Lead free keyboard fails?

[Jan Panteltje]

I had some strange messages in the log file for a few days (Linux):
Aug 8 10:47:10 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.
Aug 8 10:47:11 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.

So I looked it up with google, makes you think of keyboard sniffers.

Bumping the keyboard on the table made it come and go.
But then finally I _only_ got those messages..... no more keys.

Opened up Logitech keyboard, there is a little PCB with a chip under a blob of
epoxy, with four wires soldered to it.
Usually it is the cable in case of a malfunction, so I wanted to shorten it a bit,
but it looked really intact (and is quite new).
Took a magnifying glass and looked at the soldering of ground, data, clock, and power wires on that board.
Was shocked, did not expect that from Logitech, just a blob of solder with the wires
sort of stuck in it, no signs of proper solder flow, looked really bad.
I re-melted the solder, problem gone.
By the color of it it looks like leadless.
Now this keyboard is _not_ under thermal stress _ever_.
It is however constantly hammered on, so mechanical vibration only.

 

[MooseFET]

I had some strange messages in the log file for a few days (Linux):
Aug 8 10:47:10 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.
Aug 8 10:47:11 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.

So I looked it up with google, makes you think of keyboard sniffers.

Bumping the keyboard on the table made it come and go.
But then finally I _only_ got those messages..... no more keys.

Opened up Logitech keyboard, there is a little PCB with a chip under a blob of
epoxy, with four wires soldered to it.
Usually it is the cable in case of a malfunction, so I wanted to shorten it a bit,
but it looked really intact (and is quite new).
Took a magnifying glass and looked at the soldering of ground, data, clock, and power wires on that board.
Was shocked, did not expect that from Logitech, just a blob of solder with the wires
sort of stuck in it, no signs of proper solder flow, looked really bad.
I re-melted the solder, problem gone.
By the color of it it looks like leadless.
Now this keyboard is _not_ under thermal stress _ever_.
It is however constantly hammered on, so mechanical vibration only.

So do we blame the lack of lead or the lack of process control?

The melting temperature of the nonlead solder is higher so it could be
that the leaded solder worked when they used a rusty nail heated on
the stove but the lead free didn't.

 

[Jan Panteltje]

So do we blame the lack of lead or the lack of process control?

The melting temperature of the nonlead solder is higher so it could be
that the leaded solder worked when they used a rusty nail heated on
the stove but the lead free didn't.

Right, I should have taken a picture of those connections.
But I am now under the impression (could be wrong) that perhaps they
used 60/40 tinned cable in leadless solder.
Because even when I re-heated it, it did not flow nicely, but
maybe I should have added some flux.
If so (that cable) then it will fail again in a while......
I re-soldered at 375 C.

The rest of the PCB does not look bad, but not to be specifically happy
about either.

 

[Gary Tait]

Was shocked, did not expect that from Logitech, just a blob of solder
with the wires sort of stuck in it, no signs of proper solder flow,
looked really bad.

What I would have done was have a crimped on plug (or lump of plastic the
bared wires run over) that mates with the PCB pads by friction, like the
membrane traces probably do.

 

[Jan Panteltje]

What I would have done was have a crimped on plug (or lump of plastic the
bared wires run over) that mates with the PCB pads by friction, like the
membrane traces probably do.

Yes the membrane is held down by connectors formed of bare tracks, with some
stuff on it, on same PCB.
There are in fact 3 such connectors, 2 on top of each other (no idea how that
works).
Indeed they could have clamped copper wires, or perhaps add a telephone type
connector, or some sort of crimped on header.
Any cent counts I guess.
For the rest Logitech is pretty good, their mice are the best.

 

[James Beck]

I had some strange messages in the log file for a few days (Linux):
Aug 8 10:47:10 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.
Aug 8 10:47:11 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.

So I looked it up with google, makes you think of keyboard sniffers.

Bumping the keyboard on the table made it come and go.
But then finally I _only_ got those messages..... no more keys.

Opened up Logitech keyboard, there is a little PCB with a chip under a blob of
epoxy, with four wires soldered to it.
Usually it is the cable in case of a malfunction, so I wanted to shorten it a bit,
but it looked really intact (and is quite new).
Took a magnifying glass and looked at the soldering of ground, data, clock, and power wires on that board.
Was shocked, did not expect that from Logitech, just a blob of solder with the wires
sort of stuck in it, no signs of proper solder flow, looked really bad.
I re-melted the solder, problem gone.
By the color of it it looks like leadless.
Now this keyboard is _not_ under thermal stress _ever_.
It is however constantly hammered on, so mechanical vibration only.

I've been through a LOT of keyboards over the years.
That's what they do, they wear out, break, get stuff spilled on them,
and sometimes they just fail for their own little chiclet reasons.

 

[Nico Coesel]

I had some strange messages in the log file for a few days (Linux):
Aug 8 10:47:10 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.
Aug 8 10:47:11 grml kernel: atkbd.c: Spurious NAK on isa0060/serio1. Some program might be trying access hardware directly.

So I looked it up with google, makes you think of keyboard sniffers.

Bumping the keyboard on the table made it come and go.
But then finally I _only_ got those messages..... no more keys.

Opened up Logitech keyboard, there is a little PCB with a chip under a blob of
epoxy, with four wires soldered to it.
Usually it is the cable in case of a malfunction, so I wanted to shorten it a bit,
but it looked really intact (and is quite new).
Took a magnifying glass and looked at the soldering of ground, data, clock, and power wires on that board.
Was shocked, did not expect that from Logitech, just a blob of solder with the wires
sort of stuck in it, no signs of proper solder flow, looked really bad.

This is just a matter of a bad soldering practises which made it
through quality control. It has nothing to do with lead free. Probably
a new Chinese kid in the factory.

 

[Frithiof Andreas Jensen]

Was shocked, did not expect that from Logitech, just a blob of solder with
the wires
sort of stuck in it, no signs of proper solder flow, looked really bad.

What do you expect for like USD 10 retail (that's *after* everyone in the
supply chain have made a profit)??

It's Chinese, the "solder" probably made from burning old telephone cables
in an oil drum, the teenager soldering it gets maybe USD 2 a day (when she
gets paid) e.t.c. u.s.w. Globalisaton, I think the current plus-word is ;-)