Nodemcu part holes too small pcb order wasted


#1

I did searched for esp8266 with the looking glass and found 2 parts of nodemcu. Both of them are exactly the same.

After receiving my pcb i discovered that the holes of this part are too small.

So i wish to thank (!NOT) the one who created this.

20$ wasted.


#2

Although it doesn’t help now, I have corrected a couple of these for people. If you search for which ever one you used in the forum search bar you should find a corrected version. That is also the reason you should always export the board as gerbers and check them with a gerber viewer (in this case via the drill file as text to check the hole size).

Peter


#3

You can drill them, you just have to solder them top and bottom because it removes the plating that joins the top to the bottom.


#4

I couldn’t see it, i thought it was “just fit”.

If you have hundreds of holes you cannot see what fit and what not, even on paper.
The pins of female headers i use are small, but not smaal enough in this case.

I can imagine it’s difficult in case of esp8266 because there are many versions.

So i have from aliexpress a small version of the nodemcu and a wider version.

Do you have a link to the fpfz?


#5

Yeah you can’t just print it out and look at it, you have to open up drill.txt and check the holes manually. First set of numbers is the sizing groups, and the groups below are the positions of the holes in those groups.


#6

[Stevie Griffin voice of Family Guy and little bit of English accent] : “whaaahaaat?”

“Do i have to check every hole in the Gerber files?”


#7

That’s what it’s like with free software - I tried KiCad and parts weren’t perfect either -.

I know to make parts from the technical drawings because I’m a Mech Eng, but anyone and everyone i.e. without engineering backgrounds, can make parts which don’t get checked until something goes wrong. You can look at the drill size groups to see if any are smaller than std 0.038", and work from there.

A machine is only going to do what you tell it, GIGO, so it’s basically up to the engineer to have it right.

Just patch it up as best you can, is what most manufactures do, i.e. bodge.


#8

No just print out the SVG or PDF version of the drill file and place the part on top of it. You should be able to tell if the holes are too small (unless it is really close).


#9

First you need to remember that parts are user submitted and may be wrong. I and others tend to look over parts newly posted and check them for accuracy and obvious flaws but that isn’t always done
even in parts in core. I’m working on a python script that automates this and eventually hope to clean up core but that’s likely a ways off.

I have made changes of some kind to about 6 of them and I don’t remember which one had small holes. A search for lolin in the forum search bar (always a good place to start) turns up this one which mentions small holes but there may have been more than one.

a search for esp32 turns up even more posts. If you have the name of the part you used I can probably tell you if it is one of those that I changed. There are many different versions of these boards some in odd sizes (at least one is 1.1 in wide rather than the “standard” .9in)

No the gerber file has done that for you: In what ever this is (probably all the top view parts I made yesterday. in which case I probably should look at the hole sizes on them, given this :slight_smile: ) the lines after INCH are drill sizes

T102C0.038000 is the only correct one in the list, being the correct size for a .1 header pin which is what you want. .035 (IC leads) and .040 (1N4001 diode holes) are other common sizes. The weird ones in the table will usually get rounded up to the next higher standard drill size by the board house but are inf fact sloppy and incorrect. The leading Txxx value indexes the following table of x/y coords where that hole will be drilled, but usually you only care that the initial size table is correct.

; NON-PLATED HOLES START AT T1
; THROUGH (PLATED) HOLES START AT T100
M48
INCH
T100C0.037000
T101C0.032833
T102C0.038000
T103C0.037222
T104C0.039370
T105C0.020000
%
T100
X017489Y019218
X016489Y019218
… (many more)

Peter