Male header pins punching through pcb.

Thread Starter

Kaisha

Joined Nov 23, 2016
15
I've created two pcbs now with easyeda and had them printed off with JLCPCB. Being a self-taught/hobbyist its been a fun experience. Both have performed well except for one weird problem.

The pins on multiple male headers (from different manufacturers) under relatively light usage, seem to have disengaged from the plastic housing and have pushed through the back of the pcb, breaking the solder connection. Is this a common problem with male pin headers? Did I just somehow get really bad headers? Am I perhaps soldering them on wrong (to much heat melting the plastic or something?)?

Granted, of all the things that could go wrong faulty pin headers is rather minor, but considering this has happened multiple times now, and I have plans for a larger more permanent project in the future. I need to either figure out what I'm doing wrong, or perhaps use something else for external connections.

Any ideas?
 

dl324

Joined Mar 30, 2015
16,845
The pins on multiple male headers (from different manufacturers) under relatively light usage, seem to have disengaged from the plastic housing and have pushed through the back of the pcb, breaking the solder connection. Is this a common problem with male pin headers? Did I just somehow get really bad headers? Am I perhaps soldering them on wrong (to much heat melting the plastic or something?)?
My experience has been that the pins on the cheap ones are easily pushed through the housing. This has never caused me problems while the header is soldered to the board.

How are you applying force sufficient to break solder joints? Is this for single pins, 2 pins, more? Are you connecting ganged female pins (how many)?
 

Thread Starter

Kaisha

Joined Nov 23, 2016
15
It was with 1x6 and 2x3 male headers and I was connecting ganged female connections (again 1x6 and 2x3). The 1x6 connection is simply an adafruit FTDI 6-pin connector (this guy: https://www.adafruit.com/product/70) and the 2x3 was some generic 2x3 pin IDC connector.

The solder joints are being broken. The solder stays with the pin, but breaks away from the pcb pad. Even ample amounts of flux didn't seem to change anything.

If it was cheap pin headers. How does one differentiate between 'cheap' and 'expensive' ones??

I wasn't pushing that hard...
 

MrSoftware

Joined Oct 29, 2013
2,188
If you're talking about the headers that I think you're talking about, the plastic just holds them in place, the pins can slide in and out of the plastic relatively easily. However I'm really surprised that you were able to push the pins out of a soldered joint. Please post a close up picture of your solder joint if you can. When you solder the pins, be sure to get the pin itself hot enough to melt the solder, and keep the heat on long enough to also get the pads hot enough to melt the solder. And fill the joint completely. If the pin and pad don't get hot enough to melt solder themselves then it won't adhere properly.
 

Thread Starter

Kaisha

Joined Nov 23, 2016
15
So most likely its my poor soldering skills. Fair enough, that's something I can work on.

What temperature should I be using for soldering?
 

dl324

Joined Mar 30, 2015
16,845
The solder joints are being broken. The solder stays with the pin, but breaks away from the pcb pad.
This is a sign of a bad solder joint. When the joint is made properly, the solder forms an intermetallic bond with the pieces being joined.
If it was cheap pin headers. How does one differentiate between 'cheap' and 'expensive' ones??
The only way I've been able to tell is to remove them and see if the pins move easily. I've only had one product that I know was assembled in China and had cheap headers.

The good quality headers I have take more pressure than I can apply with my fingers to dislodge them. I have to use two pairs of pliers; one to hold the pin and one to hold the housing.
 

TeeKay6

Joined Apr 20, 2019
573
So most likely its my poor soldering skills. Fair enough, that's something I can work on.

What temperature should I be using for soldering?
I am wondering...perhaps you are using a single-sided PCB, wherein your PCB has copper pads only on one board surface--the surface opposite to the connector housing. In this case, the pin is held to the board only by (1) the connector housing and (2) the adhesion of the backside copper pad to the PCB. That adhesion would usually not be sufficient. The normal, and better, scheme is to use a double-sided board with plated-through holes. That is, pads are on both surfaces of the PCB, and solder fills any gap between the pin and the plating of the hole, usually with solder welling up on each surface of the PCB a tiny bit around the pin. Using that construction it is almost impossible to push a pin through the hole.
 

Thread Starter

Kaisha

Joined Nov 23, 2016
15
I appreciate the links SamR, now I know what to look for in the future.

I was using 2 sided boards, with the back side being primarily a ground plane. Now that I come to think of it, the ground pins were almost always the ones to fail. The ground pin being connected to the ground plant would have the highest heat dissipation, so it would be the coolest of the pins, and hence would explain what I am seeing.

Is there any easy way to ensure the ground pin pads gets hot enough without heating up the entire pcb?
 

jpanhalt

Joined Jan 18, 2008
11,087
I appreciate the links SamR, now I know what to look for in the future.

I was using 2 sided boards, with the back side being primarily a ground plane. Now that I come to think of it, the ground pins were almost always the ones to fail. The ground pin being connected to the ground plant would have the highest heat dissipation, so it would be the coolest of the pins, and hence would explain what I am seeing.

Is there any easy way to ensure the ground pin pads gets hot enough without heating up the entire pcb?
That doesn't seem to make sense. Please show a picture.
 

dl324

Joined Mar 30, 2015
16,845
Is there any easy way to ensure the ground pin pads gets hot enough without heating up the entire pcb?
Use a soldering iron tip with sufficient heat reserve so that it isn't cooled significantly by the ground plane. With the tip touching the pin and the pad, apply solder to the pad and pin on the side opposite from the tip. Let the solder flow until you have sufficient solder to form the joint. Remove the iron tip and let the joint cool without disturbing the melted solder (this isn't required for eutectic solder).

Don't touch the solder to the iron tip. That will ensure that the pin and pad are heated sufficiently.

Heat transfer is facilitated by flux. If you have some, use it.
 

TeeKay6

Joined Apr 20, 2019
573
I appreciate the links SamR, now I know what to look for in the future.

I was using 2 sided boards, with the back side being primarily a ground plane. Now that I come to think of it, the ground pins were almost always the ones to fail. The ground pin being connected to the ground plant would have the highest heat dissipation, so it would be the coolest of the pins, and hence would explain what I am seeing.

Is there any easy way to ensure the ground pin pads gets hot enough without heating up the entire pcb?
Yes, pads in a large copper area (e.g. ground plane) should have "thermal relief". Search Web for "thermal relief" or "thermal relief pad"; you will find plenty of info.
 

bassbindevil

Joined Jan 23, 2014
824
Name-brands like Molex or AMP or Samtec would be a safe bet, but not inexpensive. Unless you buy from a surplus dealer. A quick search turns up 20 pin AMP headers, a mere 4 bucks from Digikey, but 75 cents from Alltronics:
https://www.alltronics.com/cgi-bin/item/28J267/15/1-x-20-Vertical-Pin-Header
Headers on old computer boards from the '80s and '90s will be good quality, but desoldering them isn't fun; last time I just pulled out one pin at a time using pliers, then stuck 'em back in the plastic strip.
Anyway, if the pins weren't soldered solidly to the pads, that doesn't sound like a trustworthy solder joint. It should take an extreme amount of force to pull or push a pin through a soldered joint. You may need to scrape the pins clean before soldering, but that also raises doubts about the quality of the plating on the contact end.
 

JohnInTX

Joined Jun 26, 2012
4,787
I agree with the need for thermal reliefs on large ground planes. Your PCB package should have a way to connect pins/pads with thermal reliefs when you do the copper pour. Similar problems can arise when pins with a large thermal mass are soldered, too. Same solution.
 

jpanhalt

Joined Jan 18, 2008
11,087
You mean this whole"problem" is because the TS didn't properly solder the pins? Did the circuit work without "ground" pins soldered?

You don't require thermal reliefs to properly solder a pin to a ground plane. You just need to know how to solder. Pictures would help, but haven't been made available.
 

MisterBill2

Joined Jan 23, 2018
18,176
The problem is that the holes in the circuit board are excessively large. Headers need to be pressed in to the holes, if they just slide in easily the holes are far to large a diameter. This is a common problem, even on expensive laptop computers. If you have the headers with 0.025 inch square pins on 0.010 inch centers the holes must not be larger than about 0.031 inch diameter, 0.030 is better. A pressed in pin that is then soldered should not come loose without a lot of abuse.
 
Top