SMD transistor falls off ? have we been scammed ?

[[email protected]]

Hi;

I just can't explain it. Sony XBR 1997 model. Symptom whited out
negative image. Under our warranty and no print. I'm glad it's not the
busy season.

Set uses a TA1226 IC for video enhancement. Good video in, bad video
out. Lucky I had a used one because it wasn't the problem, but it came
from a working set that we own for future sale. No change in symptom at
all. I literally put the sets side by side on the same input and
started flying by the seat of my pants. Armed only with the Tektronix
and a pinout of the IC I found it.

The tippy tops of the sandcastle input to pin 10 were missing.
Following traces etc. I got a bit of a mental map of it. Three diodes
going to that pin etc... On the working set the sync tips for the SCP
are derived from Q1353 IIRC. Turning to the device under test to check
on the status of Q1353 I discovered it missing.

This set was delivered not too long ago and I'm told they have no
reason to suspect sabotage, other than what I had just told them.

Has anyone ever seen a SMD transistor pop off the board without
"assistance" ? If we've been scammed it was by someone either
incredibly lucky or who knows what they're doing. This particular part
of the video circuit seems common only to XBRs. No debris or damage to
the foil and the solder was still there. Damn I wish a defective SMD
transistor would come out so easily. Are we wasting our time
desoldering these things ?

Now playing devil's advocate here, that transistor mounts with the
collector toward the front of the set. In that position flexing of the
PC in the direction caused by up and down movement would apply
compressional/tensional force on those short leads.

I find it hard to believe that all three connections were fine and not
intermittent and just gave up all at the same time. Any one of them
open and the whole device is effectively open. Conversely could the set
have had a highly intermittent problem that we (our shop) has never
seen, and then simply disassembling the set caused the last little bit
holding it to break away ? I'm not really even ready to take a guess on
that right now. Any takers ?

How about the possibility of the customer adamantly wanting the set
replaced having someone else sabotage the set. It wouldn't be the first
time. Back when I was in business an employee of mine did that for
someone and I fired him on the spot. That's not to say there aren't any
people out there who can and will do it.

I need a bit of input, do they sometimes just fall off ? The set is
getting delivered tomorrow, I don't work Saturdays now that I've "paid
my dues", but I might like to call the boss tomorrow and tell him to
warn the customer. Being very careful not to accuse, it should be
explained that if something happens again that looks like sabotage we
will be reluctant to cover it under warranty. Maybe he didn't do it,
maybe someone else did it to him. I'm told the PIP is almost always on
for a security camera, and this could be a factor here.

I'm kinda looking for a figure here where we can tell the customer "my
guy says it's a XX% chance of sabotage. Should I say 90, or more like
75 ? IMO right now it's over 50. I have never seen an SMD just pop off,
even with a bad connection a two terminal device doesn't fall off. This
is a three terminal device, and I think that makes it pretty unlikely.
The problem is you have to tread lightly here when what you say can be
construed as an accusation of fraud.

A penny for your thoughts. Hell we're giving them away, I just found
out I'm not making quite a dollar a day. Today's dollar has about a
nickel's worth of purchasing power, $200 in an eight hour day comes out
to $25 an hour, I fall just a bit short. Perhaps I'm underpaid ? (I
mean more underpaid than others) Even if I am I don't want to see the
company get screwed.

Thanks in advance, for ANYTHING that can shed some light on this.

JURB

 

[NSM]

Has anyone ever seen a SMD transistor pop off the board without
"assistance"?

No, but it wouldn't amaze me. Probably a fault in the reflow soldering
process.

 

[Sam Goldwasser]

Has anyone ever seen a SMD transistor pop off the board without
"assistance"?

No, but it wouldn't amaze me. Probably a fault in the reflow soldering
process.[/QUOTE]

You should be able to closely examine the solder pads to determine if
it was actually unsoldered, or poorly soldered in the first place. That
is, of course, if you haven't already replaced the part!

--- sam | Sci.Electronics.Repair FAQ Mirror: http://repairfaq.ece.drexel.edu/
Repair | Main Table of Contents: http://repairfaq.ece.drexel.edu/REPAIR/
+Lasers | Sam's Laser FAQ: http://repairfaq.ece.drexel.edu/sam/lasersam.htm
| Mirror Sites: http://repairfaq.ece.drexel.edu/REPAIR/F_mirror.html

Note: These links are hopefully temporary until we can sort out the excessive
traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is
ignored unless my full name is included in the subject line. Or, you can
contact me via the Feedback Form in the FAQs.

 

[N Cook]

Hi;

I just can't explain it. Sony XBR 1997 model. Symptom whited out
negative image. Under our warranty and no print. I'm glad it's not the
busy season.

Set uses a TA1226 IC for video enhancement. Good video in, bad video
out. Lucky I had a used one because it wasn't the problem, but it came
from a working set that we own for future sale. No change in symptom at
all. I literally put the sets side by side on the same input and
started flying by the seat of my pants. Armed only with the Tektronix
and a pinout of the IC I found it.

The tippy tops of the sandcastle input to pin 10 were missing.
Following traces etc. I got a bit of a mental map of it. Three diodes
going to that pin etc... On the working set the sync tips for the SCP
are derived from Q1353 IIRC. Turning to the device under test to check
on the status of Q1353 I discovered it missing.

This set was delivered not too long ago and I'm told they have no
reason to suspect sabotage, other than what I had just told them.

Has anyone ever seen a SMD transistor pop off the board without
"assistance" ? If we've been scammed it was by someone either
incredibly lucky or who knows what they're doing. This particular part
of the video circuit seems common only to XBRs. No debris or damage to
the foil and the solder was still there. Damn I wish a defective SMD
transistor would come out so easily. Are we wasting our time
desoldering these things ?

Now playing devil's advocate here, that transistor mounts with the
collector toward the front of the set. In that position flexing of the
PC in the direction caused by up and down movement would apply
compressional/tensional force on those short leads.

I find it hard to believe that all three connections were fine and not
intermittent and just gave up all at the same time. Any one of them
open and the whole device is effectively open. Conversely could the set
have had a highly intermittent problem that we (our shop) has never
seen, and then simply disassembling the set caused the last little bit
holding it to break away ? I'm not really even ready to take a guess on
that right now. Any takers ?

How about the possibility of the customer adamantly wanting the set
replaced having someone else sabotage the set. It wouldn't be the first
time. Back when I was in business an employee of mine did that for
someone and I fired him on the spot. That's not to say there aren't any
people out there who can and will do it.

I need a bit of input, do they sometimes just fall off ? The set is
getting delivered tomorrow, I don't work Saturdays now that I've "paid
my dues", but I might like to call the boss tomorrow and tell him to
warn the customer. Being very careful not to accuse, it should be
explained that if something happens again that looks like sabotage we
will be reluctant to cover it under warranty. Maybe he didn't do it,
maybe someone else did it to him. I'm told the PIP is almost always on
for a security camera, and this could be a factor here.

I'm kinda looking for a figure here where we can tell the customer "my
guy says it's a XX% chance of sabotage. Should I say 90, or more like
75 ? IMO right now it's over 50. I have never seen an SMD just pop off,
even with a bad connection a two terminal device doesn't fall off. This
is a three terminal device, and I think that makes it pretty unlikely.
The problem is you have to tread lightly here when what you say can be
construed as an accusation of fraud.

A penny for your thoughts. Hell we're giving them away, I just found
out I'm not making quite a dollar a day. Today's dollar has about a
nickel's worth of purchasing power, $200 in an eight hour day comes out
to $25 an hour, I fall just a bit short. Perhaps I'm underpaid ? (I
mean more underpaid than others) Even if I am I don't want to see the
company get screwed.

Thanks in advance, for ANYTHING that can shed some light on this.

JURB

The only sabotaged for refund I've ever come across was a fuse replaced with
a much higher amperage rating but blown fuse. Deliberately desoldering an
SMD seems highly unlikely.
I would suggest the SMD was only held on by the glue spot and was never
soldered in the first place, if
light loading, then would function until the glue fails due to minimal
flexing from temperature or vibration.

 

[N Cook]

As an aside does anyone know if it possible to have a bubble form
in flow solder operation so failure to solder a small area ?
That is gas from outgassing from the board or just contamination forming a
bubble on the molten solder.

 

[Wild Bill]

I'd suspect that it's just a manufacturing defect. I suppose that there are
many millions (billions?) of components picked-and-placed each year.
When I look at a high-component board like a laptop main board, I'm amazed
that so many different components are present, and that all of them are
securely mounted.. then again, sometimes they aren't.

A close examination of the leads and the site might have revealed a
reasonable answer. The paste solders are generally applied/dispensed by
needles, or maybe even a faster method presently, and of all the potential
things that could go wrong, maybe the component leads might've had some
contamination on them at the time it was manufactured. There would be
numerous other possibilities.
It might've been hand rework on a small batch of boards, done with a hot air
iron.

I think you've probably been participating in the SER group long enough to
remember the thread "why can't manufacturers solder?", or something similar.

Good techniques, applied/done poorly, yield poor results. And then there's
the planned obsolesence and quality control factors.. it's only consumer
grade stuff.
Poor soldering has kept a lot of shops in business through the years.

I was completely surprised to find a desoldered HOT in a working 25" TV
years ago. It was one of several that I bought at an auction, and I was
checking their operation to see if any of 'em were working.
This particular one had apparently been worked on, and the HOT had been
desoldered and removed, then just put back with the leads thru the board
holes.

There have been several times that I've encountered contamination on
component mounting pads that resulted in oxidation and complete separation
of the solder. These appeared to be a result of unclean boards used at the
time of manufacture.

Cheers
WB
..................

 

[Franc Zabkar]

Has anyone ever seen a SMD transistor pop off the board without
"assistance" ?

Was this a power transistor or a signal transistor? It's a long shot,
but could it be that a fault in the transistor caused it to get so hot
that it desoldered itself? I've seen a bridge rectifier do that ...
after its fuse had been uprated.


- Franc Zabkar

 

[[email protected]]

I once worked on a problem on a large mainframe computer core memory
caused by one pin on a printed circuit card not having any solder on
it.

This was a sense amplifier card about 4 inches square that was flow
soldered during manufacture.
I had watched them being made in the plant. The cards were pulled
across the top of a pot of molten solder by conveyer belts on each
side. If the defect had been caused by a bubble the bubble would have
had to remain with the pin all the way across the pot. I doubt that.

The trouble was intermittent as the module pin was barely touching the
side of the plated through hole. It started failing about six months
after the system was installed. This occurred after a thunder storm
knocked out the computer room air conditioning and the room reached 90
degrees F. before the customer finally shut the system down.

I always wondered how that defect occurred. It was in the middle of
the card.

Van

 

[N Cook]

I once worked on a problem on a large mainframe computer core memory
caused by one pin on a printed circuit card not having any solder on
it.

This was a sense amplifier card about 4 inches square that was flow
soldered during manufacture.
I had watched them being made in the plant. The cards were pulled
across the top of a pot of molten solder by conveyer belts on each
side. If the defect had been caused by a bubble the bubble would have
had to remain with the pin all the way across the pot. I doubt that.

The trouble was intermittent as the module pin was barely touching the
side of the plated through hole. It started failing about six months
after the system was installed. This occurred after a thunder storm
knocked out the computer room air conditioning and the room reached 90
degrees F. before the customer finally shut the system down.

I always wondered how that defect occurred. It was in the middle of
the card.

Van

Was it a one off fault with that one board or a batch fault ?

 

[quietguy]

Many years ago I did a stint on a production line at Kriesler Radio - the
PCB was soldered in the same manner you describe, but as it came off the
solder bath it went straight to the dry joint section where all joints
were checked, and lots needed to be resoldered by hand.

Send you blind doing that checking all day, but luckily for me I only
filled in for one of the girls for a couple of days then went back to
componant replacement.

David

 

[[email protected]]

I never found out. I could have fixed the card by soldering the joint
but I replaced it and sent the bad one back to the quality control
people at the plant.

Several years earlier I had trouble shot a problem and when I removed
the card it had the words "Bad Card" written with magic marker on the
printed side. I sent it to the plant also.

Van