0603 caps' caps

[oopere]

So, yesterday I was playing building a "dead bug" style 60 MHz
oscillator. I was obtaining inconsistent signals. Seemed that depending
on touching cables the measurements changed...

Again, it turned out to be one of those 0603 capacitors whose metal caps
tend to break off quite easily, leaving an unreliable contact. Flexing
the cable to the board flexed the connector which in turn moved the
metal cap of the capacitor. I re-soldered a new capacitor and... Wait,
didn't I hear a little "click"? Indeed, the capacitor's cap went off
again (!) from the strain when the lead solidified.

It is not the first time having capacitor leads break when prototyping.
Most of the time, this came from flexing the PCB by an imperceivable
amount, but until yesterday I had no noticed this happening from the
solder cooling!

Pere

 

[Rich Grise]

So, yesterday I was playing building a "dead bug" style 60 MHz oscillator.
I was obtaining inconsistent signals. Seemed that depending on touching
cables the measurements changed...

Again, it turned out to be one of those 0603 capacitors whose metal caps
tend to break off quite easily, leaving an unreliable contact. Flexing the
cable to the board flexed the connector which in turn moved the metal cap
of the capacitor. I re-soldered a new capacitor and... Wait, didn't I hear
a little "click"? Indeed, the capacitor's cap went off again (!) from the
strain when the lead solidified.

It is not the first time having capacitor leads break when prototyping.
Most of the time, this came from flexing the PCB by an imperceivable
amount, but until yesterday I had no noticed this happening from the
solder cooling!

It sounds like you're using too much solder or too much heat. Breaking
just from soldering shouldn't happen. Use the minimum possible amount
of solder, and solder it quickly.

Good Luck!
Rich

 

[John Devereux]

Some soldering irons reach insane temps.

But I agree, a surfmount cap with one soldered to copperclad, and the
other end sticking up, is very easily damaged.

I used one like this to bypass a divider/reference voltage and got
quite large errors from the wire moving. It was creating a strain on
the capacitor body, and I guess modulating the value.

 

[oopere]

Are you using un-leaded solder ?

Michael Kellett

I am using conventional leaded solder (0.3mm diameter). The iron is
decent, suitable for SMD work, with temperature regulation and so on...

Should I expect different behavior with _un_leaded solder?

Pere

 

[oopere]

I used one like this to bypass a divider/reference voltage and got
quite large errors from the wire moving. It was creating a strain on
the capacitor body, and I guess modulating the value.

As ii pointed out in the original post, I have had _several_ problems
related to capacitors "breaking". On some prototypes I even had to
reinforce a PCB (0.8mm, though) with some perpendicular PCB strips to
avoid the board flexing and ruining capacitors. This has also happened
on small 1.6mm PCBs.

I did not note if this was related to the soldering iron temperature. It
is true that I sometimes regulate the temperature higher than the
minimum (up to 330 C) to enhance speed but, then, the process is quick.
But perhaps not so quick as to avoid the small thing to become overheated...

I have just searched for some information and found this
http://www.symmetron.ru/suppliers/epcos/ceram/00880098.pdf

"Manual soldering with a soldering iron is to be avoided, hot-air
methods are recommended for making repairs."

So, at which temperature do you regulate your iron when soldering these?

Pere

 

[TheM]

I am using conventional leaded solder (0.3mm diameter). The iron is decent, suitable for SMD work, with temperature regulation and
so on...

Should I expect different behavior with _un_leaded solder?

Pere

If you have enough space, use 2.5mm pitch ceramic caps instead of smd.
Soldering one end of smd dead bug style means it will break at minimal
tension. Use smd only where you know there won't be any tension. Not
on connecting wires, for example.

M

 

[John Devereux]

As ii pointed out in the original post, I have had _several_ problems
related to capacitors "breaking". On some prototypes I even had to
reinforce a PCB (0.8mm, though) with some perpendicular PCB strips to
avoid the board flexing and ruining capacitors. This has also happened
on small 1.6mm PCBs.

I did not note if this was related to the soldering iron
temperature. It is true that I sometimes regulate the temperature
higher than the minimum (up to 330 C) to enhance speed but, then, the
process is quick. But perhaps not so quick as to avoid the small thing
to become overheated...

I have just searched for some information and found this
http://www.symmetron.ru/suppliers/epcos/ceram/00880098.pdf

"Manual soldering with a soldering iron is to be avoided, hot-air
methods are recommended for making repairs."

So, at which temperature do you regulate your iron when soldering these?

I have never actually had any problems when manually soldering
capacitors to actual circuit boards. Things can start to come apart if
you remove again them incorrectly, or if you are not using a proper
PCB. But I would guess lower temperatures and minimal solder would be
good. I usually use 0805, so that may be a difference too.

 

[Archimedes' Lever]

Use a 600 degree tip instead of a 700. Perform the solder operation
within a 2 second window. Refrain from reflowing the solder joint
whenever possible.

If you must use the actual smd part to get reliable data for operation,
you can glue it to an insulator, like a silica transistor heat sink or a
small piece of PCB material, and solder the wires from it to the other
circuit parts. If a standard dipped ceramic cap of equiv. value could be
used to characterize the design, it would be a lot easier on you.

It isn't the solder cooling that causes the problem, per se. The
solder process itself almost detaches the terminations on SMD chip
capacitors, so it is the heat in-flow during hand soldering that does the
most damage, hence the cooler tip temp suggestion.

This is why SMD reflow ovens bring up the assembly temp on a gradual
ramp, then reflow, then ramp down steeper, but not as fast as ambient air
would.

You can reduce the issues if you pre-heat the assembly to a couple
hundred degrees F before soldering as well. It is hardier to solder hot
parts though, however, less damaging to caps.

I used to glue a whole bunch of SMD parts

 

[Archimedes' Lever]

Should I expect different behavior with _un_leaded solder?

Pere

Lead free solders are usually a hotter process, so more damage to a cap
would occur.

 

[JosephKK]

As ii pointed out in the original post, I have had _several_ problems
related to capacitors "breaking". On some prototypes I even had to
reinforce a PCB (0.8mm, though) with some perpendicular PCB strips to
avoid the board flexing and ruining capacitors. This has also happened
on small 1.6mm PCBs.

I did not note if this was related to the soldering iron temperature. It
is true that I sometimes regulate the temperature higher than the
minimum (up to 330 C) to enhance speed but, then, the process is quick.
But perhaps not so quick as to avoid the small thing to become overheated...

I have just searched for some information and found this
http://www.symmetron.ru/suppliers/epcos/ceram/00880098.pdf

"Manual soldering with a soldering iron is to be avoided, hot-air
methods are recommended for making repairs."

So, at which temperature do you regulate your iron when soldering these?

Pere

Just one question: Does the failing caps seem to be batch related?