Maker Pro
Maker Pro

Electronic failure puzzle...failure in near-vacuum?

D

David Harper

Jan 1, 1970
0
I'm engaged in a high altitude project incorporating an automatic
camera. In order to ensure that the potential cameras selected were
capable of operating in a near-vacuum, two were tested in a vacuum
chamber. One was a Kodak Ke60, and the other was a Samsung 80Ti.
Both had leads soldered to the shutter contacts, which were controlled
by solid state relays via a microcontroller. The 80Ti's only other
modification was removal of the large flash cap, as we were not
planning on using a flash (not that a flash at 100k ft would do any
good anyway). The 80Ti operated throughout the test, while the Ke60
failed at around 1 PSI. The standard "burned electronics smell" was
present after opening the chamber.

After opening the case to the Ke60 to determine what cause the Ke60
failure, it looks like there was heating near the flash circuitry.
Specifically, there is some discoloration near the contacts to a
silver, eliptical-shaped capacitor (looks wrapped) with "333KS" on the
side. Is this a pressure sensitive component? (the chamber went from
14.7 PSI to about .05 PSI in about a minute, so it was rapid
depressurization). The large flash cap was still good, so I don't
believe this was the point of failure.

Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

Thanks in advance for any advice!
Dave
 
A

Anthony Fremont

Jan 1, 1970
0
David said:
Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

It may not have failed due to the low pressure at all. It could be that
the part simply relied on conductive cooling by transferring the heat to
the air. You took the air away, viola no place to dissipate the heat
to, next thing you know out comes the magic smoke. That's my guess
anyway.

michael brown
 
S

Sir Charles W. Shults III

Jan 1, 1970
0
That is exactly what I would say- we never realize how much component
cooling relies on air conducting that heat away. You might try heat sink
addition. Replace the component and heat sink it to the case or any other
decent heat dump. It might work exactly as it should then.

Cheers!

Chip Shults
My robotics, space and CGI web page - http://home.cfl.rr.com/aichip
 
A

Active8

Jan 1, 1970
0
achilles03 said:
I'm engaged in a high altitude project incorporating an automatic
camera. In order to ensure that the potential cameras selected were
capable of operating in a near-vacuum, two were tested in a vacuum
chamber. One was a Kodak Ke60, and the other was a Samsung 80Ti.
Both had leads soldered to the shutter contacts, which were controlled
by solid state relays via a microcontroller. The 80Ti's only other
modification was removal of the large flash cap, as we were not
planning on using a flash (not that a flash at 100k ft would do any
good anyway). The 80Ti operated throughout the test, while the Ke60
failed at around 1 PSI. The standard "burned electronics smell" was
present after opening the chamber.

After opening the case to the Ke60 to determine what cause the Ke60
failure, it looks like there was heating near the flash circuitry.
Specifically, there is some discoloration near the contacts to a
silver, eliptical-shaped capacitor (looks wrapped) with "333KS" on the
side. Is this a pressure sensitive component? (the chamber went from
14.7 PSI to about .05 PSI in about a minute, so it was rapid
depressurization). The large flash cap was still good, so I don't
believe this was the point of failure.

Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

Thanks in advance for any advice!
Dave
that heat theory is indeed plausible, but the suggestion to heatsink it
doesn't make *total* sense. if you heat sink it to the "case" as was
suggested, there's still no air to dissipate the heat from the
sink/case. one minute isn't enough time for the case to get hot. think
mission time. if you heat sink it to the chamber (via the case?), the
chamber won't be there at 100k ft. the heat will need to be conducted to
the atmosphere - whatever's left of it. -46 deg C at 100k ft. brrr...
too cold for consumer components.

the obvious is staring you in the face, i think. you removed the
photoflash cap from the 80Ti but not the Ke60. i don't see many of those
caps you described, but from the number, i'm guessing that it's a .033uF
job (33x10e3pF) and though it's not neccessarily the photoflash cap, it
may be part of the photoflash circuitry and therefore not needed. at any
rate, you should probably remove all the photoflash and any other
useless circuitry.

just for grins, i would obtain various caps and throw them in the vacuum
chamber and see what happens, that will tell you if it was the heat or
rapid depressurization. you can test other parts this way as well. you
can then try them in test circuits to see if heating effects are the
culprit for sure, then you can think about heat removal. just because
something didn't smoke this time doesn't mean it won't later. no sense
trashing cameras to find out, either.

and before you go pumping hydrogen into any containers with electronics
to keep the fog off the lens or whatever... hydrogen eats aluminum.
makes it brittle. for some reason i can't remember, nitrogen was not an
option and special caps had to be used.

hope this helps,
mike
 
T

Tom MacIntyre

Jan 1, 1970
0
I'm engaged in a high altitude project incorporating an automatic
camera. In order to ensure that the potential cameras selected were
capable of operating in a near-vacuum, two were tested in a vacuum
chamber. One was a Kodak Ke60, and the other was a Samsung 80Ti.
Both had leads soldered to the shutter contacts, which were controlled
by solid state relays via a microcontroller. The 80Ti's only other
modification was removal of the large flash cap, as we were not
planning on using a flash (not that a flash at 100k ft would do any
good anyway). The 80Ti operated throughout the test, while the Ke60
failed at around 1 PSI. The standard "burned electronics smell" was
present after opening the chamber.

After opening the case to the Ke60 to determine what cause the Ke60
failure, it looks like there was heating near the flash circuitry.
Specifically, there is some discoloration near the contacts to a
silver, eliptical-shaped capacitor (looks wrapped) with "333KS" on the
side. Is this a pressure sensitive component? (the chamber went from
14.7 PSI to about .05 PSI in about a minute, so it was rapid
depressurization). The large flash cap was still good, so I don't
believe this was the point of failure.

Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

It's possible that another Ke60 would've been fine...this type of
unique operating environment may not allow for predictable results
based on such a small sample size.

Tom
 
S

Spehro Pefhany

Jan 1, 1970
0
I'm engaged in a high altitude project incorporating an automatic
camera. In order to ensure that the potential cameras selected were
capable of operating in a near-vacuum, two were tested in a vacuum
chamber. One was a Kodak Ke60, and the other was a Samsung 80Ti.
Both had leads soldered to the shutter contacts, which were controlled
by solid state relays via a microcontroller. The 80Ti's only other
modification was removal of the large flash cap, as we were not
planning on using a flash (not that a flash at 100k ft would do any
good anyway). The 80Ti operated throughout the test, while the Ke60
failed at around 1 PSI. The standard "burned electronics smell" was
present after opening the chamber.

After opening the case to the Ke60 to determine what cause the Ke60
failure, it looks like there was heating near the flash circuitry.
Specifically, there is some discoloration near the contacts to a
silver, eliptical-shaped capacitor (looks wrapped) with "333KS" on the
side. Is this a pressure sensitive component? (the chamber went from
14.7 PSI to about .05 PSI in about a minute, so it was rapid
depressurization). The large flash cap was still good, so I don't
believe this was the point of failure.

Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

The clearances in the high-voltage flash circuitry may well not be
sufficient for operation at reduced pressure. The breakdown voltage
across gaps is roughly proportional to air pressure down to a certain
point.

Removing the capacitor may actually have made things worse. Disable
the high voltage power supply-- probably removing a transformer is the
easiest thing to do.

Best regards,
Spehro Pefhany
 
B

bill brown

Jan 1, 1970
0
Another possibility is you are ionizing the air at a low pressure. I have
some real neat photos of a purple glow surrounding the input transistor of
a Geiger counter as the pressure was reduced. At full vacuum the glow
disappeared but the transistor was already cooked. It took a while to
figure that one out. -bill
 
A

Active8

Jan 1, 1970
0
The clearances in the high-voltage flash circuitry may well not be
sufficient for operation at reduced pressure. The breakdown voltage
across gaps is roughly proportional to air pressure down to a certain
point.

Removing the capacitor may actually have made things worse. Disable
the high voltage power supply-- probably removing a transformer is the
easiest thing to do.

Best regards,
Spehro Pefhany
the capacitor was removed from the unit that did not fail.

mike
 
G

George R. Gonzalez

Jan 1, 1970
0
The component sounds like a 0.033f capacitor, perhaps the capacitor that
resonates the
transformer secondary?

IMHO it's pointless speculating what happened-- how about you just kill the
oscillator entirely.
 
F

Fritz Schlunder

Jan 1, 1970
0
bill brown said:
Another possibility is you are ionizing the air at a low pressure. I have
some real neat photos of a purple glow surrounding the input transistor of
a Geiger counter as the pressure was reduced. At full vacuum the glow
disappeared but the transistor was already cooked. It took a while to
figure that one out. -bill



Yes, this is almost certainly the cause of the problem. The decreased air
pressure caused the high voltage to start ionizing the surrounding air thus
creating new conduction paths. These paths modified the circuit such that
the high voltage inverter used to generate about 330V DC for the flash
malfunctioned. Almost certainly if you test the inverter transistors
(usually one or more, say up to four, TO-92 packaged bipolar junction
transistors) you will find that at least some of them are destroyed.

330V doesn't seem like that much and shouldn't arc very far. Indeed this is
typically true. The problem occurs when transformer reset occurs. In some
designs the little transformer can sometimes produce reset voltages that
exceed 1kV. Normally the output rectifier is around a 1.5kV fast recovery
silicon type diode to handle this high voltage. This is almost certainly
plenty of voltage to cause arcs in a significantly low pressure atmosphere.

The 333KS device is a 33 nano farad probably 400V rated 10% tolerance film
capacitor. The capacitor is used in the trigger circuitry. The capacitor
is charged by the main flash voltage stored in the main photoflash capacitor
(around 330V usually) through a resistor of something in the order of 4.7Meg
ohm (although if you trace the circuit out you may find it somewhat
reversed, the 333KS gets charged rapidly from the main photoflash capacitor
when some mechanical contacts close or an SCR triggers, but it gets charged
through the trigger transformer, and then the resistor discharges the 333KS
capacitor in between flashes). In order to initially ionize the xenon the
333KS device discharges into a small usually 4kV trigger autotransformer
which produces the high voltage needed. Many cameras will use a 22nF 400V
film capacitor that looks otherwise almost identical instead of the 33nF
device your camera uses.

Since you don't need the xenon flash, you really aught to remove pretty much
all of the circuitry used for it, not just the big photoflash capacitor.
The circuitry is primarily composed of the following: xenon strobe tube,
xenon trigger transformer, 330V step up inverter transformer (usually a
miniature EE ferrite core device), primary side inverter control BJT
transistors and overvoltage control diacs or similar (usually say 1-4 of
them in TO92 package or maybe SOT23), photoflash capacitor, the trigger
transformer capacitor (the 333KS film capacitor), a high voltage diode
(usually DO41 around 1.5kV fast recovery), a few resistors, and perhaps a
few other items like a flash ready indicator (an LED or neon normally). In
lieu of all that, you could probably get by just by cutting the low voltage
trace feeding the power to the high voltage inverter circuit, although I
should think in your application lower weight would be a definite advantage.
 
S

Spehro Pefhany

Jan 1, 1970
0
the capacitor was removed from the unit that did not fail.
mike

Okay, so that's irrelevant, but the breakdown voltage issue remains.

Best regards,
Spehro Pefhany
 
A

Active8

Jan 1, 1970
0
grg2.Elide- said:
The component sounds like a 0.033f capacitor, perhaps the capacitor that
resonates the
transformer secondary?

IMHO it's pointless speculating what happened-- how about you just kill the
oscillator entirely.
i say its a .033uF cap ...

1st significant digit
2nd significant digit
multiplier

333K

i haven't seen any silver oblong caps that weren't in uF

mike
 
G

George R. Gonzalez

Jan 1, 1970
0
i say its a .033uF cap ...

1st significant digit
2nd significant digit
multiplier

333K

i haven't seen any silver oblong caps that weren't in uF

mike

Yes, that was a typo. It's 33,000pf.

But back to your problem, if you don't need the high-voltage, you're just
asking for trouble having a HV oscillator running, as things spark over much
easier at low pressures. Not to mention the RFI issues of having 300V p-p
waving around.

How about you clip out the transformer and/or the oscillator transistor?
 
Y

YD

Jan 1, 1970
0
I'm engaged in a high altitude project incorporating an automatic
camera. In order to ensure that the potential cameras selected were
capable of operating in a near-vacuum, two were tested in a vacuum
chamber. One was a Kodak Ke60, and the other was a Samsung 80Ti.
Both had leads soldered to the shutter contacts, which were controlled
by solid state relays via a microcontroller. The 80Ti's only other
modification was removal of the large flash cap, as we were not
planning on using a flash (not that a flash at 100k ft would do any
good anyway). The 80Ti operated throughout the test, while the Ke60
failed at around 1 PSI. The standard "burned electronics smell" was
present after opening the chamber.

After opening the case to the Ke60 to determine what cause the Ke60
failure, it looks like there was heating near the flash circuitry.
Specifically, there is some discoloration near the contacts to a
silver, eliptical-shaped capacitor (looks wrapped) with "333KS" on the
side. Is this a pressure sensitive component? (the chamber went from
14.7 PSI to about .05 PSI in about a minute, so it was rapid
depressurization). The large flash cap was still good, so I don't
believe this was the point of failure.

Could it have been this 333KS component? If not, what potential
components may have failed due to decreased ambient pressure?

Thanks in advance for any advice!
Dave

It's possible that expanding air inclusions in the 33nF cap deformed
it enough to short it. Unless some part of the board is essential to
operating the camera remove the whole board, saves a few grams of
weight. At least make sure the HV section is disabled.

- YD.
 
A

Active8

Jan 1, 1970
0
Yes, that was a typo. It's 33,000pf.

But back to your problem, if you don't need the high-voltage, you're just
asking for trouble having a HV oscillator running, as things spark over much
easier at low pressures. Not to mention the RFI issues of having 300V p-p
waving around.

How about you clip out the transformer and/or the oscillator transistor?
prob the best idea. i suggested wrecking out *all* the unneccessary
crap. those 2 (trans) snipped would lower the power consumption, also.

mike
 
D

David Harper

Jan 1, 1970
0
Anthony Fremont said:
It may not have failed due to the low pressure at all. It could be that
the part simply relied on conductive cooling by transferring the heat to
the air. You took the air away, viola no place to dissipate the heat
to, next thing you know out comes the magic smoke. That's my guess
anyway.

michael brown

I doubt it was the lack of convection. The amount of heat generated
by any of the components in the failed area did not produce any
significant amount of heat. In addition, based on the fact that it
was only the 3rd shot (with 20 seconds of off time in between, to
allow for radiative and conductive heat transfer from warm
components), I doubt it had enough time to induce high temperatures in
any of the components, despite their small thermal capacity. In
addition, the other camera worked. Many other cameras have worked in
this type of enviroment, therefore I'm of the opinion it was something
in the actual Ke60 design itself that was sensitive to low air
pressures, leading to the failure.

Based on another post, it seems the likely culprit maybe the fact that
lower pressure air ionizes much easier, making arcing between close
components a possibility, while at standard atmospheric pressure, that
is not a problem.

Thanks for the suggestion!
Dave
 
Top