
Status: Offline
Points: 5718
Post Options Post Options   Thanks (0) Thanks(0)   Quote Tom H Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 5:46pm
Yep, in the IPC world, the tolerances change the pattern when using the same Nominal Package Dimensions. 

They are part of the mathematical model. 

Read this:

For additional info on a popular lead form:

Stay connected - follow us! X - LinkedIn
Back to Top
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Posted: 13 Jan 2020 at 6:11pm
I will look at those forum posts.

Thanks again!

Back to Top
jnbrown View Drop Down
Active User
Active User


Joined: 13 Jan 2020
Status: Offline
Points: 40
Post Options Post Options   Thanks (0) Thanks(0)   Quote jnbrown Quote  Post ReplyReply Direct Link To This Post Posted: 14 Jan 2020 at 3:15pm
Originally posted by Tom H Tom H wrote:

Yep, in the IPC world, the tolerances change the pattern when using the same Nominal Package Dimensions. 

They are part of the mathematical model. 

Read this:

For additional info on a popular lead form:


Tom,

Those links do not seem to be working.
Would still like to read those.

Thanks
Back to Top
Tom H View Drop Down
Admin Group
Admin Group
Avatar

Joined: 05 Jan 2012
Location: San Diego, CA
Status: Offline
Points: 5718
Post Options Post Options   Thanks (0) Thanks(0)   Quote Tom H Quote  Post ReplyReply Direct Link To This Post Posted: 14 Jan 2020 at 3:20pm
I updated the links. 
Stay connected - follow us! X - LinkedIn
Back to Top
 Post Reply Post Reply Page  12>

Forum Jump Forum Permissions View Drop Down



This page was generated in 0.219 seconds.