Bad chip resistors?

[Phil Allison]

Hi,

recently came across an SMD board with lots of chip resistors (0805) and similar size ceramic caps. There was mild surface corrosion on the ends of each resistor - enough to prevent a DMM from reading the value until cleaned with a "fibre pen". For the puzzled, it's an essential servicing tool.

https://www.ebay.com.au/itm/Fibreglass-Pen-Fibre-Pen-Pencil-Abrasive-Cleaning-With-5-Fibre-Refills-/310094191492

The item itself looked new, but was in fact 9 years old.
Eventually I found 3 open resistors, all with values like 1Mohm.

Googling tells me this is a known hazard not shared with through hole parts - has to do with the use of silver plating on top surface of the resistor being subject to sulphiding and similar air borne contaminants.

Any insights?


.... Phil

[legg]

As formation of sulphides is a surface phenomenon, I doubt
it plays a role in properly reflow-soldered junctions.

Mechanical stress can.

Needle-pointed probes are quicker, though possibly painfull
when left lying around. Detachable ones make sense, if they
don't get lost.

RL

[John Crane]

I spent some time during my career as an electronics tech in the
petrochem industry. Lots of corrosive substances and vapors that will
kill electronics. It was common practice that after a board was
assembled and tested, it was bathed in a clear epoxy coat to completely
seal it.

-John

[jjhu...@gmail.com]

It's called a conformal coating and they are a number of different types depending on the environmental conditions one is trying to safeguard the electronics with.
I think Phil's question is more along the lines of the failure modes of the SMD resistors or interaction with the board material.
I have not heard of the failure mode Phil has described. IIRC, the sulphiding will form an insulation layer, so poking with needle probes is necessary to get good readings. I have not read anything about how the sulphiding changes the resistor characteristics. Sorry.
The only thing I am aware of, which is remotely related to this is the tin 'whisker' phenomenon which shorts out components. Caused by when tin (especially electroplated tin) is used as a final finish.

Could hot air reflowing the areas where the resistors are located help? or have the resistors themselves really undergone the change?
Sorry I could not be of more help
J

[Trevor Wilson]

**If I have a weird fault, I always suspect high value (>100k) or low
value (<100 Ohms) resistors. Both sometimes fail for different reasons.
That said, mostly I find such faults confined to cracked carbon resistors.

--
Trevor Wilson
www.rageaudio.com.au

--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Trevor Wilson]

On 28/06/2020 8:15 pm, Phil Allison wrote:
>
>

**Also: Get yourself a pair of these:

https://www.wagneronline.com.au/4mm-plug-on-test-probe-hck-silicone/test-leads-hck-silicone/test-measure/tools-test/6762/fl/

Pierces insulation (and fingers!) easily.

[jjhu...@gmail.com]

The pointer to the tool is interesting to me. It beats keeping an old style pencil eraser around, which tends to dry out over time...speaking of that....
What is the material in this pen? fiberglass? and does it dry up or change hardness in any way?
Thanks
J

[Fox's Mercantile]

On 6/28/20 3:40 PM, jjhu...@gmail.com wrote:
> The pointer to the tool is interesting to me. It beats keeping an old style pencil eraser around, which tends to dry out over time...speaking of that....
> What is the material in this pen? fiberglass? and does it dry up or change hardness in any way?

I have a couple in my tool box.
The brush is stiff fiberglass, and you can adjust the length.
It works amazingly well.
And no, nothing changes with age.


--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

[Terry Schwartz]

Not unusual for high value resistors to fail open. This was also true for leaded resistors. The same environmental conditions that cause the surface corrosion contribute to the resistor failures.

[Phil Allison]

Terry Schwartz wrote:
====================

>
> Not unusual for high value resistors to fail open.
>

** Well they don't fail short very often - but high rates of such in new looking equipment is very unusual.

> This was also true for leaded resistors.

** Only those with lots of DC across them, not the case here.

> The same environmental conditions that cause the surface corrosion
> contribute to the resistor failures.

** Yeah - I did figure that.

But chip resistors are unusually vulnerable cos of exposed silver.

..... Phil

[Trevor Wilson]

On 29/06/2020 12:04 pm, Phil Allison wrote:
>
> Terry Schwartz wrote:
> ====================
>
>>
>> Not unusual for high value resistors to fail open.
>>
>
> ** Well they don't fail short very often - but high rates of such in new looking equipment is very unusual.

**I've never seen them fail S/C. O/C, yes.

>
>
>> This was also true for leaded resistors.
>
> ** Only those with lots of DC across them, not the case here.

**Nope. Seen lots fail over the years. Even those in preamp sections,
when Voltages are well below 40 Volts.

>
>
>> The same environmental conditions that cause the surface corrosion
>> contribute to the resistor failures.
>
> ** Yeah - I did figure that.
>
> But chip resistors are unusually vulnerable cos of exposed silver.

**Certainly are. I've seen a few fail.

[Phil Allison]

Trevor Wilson wrote:

================

>
> **If I have a weird fault, I always suspect high value (>100k) or low
> value (<100 Ohms) resistors.

** That is why I went on a bad resistor hunt - after everything else proved OK.

> That said, mostly I find such faults confined to cracked carbon resistors.

** Small carbon film resistors regularly fail (open) due to high voltage "tracking" between turns on the spiral cut. Plate resistors in triode stages for one.

IME - metal film and cermet types ( ie chip resistors) are less prone to do the same.


.... Phil

[Trevor Wilson]

On 29/06/2020 12:24 pm, Phil Allison wrote:
> Trevor Wilson wrote:
>
> ================
>
>>
>> **If I have a weird fault, I always suspect high value (>100k) or low
>> value (<100 Ohms) resistors.
>
> ** That is why I went on a bad resistor hunt - after everything else proved OK.

**The most painful ones I've seen were popular in Sansui products a few
decades back. They were typically in the range of 100 ohms ~ 47k and
would go high. Frequently to around double the value printed on them.
Fault finding was very difficult, since Voltages were almost what they
were supposed to be. Almost, but not quite.

>
>
>> That said, mostly I find such faults confined to cracked carbon resistors.
>
>
> ** Small carbon film resistors regularly fail (open) due to high voltage "tracking" between turns on the spiral cut. Plate resistors in triode stages for one.
>
> IME - metal film and cermet types ( ie chip resistors) are less prone to do the same.

**I've seen plenty of >100k resistors go high, despite being subject o
Voltages of less than 40 Volts.

[Phil Allison]

Trevor Wilson wrote:

===================
>
> >
> >>
> >> Not unusual for high value resistors to fail open.
> >>
> >
> > ** Well they don't fail short very often - but high rates of such in new looking equipment is very unusual.
>
> **I've never seen them fail S/C.

** ROTFL....

> >
> >> This was also true for leaded resistors.
> >
> > ** Only those with lots of DC across them, not the case here.
>
> **Nope. Seen lots fail over the years.
>

** Yawnnnn......

TW claims all kinds of impossible things.

>
> > But chip resistors are unusually vulnerable cos of exposed silver.
>
> **Certainly are. I've seen a few fail.
>

** The amp in question is a "DV Mark" combo with 3 x PCBs covered in SMD.

Every chip resistor ( 100+ of them ) is corroded after storage in its original packaging for 6 years or so.

It uses just 2 obsolete Sanken SAP 5-pin darlingtons, has no VI limiting or rail fuses but has an extension speaker jack just waiting to be shorted. Peak available output current is +/- 30 amps double the max Ic of the transistor.

SMPS with the markings ground off the control chip.

Will drive 200W @ 4ohms but has totally inadequate cooling for this condition and no thermal shut down.

The fitted speaker is 16 ohms with a *tin* neo magnet.

No schem is available.

Total PITA.


..... Phil

[John-Del]

I've run into a lot of >500K resistors open in SMPS "start" circuits - where an IC gets its start Vcc from the rectified mains and divided down to 2.5V with half a dozen high value resistors and one low value resistor at ground. The same arrangement was also used to monitor the output and use the divided voltage for feedback. I can't recall seeing a high value smd resistor open with low voltage across them.

Whenever I see any inop SMPS supply with no blown output devices, I look for these first.

Interestingly, I've also found these resistors can be bothered by the adhesive beneath them and read *lower* value than their printed values. When I run into a supply that drifts, I remove these resistors, clean the adhesive off the board beneath them, and reinstall them. They read on the dot after that.

[Phil Allison]

legg wrote:

============
Phil Allison

>
> >
> >The item itself looked new, but was in fact 9 years old.
> >Eventually I found 3 open resistors, all with values like 1Mohm.
> >
> >Googling tells me this is a known hazard not shared with through hole parts - has to do with the use of silver plating on top surface of the resistor being subject to sulphiding and similar air borne contaminants.
> >
>>

> As formation of sulphides is a surface phenomenon, I doubt
> it plays a role in properly reflow-soldered junctions.
>

** Never said it did.

As the silver layer is on TOP of the SMD chip, failure occurs at the silver to resistive element interface.

You would need a microscope to see evidence of it.

My concern is that many more of the same resistors will fail open in time.


..... Phil

[Jeff Urban]

Well the best thing to do is probably keep it cool and don't bend it.

What else ? Change them all ?

[Phil Allison]

Jeff Urban wrote:

=======================

> Well the best thing to do is probably keep it cool and don't bend it.
>

** FFS temp was never and issue nor flexing.

> What else ? Change them all ?

** Soon as pigs can fly...

[legg]

On Mon, 29 Jun 2020 11:59:40 -0700 (PDT), John-Del <ohg...@gmail.com>
wrote:

>I've run into a lot of >500K resistors open in SMPS "start" circuits - where an IC gets its start Vcc from the rectified mains and divided down to 2.5V with half a dozen high value resistors and one low value resistor at ground. The same arrangement was also used to monitor the output and use the divided voltage for feedback. I can't recall seeing a high value smd resistor open with low voltage across them.
>
>Whenever I see any inop SMPS supply with no blown output devices, I look for these first.
>
>Interestingly, I've also found these resistors can be bothered by the adhesive beneath them and read *lower* value than their printed values. When I run into a supply that drifts, I remove these resistors, clean the adhesive off the board beneath them, and reinstall them. They read on the dot after that.

Adhesive is used when SMDs are wave-soldered, or for inverted
or repeated reflow. All those processes require additional
skill and technical knowledge to succeed, with a reasonable
service life.

There are a lot of other factors influencing parasitic
terminal leakage in those cases. You'd have to measure
the part and the board, before and after cleaning, before
and after each stage of rework, to get any idea of the
leakage source.

Reduction of R in higher voltage applications is also associated
with tracking and bridging of the part's element and board
surfaces. All of those were affected by your simple before and
after comparison.

RL

[Jeff Urban]

>** FFS temp was never and issue nor flexing.

Sorry for not reading your mind.

>** Soon as pigs can fly...

They put monkeys up in orbit, why not pigs ?

Anyway so that leaves us with what ? They failed because they were pieces of shit, substandard junk ? What else are we to believe here ?

[Phil Allison]

Jeff Urban wrote:

==============

> >** FFS temp was never and issue nor flexing.
>
> Sorry for not reading your mind.

** Try reading my fucking posts - idiot.

>
> >** Soon as pigs can fly...
>
> They put monkeys up in orbit,

** Are you next?

>
> Anyway so that leaves us with what ?
>

** Exactly what my initial post and the one yesterday said.

[peterw...@gmail.com]

Ladies and Gentleman, Boys and Girls, and Children of all ages: You will now be able to observe what happens when two (2) "pigs" mud-wrestle.


Peter Wieck
Melrose Park, PA

[John-Del]

I unashamedly count myself in the third group..