how to test for flaky solder connections?

Go To Last Post
20 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

When I hand-assemble boards, I always use lots of flux, and test connections for continuity.  Nevertheless, I still encounter flaky connections, particularly with through-hole header pin connectors.  Over the past few days, I was debugging a seemingly random problem that would usually (but not always) go away when I connected my scope probes.  The torque from the probe clamp on the header pin was improving the connection, making the problem go away.

 

Is there any systematic way to test for these kind of problems?

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Working with tubeamps I used a chopstick to poke at connections to check them ... just like you did with the probe..

I’ve repaired some amps and guitarpedals with ‘flaky connections’ and used the ‘chopstick method’ to find the part/component causing the issues and then reheat/flow all the solderjoints on that bit and ... well ... rewind and repeat ? 
 

Don’t know if you can call that ‘systematic’, but it’s worked for me ! ;)

Last Edited: Sun. Jan 17, 2021 - 04:04 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

I feel your pain, I too have had a hard time soldering these pins, for what ever reason they don't want to take solder very well.

I use a 10x magnifier to examine each pin, looking for smooth joints and fillets between pin and pad, reflux and reheat as needed, take your time when soldering, don't rush it.

And don't use too much solder, if the joint looks like a blob, remove the excess and try again.

 

Jim

 

 

(Possum Lodge oath) Quando omni flunkus, moritati.

"I thought growing old would take longer"

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Probably nothing new here, but I’ll mention some thoughts:

  1. Make sure that the parts are “clean”.  Old parts with an oxide layer won’t solder well.
  2. Use the right solder.  I find rosin core with lead MUCH easier to use than the leadless solder.
  3. Use the right heat, and make sure you are heating both the lead (pin) and the pad, not just one of them. 
  4. Touch your solder to the lead & pad, NOT to the soldering iron tip.  The solder should flow easily when touching the hot lead and pad.
  5. Watch the solder.  It should nicely flow around the components.  If it sort of blobs up, and doesn’t flow and spread, then either the lead had an oxide layer, or the lead / pad wasn’t hot enough.
  6. Think about using solder paste on your hand assembled boards.  I use a syringe with a cut off (shortened) needle to put a small amount of paste on each pad, then place all of the parts, and then fry them on an electric frying pad.  For non-commercial one-offs and home use, one doesn’t need a reflow oven. 
  7. I, too, use a jeweler’s loupe, sometimes even with a headband magnifier, under a bright light bulb, to visually inspect the boards and soldering after assembly.
  8. Besides stressing the joints mechanically, (with an O’scope probe, or a chip stick), you can also stress them thermally, with a aerosol can of cooling spray and a small straw to direct it.
  9. When I have to route a trace under a chip I try to: 1) Make sure there is no pad on the bottom of the chip… (Don’t ask), 2) Provide as much clearance as I can from the chip’s pins, as one can’t see any solder bridges under the chip once it is assembled.
  10. If you solder has embedded flux, you really shouldn’t need to add additional flux for most soldering.    

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the tips.  I use made in Canada 60/40 SN/PB solder with 3% flux.  I find it much harder to solder parts if I don't flux them first.  I'll definitely look for a jeweler's loupe.  Currently I use a hand-held magnifying glass and a 12-LED desk lamp for inspecting boards.

 

I was wondering if there is any way of measuring the quality of the solder connections.  Perhaps an avalanche pulse generator and a high-speed scope to look for reflections?

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ralphd wrote:
Perhaps an avalanche pulse generator and a high-speed scope to look for reflections?

Those tools are usually out of reach for most hobbyists, but if you have access and a known bad board, worth a shot to see what you find.

Another possible test is to xray or electron microscope the joint.

 

Jim

 

 

(Possum Lodge oath) Quando omni flunkus, moritati.

"I thought growing old would take longer"

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ralphd wrote:

I was wondering if there is any way of measuring the quality of the solder connections.  Perhaps an avalanche pulse generator and a high-speed scope to look for reflections?

 

 

Just out of curiosity : How would You use them the check the quality of solder connections ? Measure the delay between the sent pulse(s) and the reflections or ? 

But/and : wouldn't that require hooking up some probes to the board, which then might make a flaky connection connect as long as the probe is connected ? or ? 

 

And what if the problem you described was the other way around ? Ie. that the connection was OK as long as there's nothing connected to it / nothing probing it ? 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It has been 15 years since I worked on an SMD line; they had both AOI and ICT.

 

https://en.wikipedia.org/wiki/Automated_optical_inspection

 

https://en.wikipedia.org/wiki/In-circuit_test

 

ICT can miss what we called cold solder joints, but even at that time, AOI could catch some of them if the solder joint was visible. I have no idea of state of the art now. I use a 10x jeweler's loupe to look over the (SMD) boards from my toaster oven and touch up the pads that look odd; my hacker version of ICT is to test for the ESD diodes (and nothing else). Of the dozens of boards I have toasted, I have only had one act truly wonky; I think a via may have been the problem; that is something that both ICT and AOI can miss; it pretty much is a case of sending it to the customer and let them find the problem (aggravating).

 

ralphd wrote:
Perhaps an avalanche pulse generator and a high-speed scope to look for reflections?

 

Since I know ICT can pass current through a tiny whisker in a cold solder junction, I would have more hope for putting the board in a water bath, with a dolphin trained to ping for defects.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok I can understand if you have problems with cheap lead free solder, but 60/40 SN/PB should flow fine, if you really have problems use solder paste even if you use a solder iron. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ron_sutherland wrote:

 my hacker version of ICT is to test for the ESD diodes (and nothing else).

I like this idea.  When I test for continuity, I have to push on the leads a bit, which could create a connection when there isn't good solder contact with the pad.  Testing the VCC & GND protection diodes will allow me to test from the 0.1" headers.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:

Make sure that the parts are “clean”.  Old parts with an oxide layer won’t solder well.

 

 

I have a lot of parts left over from a very salty air environment, and they don't solder worth a darn, but a little scraping with an X-acto (or any sharp blade) works wonders.  Optically they shine after scraping, so it's easy to tell which parts have been and which have not.

 

Also seconding looking at the joint shape.  I inspect all my solder joints, sometimes several times, and with magnification.  A ball (or worse, a doughnut) around the pin is bad.  Ideally they should look like a little circular tent, with the pin being the center pole.  S.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Surely testing is a backstop, and the solution is to improve the quality of the solder joints?

What temperature is your iron and what wattage is it?

#1 Hardware Problem? https://www.avrfreaks.net/forum/...

#2 Hardware Problem? Read AVR042.

#3 All grounds are not created equal

#4 Have you proved your chip is running at xxMHz?

#5 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand."

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Brian Fairchild wrote:
Surely testing is a backstop

 

Backstop? Interesting... If I were hoping to ditch testing, I would start by getting rid of the iron and using a process where a proper reflow profile has at least some chance of being followed.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ron_sutherland wrote:

Brian Fairchild wrote:

Surely testing is a backstop

 

 

Backstop? Interesting... If I were hoping to ditch testing, I would start by getting rid of the iron and using a process where a proper reflow profile has at least some chance of being followed.

 

And, of course, if you had unlimited time and money you'd just hire someone to do it for you.  I'm presuming here that the OP doesn't have unlimited resources to 'do it right', as it were.  S.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Scroungre wrote:

And, of course, if you had unlimited time and money you'd just hire someone to do it for you.  I'm presuming here that the OP doesn't have unlimited resources to 'do it right', as it were.  S.

 

Part of the problem is the board design.  I've bought some modules that have rather small pads around the header pin holes.  I like nice big oval pads that make it easy to tell when the solder has good contact between the pad and pin.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Poor solder joints normally means that your iron is too cold and too low powered.

#1 Hardware Problem? https://www.avrfreaks.net/forum/...

#2 Hardware Problem? Read AVR042.

#3 All grounds are not created equal

#4 Have you proved your chip is running at xxMHz?

#5 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand."

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

if the header pins are square and the holes are round and via's (cobber inside the holes) I would fill the holes with solder paste so some stay in the hole, heat the pin and the hole thing is soldered.

 

And make sure that the header pins are clean! 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ralphd wrote:

Part of the problem is the board design.  I've bought some modules that have rather small pads around the header pin holes.  I like nice big oval pads that make it easy to tell when the solder has good contact between the pad and pin.

 

Personally, I loathe those things.  My 'Eagle' PCB layout program always has those as defaults, and I can't route traces around them!  I'm incessantly going into the libraries and making the pads small again...  Except for big connectors, those I like the mechanical strength of a big pad and lots of solder.  

 

I'm not sure I can trace any board failures of mine to lack of solder contact.  Much more often are solder bridges - accidentally connecting two pins together.  Most of those I catch during inspection.   S.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Scroungre wrote:
lots of solder.  

 

err... I have that down... my latest fresh from the toaster.

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

In addition to process is education, training, mentoring, and practicing by the operator; feedback aids.

Hand Soldering - CV-5210 & CV-500 Connection Validation Soldering Station (Metcal)

[third paragraph]

Currently, the industry standard relies on a visual inspection of the solder joint. ...

[fourth paragraph]

...

CV complements the skill of the operator to judge the quality of a solder joint by introducing an objective method of evaluating solder joint quality [intermetallics]. 

...

CV-510 Metcal | Mouser via CV-510 Connection Validation™ Soldering System - Metcal | Mouser

 

"Dare to be naïve." - Buckminster Fuller