 |
|
| Science Forum Index » Optics Forum » Photodiode leakage current sources?... |
|
Page 1 of 1 |
|
| Author |
Message |
| john Norlie... |
 Posted: Sun Sep 27, 2009 5:47 pm |
|
|
|
Guest
|
This is a photodiode manufacturing process issue -
We are customer for a custom photodiode, and recently we have been
receiving PDs with a leakage current that drifts up over time, to the
point where the comparator circuit monitoring it false triggers. The
supplier is unable to sort out the cause - so I'm looking for help.
There is nothing fancy about this PD - room temp operating range,
photovoltaic mode, 12V reverse bias through a 2M resistor feeding a
comparator circuit. This part has been successfully made for us for
many years now. Leakage current (with no light) is usually <1nA.
Failing parts can exceed 100nA, but do so over time (after final
test).
Any suggestions on things to look for in the manufacturing process
that can cause excessive leakage currents?
Any suggestions for a means to test parts to sort out good from bad
(where "bad" is a part that looks good initially, but begins to
increase leakage current over time)?
- John |
|
|
| Back to top |
|
|
|
| Helmut Wabnig... |
| Posted: Mon Sep 28, 2009 12:09 am |
|
|
|
Guest
|
On Sun, 27 Sep 2009 20:47:54 -0700 (PDT), john Norlie
<jnorlie at (no spam) gmail.com> wrote:
[quote:3e2adfbe21]This is a photodiode manufacturing process issue -
We are customer for a custom photodiode, and recently we have been
receiving PDs with a leakage current that drifts up over time, to the
point where the comparator circuit monitoring it false triggers. The
supplier is unable to sort out the cause - so I'm looking for help.
There is nothing fancy about this PD - room temp operating range,
photovoltaic mode, 12V reverse bias through a 2M resistor feeding a
comparator circuit. This part has been successfully made for us for
many years now. Leakage current (with no light) is usually <1nA.
Failing parts can exceed 100nA, but do so over time (after final
test).
Any suggestions on things to look for in the manufacturing process
that can cause excessive leakage currents?
Any suggestions for a means to test parts to sort out good from bad
(where "bad" is a part that looks good initially, but begins to
increase leakage current over time)?
- John
search for another manufacturer,[/quote:3e2adfbe21]
if they don't know what they are doing.
Things can go wrong in many ways.
One company used to make transistors with metal case.
Suddenly those transistors would start to degrade in the field.
......long story.....
In the laboratory where the transistor cases had to be cleaned
before mounting, they used glass beakers for mixing the washing
fluids.
The new cleaning lady used to clean the beakers with a kitchen
dish washer. Dish washers leave a film on the glass to make it
"shiny".
w. |
|
|
| Back to top |
|
|
|
| Phil Hobbs... |
| Posted: Mon Sep 28, 2009 8:12 am |
|
|
|
Guest
|
john Norlie wrote:
[quote:7bbe82caa8]This is a photodiode manufacturing process issue -
We are customer for a custom photodiode, and recently we have been
receiving PDs with a leakage current that drifts up over time, to the
point where the comparator circuit monitoring it false triggers. The
supplier is unable to sort out the cause - so I'm looking for help.
There is nothing fancy about this PD - room temp operating range,
photovoltaic mode, 12V reverse bias through a 2M resistor feeding a
comparator circuit. This part has been successfully made for us for
many years now. Leakage current (with no light) is usually <1nA.
Failing parts can exceed 100nA, but do so over time (after final
test).
Any suggestions on things to look for in the manufacturing process
that can cause excessive leakage currents?
Any suggestions for a means to test parts to sort out good from bad
(where "bad" is a part that looks good initially, but begins to
increase leakage current over time)?
- John
[/quote:7bbe82caa8]
Check the passivation and the pad metal. Did the fab change over to a
ROHS-compliant process, or change the passivation material?
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net |
|
|
| Back to top |
|
|
|
| john Norlie... |
| Posted: Thu Oct 08, 2009 5:33 pm |
|
|
|
Guest
|
On Sep 28, 8:41 pm, Russell Shaw <rjshawN_o at (no spam) s_pam.netspace.net.au>
wrote:
[quote:b07c55f365]john Norlie wrote:
On Sep 28, 1:26 pm, Phil Hobbs
pcdhSpamMeSensel... at (no spam) electrooptical.net> wrote:
John Devereux wrote:
john Norlie <jnor... at (no spam) gmail.com> writes:
On Sep 28, 9:12 am, Phil Hobbs
pcdhSpamMeSensel... at (no spam) electrooptical.net> wrote:
john Norlie wrote:
This is aphotodiodemanufacturing process issue -
We are customer for a customphotodiode, and recently we have been
receiving PDs with aleakagecurrentthat drifts up over time, to the
point where the comparator circuit monitoring it false triggers. The
supplier is unable to sort out the cause - so I'm looking for help..
There is nothing fancy about this PD - room temp operating range,
photovoltaic mode, 12V reverse bias through a 2M resistor feeding a
comparator circuit. This part has been successfully made for us for
many years now. Leakagecurrent(with no light) is usually <1nA.
Failing parts can exceed 100nA, but do so over time (after final
test).
Any suggestions on things to look for in the manufacturing process
that can cause excessiveleakagecurrents?
Any suggestions for a means to test parts to sort out good from bad
(where "bad" is a part that looks good initially, but begins to
increaseleakagecurrentover time)?
- John
Check the passivation and the pad metal. Did the fab change over to a
ROHS-compliant process, or change the passivation material?
Cheers
Phil Hobbs
Unfortunately the details of the die manufacturing process are
proprietary. We can only specify the performance of the end
component. Even if we change suppliers, we will have the same problem
(no access to a proprietary process).
We would like to have additional means to qualify these components -
even if we have to sample some for a detailed material analysis from
time to time.
No idea really... But perhaps you could use high temperature and/or high
voltage to promote theleakageon units that have not yet "failed"? Use
some new parts from the suspect batch and a few units from the original
"good" batch, see if there is an observable difference.
I'd try boiling a few in water for a half hour.
Cheers
Phil Hobbs
Is this intended to be a cleansing process for surface contaminated
devices? If so, then I would take some known-bad actors, boil 'em up
and look for an improvement inleakagecurrent.
We have tried baking them - at 125c for 60 minutes. The idea being
that if they had water trapped under the delamination, this would
drive it out. This did improve things for the failing assemblies.
But now some are failing again - so this was not a permanent cure.
Do you think I need to do anything after the hard boiling step (e.g.
alcohol rinse and/or high temp bake) to drive off residual water?
After baking a dozen, put half in a bag with dessicant and wait.
If the ones in the bag don't drift, then it's a moisture absorbtion
problem.
[/quote:b07c55f365]
The drift takes too long at room temp/RH (months). So we placed a
dozen in a high temp high RH atmosphere (90c/90%) with power applied
to the electronics (+12V bias applied to the photodiode). After cool
down to room temp, all had high leakage currents. We then baked them
again at 125 at low RH to see if the the absorbed moisture cold be
driven off again. There was no change in the (high) leakage current -
whatever reaction occurred on the surface in the presence of RH +
voltage + contaminants had become a permanent leakage path. We had
two "control" devices that also went through this recipe - no change
in performance. Clearly there is a contamination of some sort with
this lot.
We will next try to seal these devices against the effects of RH - any
suggestions for an overcoating material that will not interfere with
the electrical & optical performance?
Any suggestions for a means to analyze the surface contaminant
composition?
Thanks, ]
John. |
|
|
| Back to top |
|
|
|
| Glen Walpert... |
| Posted: Fri Oct 09, 2009 9:14 am |
|
|
|
Guest
|
On Thu, 8 Oct 2009 20:33:27 -0700 (PDT), john Norlie
<jnorlie at (no spam) gmail.com> wrote:
<clip>
[quote:5ab62456d8]
We will next try to seal these devices against the effects of RH - any
suggestions for an overcoating material that will not interfere with
the electrical & optical performance?
[/quote:5ab62456d8]
Parylene C might work, although you would need to send your parts or
assemblies to a coater (or spend a lot of money on equipment):
http://engineering.tufts.edu/microfab/index_files/docs/Parylene.pdf
search on parylene conformal coating for suppliers & more info. |
|
|
| Back to top |
|
|
|
| Joseph Gwinn... |
| Posted: Sun Oct 11, 2009 12:49 pm |
|
|
|
Guest
|
In article
<a8163ead-b895-40f2-b233-c159e4248c05 at (no spam) p9g2000vbl.googlegroups.com>,
john Norlie <jnorlie at (no spam) gmail.com> wrote:
[quote:2fa7247817]On Sep 28, 8:41 pm, Russell Shaw <rjshawN_o at (no spam) s_pam.netspace.net.au
wrote:
john Norlie wrote:
On Sep 28, 1:26 pm, Phil Hobbs
pcdhSpamMeSensel... at (no spam) electrooptical.net> wrote:
John Devereux wrote:
john Norlie <jnor... at (no spam) gmail.com> writes:
On Sep 28, 9:12 am, Phil Hobbs
pcdhSpamMeSensel... at (no spam) electrooptical.net> wrote:
john Norlie wrote:
This is aphotodiodemanufacturing process issue -
We are customer for a customphotodiode, and recently we have been
receiving PDs with aleakagecurrentthat drifts up over time, to the
point where the comparator circuit monitoring it false triggers. The
supplier is unable to sort out the cause - so I'm looking for help.
There is nothing fancy about this PD - room temp operating range,
photovoltaic mode, 12V reverse bias through a 2M resistor feeding a
comparator circuit. This part has been successfully made for us for
many years now. Leakagecurrent(with no light) is usually <1nA.
Failing parts can exceed 100nA, but do so over time (after final
test).
Any suggestions on things to look for in the manufacturing process
that can cause excessiveleakagecurrents?
Any suggestions for a means to test parts to sort out good from bad
(where "bad" is a part that looks good initially, but begins to
increaseleakagecurrentover time)?
- John
Check the passivation and the pad metal. Did the fab change over to a
ROHS-compliant process, or change the passivation material?
Cheers
Phil Hobbs
Unfortunately the details of the die manufacturing process are
proprietary. We can only specify the performance of the end
component. Even if we change suppliers, we will have the same problem
(no access to a proprietary process).
We would like to have additional means to qualify these components -
even if we have to sample some for a detailed material analysis from
time to time.
No idea really... But perhaps you could use high temperature and/or high
voltage to promote theleakageon units that have not yet "failed"? Use
some new parts from the suspect batch and a few units from the original
"good" batch, see if there is an observable difference.
I'd try boiling a few in water for a half hour.
Cheers
Phil Hobbs
Is this intended to be a cleansing process for surface contaminated
devices? If so, then I would take some known-bad actors, boil 'em up
and look for an improvement inleakagecurrent.
We have tried baking them - at 125c for 60 minutes. The idea being
that if they had water trapped under the delamination, this would
drive it out. This did improve things for the failing assemblies.
But now some are failing again - so this was not a permanent cure.
Do you think I need to do anything after the hard boiling step (e.g.
alcohol rinse and/or high temp bake) to drive off residual water?
After baking a dozen, put half in a bag with dessicant and wait.
If the ones in the bag don't drift, then it's a moisture absorbtion
problem.
The drift takes too long at room temp/RH (months). So we placed a
dozen in a high temp high RH atmosphere (90c/90%) with power applied
to the electronics (+12V bias applied to the photodiode). After cool
down to room temp, all had high leakage currents. We then baked them
again at 125 at low RH to see if the the absorbed moisture cold be
driven off again. There was no change in the (high) leakage current -
whatever reaction occurred on the surface in the presence of RH +
voltage + contaminants had become a permanent leakage path. We had
two "control" devices that also went through this recipe - no change
in performance. Clearly there is a contamination of some sort with
this lot.
We will next try to seal these devices against the effects of RH - any
suggestions for an overcoating material that will not interfere with
the electrical & optical performance?
[/quote:2fa7247817]
I'm no expert here, but it sounds like sodium contamination, as from
human sweat.
[quote:2fa7247817]Any suggestions for a means to analyze the surface contaminant
composition?
[/quote:2fa7247817]
Some kinds of scanning electron microscope can analyze the chemical
composition of what is being looked at, by analyzing the scattered
electrons. This is one way contaminants are tracked in semiconductor
production.
One would expect the photodiode maker to have this equipment.
Joe Gwinn |
|
|
| Back to top |
|
|
|
| Bob May... |
| Posted: Sun Oct 11, 2009 3:13 pm |
|
|
|
Guest
|
You may also want to clean the sockets and so forth or, at the least,
evaluatee them for contamination. It takes a lot of cleaning work to get
good results at the pA level.
--
Bob May
rmay at nethere.com
http: slash /nav.to slash bobmay
http: slash /bobmay dot astronomy.net |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Sat Nov 28, 2009 9:24 pm
|
|