Author Topic: Common PCB FHS and pad diamaters  (Read 7935 times)

0 Members and 1 Guest are viewing this topic.

Offline ampdoctorTopic starter

  • Frequent Contributor
  • **
  • Posts: 266
  • Country: us
Common PCB FHS and pad diamaters
« on: October 10, 2014, 01:08:48 am »
I was doing some organizing over here the past day or two and stumbled across an old spreadsheet from the mid 90's, and I have no idea where it came from. It contains an exhaustive list of lead, finished hole, and pad sizes ranging from to 0.012 to 0.216 inches. 

In the table below I listed some of the more common sizes people are likely to encounter, and the holes and preferred pads seem a bit on the large side. Possibly due to old school mfg tolerances? Since everything was in imperial measures I added the metric equivalents. Anyway, does this seem like a workable list for general purpose usage? If anything seems uniquely out of whack, and if so what changes or modifications might be recommended?


Nominal lead         min FHS +/- 0.003"  min. pad dia          pref. pad dia

20 mil / 0.5 mm    34 mil / 0.9 mm       58 mil / 1.5 mm     68 mil / 1.8 mm
24 mil / 0.6 mm    38 mil / 1.0 mm       64 mil / 1.6 mm     76 mil / 1.9 mm
28 mil / 0.7 mm    43 mil / 1.1 mm       66 mil / 1.7 mm     86 mil / 2.2 mm
31 mil / 0.8 mm    46 mil / 1.2 mm       70 mil / 1.8 mm     92 mil / 2.3 mm
39 mil / 1.0 mm    58 mil / 1.5 mm       82 mil / 2.1 mm     106 mil / 2.7 mm
71 mil / 1.8 mm    90 mil / 2.3 mm       106 mil / 2.7 mm   180 mil / 4.6 mm
79 mil / 2.0 mm    98 mil / 2.5 mm       122 mil / 3.1 mm   196 mil / 5.0 mm
91 mil / 2.3 mm   106 mil / 2.7 mm      130 mil / 3.3 mm   260 mil / 6.6 mm
 

Offline pstemari

  • Contributor
  • Posts: 28
Re: Common PCB FHS and pad diamaters
« Reply #1 on: October 14, 2014, 03:54:12 pm »
Those are pretty huge compared to the restring settings in Eagle from the Laenpcb or Sparkfun configs.

Single-sided vs double-sided, perhaps, and more importantly, plated-through holes?
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21681
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Common PCB FHS and pad diamaters
« Reply #2 on: October 14, 2014, 05:14:29 pm »
Those sound pretty reasonable.  A little on the sloppy side, maybe something about automatic insertion (or hand insertion, for that matter)?

I've done THT designs with 30-35 mil holes for most stuff (resistors, etc.), >= 10 mil annular ring.  This comes in just a little bit below what's in the table.  Seems okay.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline ampdoctorTopic starter

  • Frequent Contributor
  • **
  • Posts: 266
  • Country: us
Re: Common PCB FHS and pad diamaters
« Reply #3 on: October 15, 2014, 01:30:48 am »
As I commented initially, the sizes did seem a bit loose but not horribly so. I'd say there's a couple of factors at play. First, given that the spreadsheet was created in '97 I don't think the machine tolerances back then were nearly what they are today even at the cheapest low end board houses, so I'm sure they had to figure in a bit of slop into the sizes. If the CNC drill is off by a mil or two you could miss the pad entirely. Same with insertion machines. You could wind up with a lot of bent leads. Another thing is that the finished hole sizes square up nicely with standard imperial drill rack sizes which make things quite convenient.

Quote
Those are pretty huge compared to the restring settings in Eagle from the Laenpcb or Sparkfun configs.

Single-sided vs double-sided, perhaps, and more importantly, plated-through holes

I don't always trust the pre-built components from any of the software vendors. Invariably, the pads are grossly undersized and the components themselves often need adjustment, so I always wind up building my own libraries. NI Ultiboard are particularly rancid with annular rings typically in the range of 5-7 mil. As T3sl4co1l brought up in passing, in my experience you really want to be >= 10 mil but not so large you can't get a good fill.

Given that these were listed as FHS I'm going to assume they're plated through holes. If they were single sided non-plated holes I'd want the through hole as close to the size of the component lead as I could get them.

Since I didn't get any OMFG are they insane comments, I suppose I'll leave the list unchanged for reference. There are also listings for solid and stranded wire sizes from about 28awg up to 12awg along with most imperial mounting hardware sizes. If somebody would like to see something specific, let me know and I can edit my original post and add them as well.
 

Offline pstemari

  • Contributor
  • Posts: 28
Re: Common PCB FHS and pad diamaters
« Reply #4 on: October 23, 2014, 03:57:05 am »
The closest standard US number drill sizes are

 34 mil: #65 35 mils or #66 33 mils
 38 mil: #62
 43 mil: #57
 46 mil: #56 46.5 mils
 58 mil: #53 59.5 mils
 90 mil: #43 89 mils
 98 mil: #40
106 mil: #36 106.5 mils

which isn't particularly close.
« Last Edit: October 23, 2014, 04:04:49 am by pstemari »
 

Offline ampdoctorTopic starter

  • Frequent Contributor
  • **
  • Posts: 266
  • Country: us
Re: Common PCB FHS and pad diamaters
« Reply #5 on: October 24, 2014, 02:51:59 am »
Remember it says MINIMUM, which implies a little bigger would be better. You're also not taking into account that these are finish hole sizes so you have to add a few thou to allow for plating. For example, a 34mil fhs would require say 2 mil oversize so you'd use a 36 mil hole before plating which is a #64, and on down the list. These days most of the houses want FHS and not drill size, unless you have very specific requirements, and they'll make the adjustment in the setup to allow for plating thickness or whatever other factors may apply to their particular mfg process.

The purpose of this post isn't to argue semantics or how long is a piece of string. The point is to present a set of "safe" values that people might be able to use as a universal baseline for layouts or parts creation. In other words you won't get DRM errors due to undersize annular rings, pads big enough that they wont lift under rework and not so big that you can't get a good fill and so on. Or if a mfg decides to change your drill file to match their tooling you won't get an unusable board back due to rounding or conversion errors. Kind of an "if you're in doubt use this and you won't get in trouble" type of list.
 

Offline pstemari

  • Contributor
  • Posts: 28
Re: Common PCB FHS and pad diamaters
« Reply #6 on: December 28, 2014, 06:37:15 am »
When I said "a bit sloppy", I was mostly referring to the component lead vs FHS.  Yes, I agree that Eagle's restring settings are pretty aggressive and make for tiny pads.

FWIW, the default drill list in Eagle is 0.5mm-1.5mm by tenths, plus a few larger holes, which seems fairly reasonable.  I was fairly confused about the odd sizes it displayed until I went through the appropriate handsprings to display the list in millimeters instead of inches

I would assume most of the board houses are using metric drills--US number drills in that size range are a mess--and that they bump up to the next larger standard drill size when allowing for plating. For example, if you call for a 0.6mm hole and plating adds 1 mils == 0.0254mm, that they drill a 0.65mm or a 0.7mm and call it good.

Looking at Harvey Tool (selected purely at random), they have carbide drills down to 0.05mm, but once you get to 0.5mm the metric sizes are all in 0.1mm increments.

What is a good allowance between hole & lead?  0.1mm, 0.2mm?  Seems like I've been filling a fair number of over-sized holes with solder of late.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf