Lead-Free Solder: Weird Behavior

[Daniel Haude]

Hello,

I've been trying to solder wires to copper-plated piezo actuators using
lead-free electronics solder. It turns out that within a fraction of a
second I get something that resembles a solder joint, but then the solder
immediately retracts to other parts of the surface, leaving behind a
dullish-grey area that won't wet any more no matter what I try.

I've been soldering piezos a lot, and I've been using lead-free solder a
lot (the stuff used for plumbing; it has a higher melting point than the
electronics stuff but is otherwise a similar alloy). Only I've been using
nickel- and gold-over-nickel plated piezos which solder like a charm.

Problem is that in my application I can't have lead, and I can't have
magnetic material.

I've tried to scrape off that grey layer, but it seems like I hit the
piezo material right away. I can't tell if I have any copper left because
the copper is only some three microns in the first place.

Of course the stuff solders as expected with lead solder (which, however,
won't stick to the surface spoilt by the lead-free stuff).

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.

The flux is not the culprit; I'm using both resin and a phosphorous
acid based stuff, and both don't affect soldering with lead-containing
solder.

Any hints appreciated,
--Daniel

 

[Roger Hamlett]

"Daniel Haude" <haude@kir.physnet.uni-hamburg.de> wrote in message
news:slrnd3ts9t.ep4.haude@kir.physnet.uni-hamburg.de...

Hello,

I've been trying to solder wires to copper-plated piezo actuators using
lead-free electronics solder. It turns out that within a fraction of a
second I get something that resembles a solder joint, but then the
solder
immediately retracts to other parts of the surface, leaving behind a
dullish-grey area that won't wet any more no matter what I try.

I've been soldering piezos a lot, and I've been using lead-free solder a
lot (the stuff used for plumbing; it has a higher melting point than the
electronics stuff but is otherwise a similar alloy). Only I've been
using
nickel- and gold-over-nickel plated piezos which solder like a charm.

Problem is that in my application I can't have lead, and I can't have
magnetic material.

I've tried to scrape off that grey layer, but it seems like I hit the
piezo material right away. I can't tell if I have any copper left
because
the copper is only some three microns in the first place.

Of course the stuff solders as expected with lead solder (which,
however,
won't stick to the surface spoilt by the lead-free stuff).

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.

The flux is not the culprit; I'm using both resin and a phosphorous
acid based stuff, and both don't affect soldering with lead-containing
solder.

Any hints appreciated,
--Daniel
Are you using a different iron?.

Lead free solders, except special varieties designed to handle lead
contaimination, hate having any lead present. if you are using an iron
that still has a tinned coating on the tip from using a lead based solder,
it could give the behaviour you are seeing...

Best Wishes

 

[John Woodgate]

I read in sci.electronics.design that Daniel Haude
<haude@kir.physnet.uni-hamburg.de> wrote (in
<slrnd3ts9t.ep4.haude@kir.physnet.uni-hamburg.de&gt about 'Lead-Free
Solder: Weird Behavior', on Mon, 21 Mar 2005:

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.


Yes, your experience seems to indicate that. But it may not be just

Cu/Pb. I doubt that the Cu is plated directly on to the piezo material,
and if there is, for example, a Ni interlayer there, the possibility of
getting a non-wetting alloy is much greater. Especially if you are using
phosphoric acid.

Try using lead-free solder made for **electronics**, not the Sn/Cu stuff
used for plumbing.
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[John Popelish]

Daniel Haude wrote:

Hello,

I've been trying to solder wires to copper-plated piezo actuators using
lead-free electronics solder. It turns out that within a fraction of a
second I get something that resembles a solder joint, but then the solder
immediately retracts to other parts of the surface, leaving behind a
dullish-grey area that won't wet any more no matter what I try.

I've been soldering piezos a lot, and I've been using lead-free solder a
lot (the stuff used for plumbing; it has a higher melting point than the
electronics stuff but is otherwise a similar alloy). Only I've been using
nickel- and gold-over-nickel plated piezos which solder like a charm.

Problem is that in my application I can't have lead, and I can't have
magnetic material.

I've tried to scrape off that grey layer, but it seems like I hit the
piezo material right away. I can't tell if I have any copper left because
the copper is only some three microns in the first place.

Of course the stuff solders as expected with lead solder (which, however,
won't stick to the surface spoilt by the lead-free stuff).

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.

The flux is not the culprit; I'm using both resin and a phosphorous
acid based stuff, and both don't affect soldering with lead-containing
solder.

Any hints appreciated,
--Daniel

I had something very similar happen when I tried to solder positive
tempco thermistor pellets between two brass sheets. I was using tin
lead solder, and the result was similar to what you are getting. The
extremely thin silver metalization layer was simply dissolving in the
solder and leaving me with an ummetalized die. In my case, it was not
necessary to eliminate lead, so the solution (pardon the pun) was to
use 2% silver solder, minimum temperature and soldering time.

Evidently your metalization layer is quite soluble (given the
temperature and time) in the solder you are using. There is some low
temperature tin bismuth solder that might work better.

--
John Popelish

 

[John Woodgate]

I read in sci.electronics.design that John Popelish <jpopelish@rica.net>
wrote (in <423EFDE8.38C69132@rica.net&gt about 'Lead-Free Solder: Weird
Behavior', on Mon, 21 Mar 2005:

There is some low temperature tin bismuth solder that might work
better.

Back in the Dark Ages, before 1930, people used Woods Metal to fix
galena crystals for crystal receivers into their holders. maybe it
should make a come-back. However, it's 25% Pb, so no deal....
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Robert Baer]

Daniel Haude wrote:

Hello,

I've been trying to solder wires to copper-plated piezo actuators using
lead-free electronics solder. It turns out that within a fraction of a
second I get something that resembles a solder joint, but then the solder
immediately retracts to other parts of the surface, leaving behind a
dullish-grey area that won't wet any more no matter what I try.

I've been soldering piezos a lot, and I've been using lead-free solder a
lot (the stuff used for plumbing; it has a higher melting point than the
electronics stuff but is otherwise a similar alloy). Only I've been using
nickel- and gold-over-nickel plated piezos which solder like a charm.

Problem is that in my application I can't have lead, and I can't have
magnetic material.

I've tried to scrape off that grey layer, but it seems like I hit the
piezo material right away. I can't tell if I have any copper left because
the copper is only some three microns in the first place.

Of course the stuff solders as expected with lead solder (which, however,
won't stick to the surface spoilt by the lead-free stuff).

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.

The flux is not the culprit; I'm using both resin and a phosphorous
acid based stuff, and both don't affect soldering with lead-containing
solder.

Any hints appreciated,
--Daniel
Firstly, it is known that there is a problem with (standard tin-lead)

solder on gold.
It works very well, and looks super when done.
However, as time goes on, the solder will seperate from the gold as
if nothing was done.
The solution to this is a standard practice: slobber solder on,
remove and sloober again - until all the damn gold is leached from the
surface; *then* solder the connection.

Secondly: plumbers solder is not lead free, and usually uses ACID
(!!); there are numerous strong recommendations against use of plumbers
solder. And it is *not* considered fit for electronics use.

There are numerous *electronic* lead-free solders:
* Sn 95.5, Ag 3.8, Cu 0.7 - MP 217C / 423 F; adopted by NEMI adn many
high volume consumer OEMs
* Sn 96.5, Ag 3.5 - MP 221C / 430F
* Sn 95, Sb 5 - MP 232C / 450F
Of all of these, i would strongly recommend the first, because it
does have copper and would be less likely to leach copper from your parts.
And it is clear from your description that leaching is the primary
culprit.

 

[Robert Baer]

Tim Wescott wrote:

John Popelish wrote:

Daniel Haude wrote:

Hello,

I've been trying to solder wires to copper-plated piezo actuators using
lead-free electronics solder. It turns out that within a fraction of a
second I get something that resembles a solder joint, but then the
solder
immediately retracts to other parts of the surface, leaving behind a
dullish-grey area that won't wet any more no matter what I try.

I've been soldering piezos a lot, and I've been using lead-free solder a
lot (the stuff used for plumbing; it has a higher melting point than the
electronics stuff but is otherwise a similar alloy). Only I've been
using
nickel- and gold-over-nickel plated piezos which solder like a charm.

Problem is that in my application I can't have lead, and I can't have
magnetic material.

I've tried to scrape off that grey layer, but it seems like I hit the
piezo material right away. I can't tell if I have any copper left
because
the copper is only some three microns in the first place.

Of course the stuff solders as expected with lead solder (which,
however,
won't stick to the surface spoilt by the lead-free stuff).

Could it be that the lead-free solder dissolves copper to form some
un-solderable alloy? Sounds odd since the solder alloy contains some
copper itself.

The flux is not the culprit; I'm using both resin and a phosphorous
acid based stuff, and both don't affect soldering with lead-containing
solder.

Any hints appreciated,
--Daniel



I had something very similar happen when I tried to solder positive
tempco thermistor pellets between two brass sheets. I was using tin
lead solder, and the result was similar to what you are getting. The
extremely thin silver metalization layer was simply dissolving in the
solder and leaving me with an ummetalized die. In my case, it was not
necessary to eliminate lead, so the solution (pardon the pun) was to
use 2% silver solder, minimum temperature and soldering time.
Evidently your metalization layer is quite soluble (given the
temperature and time) in the solder you are using. There is some low
temperature tin bismuth solder that might work better.

You can (or could) get tin-lead solder that came pre-saturated with
copper, to prevent the solder from dissolving your tip or other valuable
things -- is it possible to get this same thing in lead-free solder?

Yes; see my earlier posting in this thread.

 

[budgie]

On Mon, 21 Mar 2005 21:06:10 +0000, John Woodgate
<jmw@jmwa.demon.contraspam.yuk> wrote:

Back in the Dark Ages, before 1930, people used Woods Metal to fix
galena crystals for crystal receivers into their holders. maybe it
should make a come-back.

It also used to be used for "trick" teaspoons that melted when you tried to stir
your hot cuppa.

However, it's 25% Pb, so no deal....

That didn't stop them then ;-)

 

[Daniel Haude]

On Mon, 21 Mar 2005 08:34:12 -0800,
John Larkin <jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote
in Msg. <emtt319m25juoegttcvgp9io6n28elsbfn@4ax.com>

It probably just dissolved all the copper away, then beaded up.

Considered conductive epoxy?

Yes, I'd use H20E. But this will take ages to do because I can't do all
the joints at the same time. Grrr. I'm reconsidering tin-lead solder, but
since this is a ultrahigh vacuum device lead is some sort of a no-no (too
high vapor pressure). Ultimately this is a low-temperature apparatus,
rendering vapor pressure a non-issue, but it also has to be
contaminant-free at room temp.

--Daniel

 

[Daniel Haude]

On Mon, 21 Mar 2005 16:32:27 +0000,
John Woodgate <jmw@jmwa.demon.contraspam.yuk> wrote
in Msg. <5bOkIyFbcvPCFw8d@jmwa.demon.co.uk>

Yes, your experience seems to indicate that. But it may not be just
Cu/Pb. I doubt that the Cu is plated directly on to the piezo material,
and if there is, for example, a Ni interlayer there

The mfg (Staveley Sensors) says no. They usually go Piezo-Cu-Ni-Au (if you
want gold). Nickel solders by far the best, but it's magnetic.

Try using lead-free solder made for **electronics**, not the Sn/Cu stuff
used for plumbing.

I tried both (two kinds of electronics stuff and one of the other)
with the same results.

--Daniel

 

[Daniel Haude]

On Mon, 21 Mar 2005 12:01:28 -0500,
John Popelish <jpopelish@rica.net> wrote
in Msg. <423EFDE8.38C69132@rica.net>

Evidently your metalization layer is quite soluble (given the
temperature and time) in the solder you are using.

Yes, it goes away almost immediately even when I set the iron to a
temperature where the solder barely melts.

--Daniel

 

[Daniel Haude]

On Mon, 21 Mar 2005 21:36:07 GMT,
Robert Baer <robertbaer@earthlink.net> wrote
in Msg. <b7H%d.1254$H06.109@newsread3.news.pas.earthlink.net>

Secondly: plumbers solder is not lead free, and usually uses ACID
(!!); there are numerous strong recommendations against use of plumbers
solder. And it is *not* considered fit for electronics use.

Actually what I used isn't really plumber's solder. I just called it that
because our workshop always uses it when they solder copper pipes, but it
is lead-free solder.

There are numerous *electronic* lead-free solders:
* Sn 95.5, Ag 3.8, Cu 0.7 - MP 217C / 423 F; adopted by NEMI adn many
high volume consumer OEMs

That's what I have here and what I wanted to use. After it showed those
weird effects I went back to my "old" Pb-free solder only to find that it
behaved the same.

--Daniel

 

[Daniel Haude]

On Mon, 21 Mar 2005 16:21:05 GMT,
Roger Hamlett <rogerspamignored@ttelmah.demon.co.uk> wrote
in Msg. <RvC%d.1025$Ap1.808@newsfe3-gui.ntli.net>

Are you using a different iron?.
Lead free solders, except special varieties designed to handle lead
contaimination, hate having any lead present. if you are using an iron
that still has a tinned coating on the tip from using a lead based solder,
it could give the behaviour you are seeing...

That isn't it. The solder wets fine when applied to a thicker Cu surface.

--D.

 

[John Woodgate]

I read in sci.electronics.design that Daniel Haude
<haude@kir.physnet.uni-hamburg.de> wrote (in
<slrnd3voob.fgt.haude@kir.physnet.uni-hamburg.de&gt about 'Lead-Free
Solder: Weird Behavior', on Tue, 22 Mar 2005:

The mfg (Staveley Sensors) says no. They usually go Piezo-Cu-Ni-Au (if
you want gold). Nickel solders by far the best, but it's magnetic.

Not very (and if it's alloyed it may even not be ferromagnetic). You
would likely get more magnetic effects from the currents in the
interconnections.
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Daniel Haude]

On Tue, 22 Mar 2005 12:04:11 +0000,
John Woodgate <jmw@jmwa.demon.contraspam.yuk> wrote
in Msg. <dA2cyTU7mAQCFwwe@jmwa.demon.co.uk>

The mfg (Staveley Sensors) says no. They usually go Piezo-Cu-Ni-Au (if
you want gold). Nickel solders by far the best, but it's magnetic.

Not very (and if it's alloyed it may even not be ferromagnetic).

I know it's definetely "not very", but what little there might be could be
too much. This is an STM that hangs inside a 14T magnet like a pendulum.
We found that the current model gets deflected by 1mm to the side already
at 4T. My hunch is that this is due to the nickel interlayer that was
carelessly applied when the whole thing was gold-plated. Now I'm building
a replacement and am trying to ban all ferromagnetic materials whatsoever.

That said -- I tried to make a nickel-plated 1/2" piezo scanner tube roll
on a flat surface by holding a strong permanent magnet (from a harddisk)
next to it. Nothing. That was of course at room temperature; who knows
what happens at .3K. After a few years of working with materials under
such conditions you develop a superstition.

--Daniel

 

[Ken Smith]

In article <dA2cyTU7mAQCFwwe@jmwa.demon.co.uk>,
John Woodgate <noone@yuk.yuk> wrote:

I read in sci.electronics.design that Daniel Haude
haude@kir.physnet.uni-hamburg.de> wrote (in
slrnd3voob.fgt.haude@kir.physnet.uni-hamburg.de&gt about 'Lead-Free
Solder: Weird Behavior', on Tue, 22 Mar 2005:
The mfg (Staveley Sensors) says no. They usually go Piezo-Cu-Ni-Au (if
you want gold). Nickel solders by far the best, but it's magnetic.

Not very (and if it's alloyed it may even not be ferromagnetic). You
would likely get more magnetic effects from the currents in the
interconnections.

This isn't likely in a piezo device since the DC currents are very very
small. There is nothing in the universe more magnetic per pound than a Ni
plating.

--
--
kensmith@rahul.net forging knowledge

 

[John Larkin]

On 22 Mar 2005 09:15:14 GMT, Daniel Haude
<haude@kir.physnet.uni-hamburg.de> wrote:

On Mon, 21 Mar 2005 08:34:12 -0800,
John Larkin <jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote
in Msg. <emtt319m25juoegttcvgp9io6n28elsbfn@4ax.com

It probably just dissolved all the copper away, then beaded up.

Considered conductive epoxy?

Yes, I'd use H20E. But this will take ages to do because I can't do all
the joints at the same time. Grrr. I'm reconsidering tin-lead solder, but
since this is a ultrahigh vacuum device lead is some sort of a no-no (too
high vapor pressure). Ultimately this is a low-temperature apparatus,
rendering vapor pressure a non-issue, but it also has to be
contaminant-free at room temp.

The UHV guys tend to hate epoxy more than lead. The epoxy outgasses at
room temp and deposits organic crud everywhere, and some of that leaks
out later. There are all sorts of exotic lead-free solders. Some
people solder with pure indium which (I think) has low vapor pressure.

Try Kester or Multicore or Aim. They may have a lead-free solder with
some copper content, which might help the leaching.

UHV is a huge nuisance. The people who do this tend to be unnaturally
patient. "Well, we may have a leak or we may have a fingerprint
somewhere. Let's pump it down for a week or so and see what happens."

Or use a spring.

John

 

[Daniel Haude]

On Tue, 22 Mar 2005 07:50:37 -0800,
John Larkin <jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote
in Msg. <48f041p5h5g263fukkisnl34kb33om4k2g@4ax.com>

The UHV guys tend to hate epoxy more than lead. The epoxy outgasses at
room temp and deposits organic crud everywhere, and some of that leaks
out later.

All possible, but we've been using epoxy forever and never had had
problems with it. Of course we try to not put the glue joints within
direct sight of the sample surface.

There are all sorts of exotic lead-free solders. Some
people solder with pure indium which (I think) has low vapor pressure.

Yes but it melts when you bake out the system at 150 degrees. And it makes
terrible joints.

UHV is a huge nuisance.

Yup.

The people who do this tend to be unnaturally
patient. "Well, we may have a leak or we may have a fingerprint
somewhere. Let's pump it down for a week or so and see what happens."

Yup. That's me. And this is the reason why you develop superstitions and
hate to do something that you haven't done before, even if by any rational
argument it should be safe (like: If epoxy is OK, lead should be OK).

BTW the epoxies we use are those horrenduously expensive compounds from
Epotek, not your home-depot two component glue. It's semiconductor
industry stuff I believe.

Or use a spring.

No space.

--Daniel

 

[Guy Macon]

John Larkin wrote:

UHV is a huge nuisance. The people who do this tend to be unnaturally
patient. "Well, we may have a leak or we may have a fingerprint
somewhere. Let's pump it down for a week or so and see what happens."

Or even "Let's heat it red-hot and pump it down for a week
or so while hitting it with a hammer every couple of hours."

 

[John Woodgate]

I read in sci.electronics.design that Ken Smith
<kensmith@green.rahul.net> wrote (in <d1pc8f$4tm$2@blue.rahul.net&gt
about 'Lead-Free Solder: Weird Behavior', on Tue, 22 Mar 2005:

This isn't likely in a piezo device since the DC currents are very very
small.

But what about the supplies to whatever is connected to the piezo?

There is nothing in the universe more magnetic per pound than a Ni
plating.

I find that astonishing. Ni plating is more permeable than, say,
Permalloy C, or has a higher coercivity than barium ferrite, or what?
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk