Welcome, Guest
Username: Password: Secret Key Remember me

TOPIC: Estimating Footprints from a PCB

Re: Estimating Footprints from a PCB 6 years 5 months ago #16422

  • Jon Chandler
  • Jon Chandler's Avatar
  • Offline
  • Moderator
  • Posts: 365
  • Thanks received: 352
article on Hack-A-Day: Reverse engineering salvaged part footprints.  I show a method to accomplish this technique using Infanview, a image manipulation program I highly recommend.

Read full article...

Re: Estimating Footprints from a PCB 6 years 5 months ago #12898

I don't have a PCB layout to reverse engineer right now, but I gave this a go out of interest. Great tip Jon, Infranview makes it simple to get measurements etc; nice application of use for electronics!

Re: Estimating Footprints from a PCB 6 years 5 months ago #12899

  • W4GNS
  • W4GNS's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 11
  • Thanks received: 43
I look forward to looking into Infranview also, if I can manage to stay awake till midnight sometime. Which is the time when I can start serious downloading.

Edit: It's a very small download. 2 Mb

Re: Estimating Footprints from a PCB 6 years 5 months ago #12902

  • Jon Chandler
  • Jon Chandler's Avatar
  • Offline
  • Moderator
  • Posts: 365
  • Thanks received: 352
It should go without saying, but...

Resolution is important to this process. The resolution for the demo board I've pictured here is about 1200 pixels across the width. This is ok to measure larger objects on the board. But let's say I want to measure the dimensions of one of the quad flat packs in the lower right corner.

To make decent measurements, I'd need a high resolution picture of that area - taken with a macro lens, through a microscope or maybe with a scanner limited to that area and set to maximum resolution.

As it is in the current picture, the entire area of one of the quad flat packs is about 100 pixels wide, and each pad spans maybe 10 pixels. There's not enough resolution to get good detail in a measurement.

Re: Estimating Footprints from a PCB 6 years 5 months ago #12904

  • Jon Chandler
  • Jon Chandler's Avatar
  • Offline
  • Moderator
  • Posts: 365
  • Thanks received: 352
I used a macro lens on my dSLR to take a closeup of the quad flat pack.



Measuring as close as possible with a ruler, from extreme end of one pad to the extreme end of the other was 13 mm.

To calibrate, Irfanview told me this distance was 738 pixels. So

738 pixels / 13 mm = 56.5 pixels per mm.

I defined a new unit called scaled mm, and entered 56.5 as the calibration factor. Measuring across the pads using this calibration yielded a size of 12.98 mm. I think we can call that close enough.

Next I measured one of the pads using Irfanview: 1.54 mm x 0.44 mm

Checking the Eagle library that has this footprint, I found the actual pad size was 1.5 mm x 0.5 mm.

If I had re-created this part using measurements taken from the photo, it would work well. This is a great technique using software I already use frequently.
Attachments:

Re: Estimating Footprints from a PCB 6 years 4 months ago #13054

  • Jon Chandler
  • Jon Chandler's Avatar
  • Offline
  • Moderator
  • Posts: 365
  • Thanks received: 352
Of course it goes without saying (sadly, much of what shouldn't need saying seems to!) that this technique is most useful to duplicate an unusual footprint from a circuit board.

If you're designing a footprint for a new component, the data sheet should be the first and ultimate reference. Manufacturers always provide information about recommended PCB layouts. Sometimes you may want to make some customizations, but this is the place to start.

For those rare cases where chip manufacturers don't provide any information on recommended layouts Microchip's Packaging Specifications is almost the Bible on the subject.

Attachments:
Time to create page: 0.259 seconds