Re: [time-nuts] Low noise power supplies - dont' use Electrolytics
Electrolytic caps have an extremely poor lifetime (MTBF). Sanyo on their
website state "50K Hrs at 50C".
This means only 6250 MTBF hours at 80C for one single cap. MTBF gets worse
the more caps are being used of course. I have seen some Panasonic
electrolytics state only 2000 hours MTBF at temperature in their datasheets.
That's not even one year before a mean failure occurs making these useless
in high-reliability applications.
Note also that caps in high AC current situations (Buck DC-DC switcher
input cap for example) will self heat due to internal resistance, making things
even worse. This is probably one of the main failure reasons for PC
motherboards.
And that's with name brand parts, it's even worse if one ends up buying
counter-fit or non-name-brand Electrolytics.
Some of our competitors use Electrolytics all over the place (we don't use
any electrolytics) - that's been good for our business.
bye,
Said
In a message dated 11/24/2011 17:08:42 Pacific Standard Time,
lists@lazygranch.com writes:
I'm not familiar with rubycon caps. The low ESR large value caps are
"organic semiconductor." OSCON is a common brand from Sanyo. Finding the
ultimate cap is nearly as much fun as finding the ultimate LDO. Check
out Nichicon. Or you can stick with the Rubycon. Glancing at their
website, they seem to copy the Nichicon product line.
Many of us have seen electronic equipment last longer then one year.
Some of use even have still working antiques with old eletro caps.
Those short lifetimes assume a worse case, usually with a very high
ripple current. IOf you can reduce the ripple the MTBF goes up.
One question: How does one avoid using electrolytic caps if you need
(say) 1,000uF or even 100uF. Those would be some mighty big film
caps.
On Fri, Nov 25, 2011 at 10:06 PM, SAIDJACK@aol.com wrote:
Electrolytic caps have an extremely poor lifetime (MTBF). Sanyo on their
website state "50K Hrs at 50C".
This means only 6250 MTBF hours at 80C for one single cap. MTBF gets worse
the more caps are being used of course. I have seen some Panasonic
electrolytics state only 2000 hours MTBF at temperature in their datasheets.
That's not even one year before a mean failure occurs making these useless
in high-reliability applications.
Note also that caps in high AC current situations (Buck DC-DC switcher
input cap for example) will self heat due to internal resistance, making things
even worse. This is probably one of the main failure reasons for PC
motherboards.
And that's with name brand parts, it's even worse if one ends up buying
counter-fit or non-name-brand Electrolytics.
Some of our competitors use Electrolytics all over the place (we don't use
any electrolytics) - that's been good for our business.
bye,
Said
In a message dated 11/24/2011 17:08:42 Pacific Standard Time,
lists@lazygranch.com writes:
I'm not familiar with rubycon caps. The low ESR large value caps are
"organic semiconductor." OSCON is a common brand from Sanyo. Finding the
ultimate cap is nearly as much fun as finding the ultimate LDO. Check
out Nichicon. Or you can stick with the Rubycon. Glancing at their
website, they seem to copy the Nichicon product line.
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
--
Chris Albertson
Redondo Beach, California
To second the older electronics:
I maintain nearly 100 analytical instruments. The old designs(1970-late
80's) are almost all electrolytic caps and none of the caps have ever
failed. When I do find a bad cap it's always in a modern design. A high
frequency switcher with under rated caps. When i say under rated i mean
take the peak figure and multiple it by two, that's the real part value.
My opinion, bad designs.
Steve
On Sat, Nov 26, 2011 at 1:25 AM, Chris Albertson
albertson.chris@gmail.comwrote:
Many of us have seen electronic equipment last longer then one year.
Some of use even have still working antiques with old eletro caps.
Those short lifetimes assume a worse case, usually with a very high
ripple current. IOf you can reduce the ripple the MTBF goes up.
One question: How does one avoid using electrolytic caps if you need
(say) 1,000uF or even 100uF. Those would be some mighty big film
caps.
On Fri, Nov 25, 2011 at 10:06 PM, SAIDJACK@aol.com wrote:
Electrolytic caps have an extremely poor lifetime (MTBF). Sanyo on their
website state "50K Hrs at 50C".
This means only 6250 MTBF hours at 80C for one single cap. MTBF gets
worse
the more caps are being used of course. I have seen some Panasonic
electrolytics state only 2000 hours MTBF at temperature in their
datasheets.
That's not even one year before a mean failure occurs making these
useless
in high-reliability applications.
Note also that caps in high AC current situations (Buck DC-DC switcher
input cap for example) will self heat due to internal resistance, making
things
even worse. This is probably one of the main failure reasons for PC
motherboards.
And that's with name brand parts, it's even worse if one ends up buying
counter-fit or non-name-brand Electrolytics.
Some of our competitors use Electrolytics all over the place (we don't
use
any electrolytics) - that's been good for our business.
bye,
Said
In a message dated 11/24/2011 17:08:42 Pacific Standard Time,
lists@lazygranch.com writes:
I'm not familiar with rubycon caps. The low ESR large value caps are
"organic semiconductor." OSCON is a common brand from Sanyo. Finding the
ultimate cap is nearly as much fun as finding the ultimate LDO. Check
out Nichicon. Or you can stick with the Rubycon. Glancing at their
website, they seem to copy the Nichicon product line.
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
and follow the instructions there.
--
Chris Albertson
Redondo Beach, California
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
At sane temperatures, OSCONs are very good. Who runs their gear hot
enough to boil water?
In message CABbxVHvNSAxBx5uOrjTQeSTTd50yEDyAVkGXh52bPsPUAYtyWw@mail.gmail.com
, Chris Albertson writes:
One question: How does one avoid using electrolytic caps if you need
(say) 1,000uF or even 100uF. Those would be some mighty big film caps.
One detail, often overlooked, is that electrolytics seldom are
dimensioned very precisely, mostly because they do come with such
big capacity but also because them have very big tolerances, +/-
50% is not uncommon.
Another effect that causes overdimensioning is that they are not very
good capacitors, in particular at higher frequency.
I have in a couple of instances replaced electrolytics with film-caps
and gotten away with less than 1% of the original capacitance by doing
a bit of calculations and measurements on the actual circuit.
In one case, an audio circuit had 1000uF for a handful of opamps,
using 4.7uF of good film capacitors instead reduced THD by 80%.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
On Sat, Nov 26, 2011 at 2:13 AM, gary lists@lazygranch.com wrote:
At sane temperatures, OSCONs are very good. Who runs their gear hot enough
to boil water?
National Fire Protection Agency (NFPA) 2007 edition of their design
regulations state the electronics worn by Fire Fighters must work at
500'F for five minutes.
In the Paper Pusher's mind the Kevlar clothing can withstand those
temperatures, therefor anything else in the Universe can too.
Never under estimate the bureaucratic mind...
I molt them in a high-power charge-pump. The same with WIMA MKS.
At normal usage they will last forever and work even at low temperature
whereas normal Al caps won't.
- Henry
gary schrieb:
At sane temperatures, OSCONs are very good. Who runs their gear hot
enough to boil water?
--
ehydra.dyndns.info
On Fri, 25 Nov 2011 22:25:37 -0800
Chris Albertson albertson.chris@gmail.com wrote:
One question: How does one avoid using electrolytic caps if you need
(say) 1,000uF or even 100uF. Those would be some mighty big film
caps.
I think, aluminium electrolytics are meant, as these have a lot of
"wear" and can die.
Hence, you can use tantal or niob electrolytics, which have a dry
eletrolyt.
Or you can use ceramics which are already available at 100uF.
Attila Kinali
--
Why does it take years to find the answers to
the questions one should have asked long ago?
Ceramic caps can be microphonic. Just something to be on the look out.
Not so much with leaded ceramics, but more of a problem with surface mounts.
Tantalums are prone to overvoltage failure. Best to really overspec them
regarding voltage.
Note that some LDOs are not stable with really low ESR bypass. There are
app notes on this. I've stated my preference for p-fet pass devices, or
better yet, shunt regulators. The PNP pass devices will need feedforward
compensation if the ESR is too low.
On 11/27/2011 11:21 PM, Attila Kinali wrote:
On Fri, 25 Nov 2011 22:25:37 -0800
Chris Albertsonalbertson.chris@gmail.com wrote:
One question: How does one avoid using electrolytic caps if you need
(say) 1,000uF or even 100uF. Those would be some mighty big film
caps.
I think, aluminium electrolytics are meant, as these have a lot of
"wear" and can die.
Hence, you can use tantal or niob electrolytics, which have a dry
eletrolyt.
Or you can use ceramics which are already available at 100uF.
Attila Kinali
On Mon, 28 Nov 2011 01:05:22 -0800
gary lists@lazygranch.com wrote:
Ceramic caps can be microphonic. Just something to be on the look out.
Not so much with leaded ceramics, but more of a problem with surface mounts.
Ceramic caps have a lot of other problems too, like capacitance
derating on increasing voltage. Some types go down to 80% of rated
capacity when driven to the max rated voltage.
Tantalums are prone to overvoltage failure. Best to really overspec them
regarding voltage.
Uhmm.. overspec? The rated voltage is the maximum allowed voltage.
Like something you should never cross. That's why we have these
odd numbers like 16V. They are not intended to be used at 16V but
rather at standard voltages like 12V.
Note that some LDOs are not stable with really low ESR bypass. There are
app notes on this.
Actually, the datasheet of any LDO (or power regulator) should define
the stable operation output capacitance and ESR. If you work outside
the speced range.. well.. your own fault if you get oscillation ;-)
Attila Kinali
--
The trouble with you, Shev, is you don't say anything until you've saved
up a whole truckload of damned heavy brick arguments and then you dump
them all out and never look at the bleeding body mangled beneath the heap
-- Tirin, The Dispossessed, U. Le Guin