RE: [CAD_CAM_EDM_DRO] Capacitor selection for power supply design
Posted by
Andy Wander
on 2005-01-26 19:02:36 UTC
Interesting.
I would bet, however, that when you parallel two capacitors you can't just
double the ripple current rating. To do that, you'd have to make sure the
caps were exactly identical, so that the current would be split perfectly
evenly between them.
Given the tolerance specs on capacitors, I wouldn't bet on it.
Andy Wander
Verrex Corporation
-----Original Message-----
From: cnc_4_me [mailto:cnc4me@...]
Sent: Wednesday, January 26, 2005 5:19 PM
To: CAD_CAM_EDM_DRO@yahoogroups.com
Subject: [CAD_CAM_EDM_DRO] Capacitor selection for power supply
design
The reason for multiple capacitors in power supply design is for
longer life of the capacitor. Multiple capacitors decrease the
ripple current in each capacitor thereby increasing its life.
The "ripple current spec" of capacitors is one of the most important
specs for capasitor life after voltage, and is almost unknown by most
electronic engineers....
This spec is so unknown that many electronic catalogs leave it out.
To find this spec do as i did...Pick a mfg and capasitor type from
catalog. Then go to google and search for something like "mallory
CGS ripple current" CGS was capasitor type...In the first few links
you should see...
www.cornell-dubilier.com/mpdf/cgs.pdf
try it with "vishay 36DY ripple current" 36DY was capasitor type...In
the first few links you should see...
www.vishay.com/capacitors/aluminum/volt-101-500/....Scrool down to
the 36DY pdf...
With those 2 pdf's you have most of specs for the powersupply caps
that digikey and mouser have...Print them out and put them in your
powersupply design folder...
Now look at some caps like Mallory 150WVDC
11,000uF cap has 12.1 amp max ripple.
22,000uF cap has 15.6 amp max ripple.
Notice how 2 11,000 uF caps have 2 x 12.1 or 24.2 amps of ripple
handling capacity compared to the 15.6 amps of the 22,000 uF
capacitor...
This extra capacity of ripple current = longer capacitor life by less
internal heating of the capacitor...
Next question...How much ripple current do the capacitors have to be
sized for...They must handle the total current of the power
supply...A 20 amp power supply has 20 amps of ripple...
I did a quick search on google for the following "capacitor ripple
current life" and a zillion links came up.
The following link says...
http://industrial.panasonic.com/ww/news_e/nr200304MC002_e/nr200304MC00
2_e.html
"Allowed Ripple Current
When a pulsating flow (AC component) is added to voltage applied to
the capacitor, the resulting current is referred to as "ripple
current." This ripple current affects the component life of the
capacitor. "Allowed ripple current" is a specification value
indicating how much ripple current may be applied to the capacitor.
The higher this value, the higher the component's reliability may be
said to be."
The following pdf explains how capacitors are one of the main source
of failures in VFD drives...And says
"All the degradation mechanisms are exacerbated by ripple current
heating."
http://www.wempec.wisc.edu/reports/2002/2002_35.pdf
Wally
I would bet, however, that when you parallel two capacitors you can't just
double the ripple current rating. To do that, you'd have to make sure the
caps were exactly identical, so that the current would be split perfectly
evenly between them.
Given the tolerance specs on capacitors, I wouldn't bet on it.
Andy Wander
Verrex Corporation
-----Original Message-----
From: cnc_4_me [mailto:cnc4me@...]
Sent: Wednesday, January 26, 2005 5:19 PM
To: CAD_CAM_EDM_DRO@yahoogroups.com
Subject: [CAD_CAM_EDM_DRO] Capacitor selection for power supply
design
The reason for multiple capacitors in power supply design is for
longer life of the capacitor. Multiple capacitors decrease the
ripple current in each capacitor thereby increasing its life.
The "ripple current spec" of capacitors is one of the most important
specs for capasitor life after voltage, and is almost unknown by most
electronic engineers....
This spec is so unknown that many electronic catalogs leave it out.
To find this spec do as i did...Pick a mfg and capasitor type from
catalog. Then go to google and search for something like "mallory
CGS ripple current" CGS was capasitor type...In the first few links
you should see...
www.cornell-dubilier.com/mpdf/cgs.pdf
try it with "vishay 36DY ripple current" 36DY was capasitor type...In
the first few links you should see...
www.vishay.com/capacitors/aluminum/volt-101-500/....Scrool down to
the 36DY pdf...
With those 2 pdf's you have most of specs for the powersupply caps
that digikey and mouser have...Print them out and put them in your
powersupply design folder...
Now look at some caps like Mallory 150WVDC
11,000uF cap has 12.1 amp max ripple.
22,000uF cap has 15.6 amp max ripple.
Notice how 2 11,000 uF caps have 2 x 12.1 or 24.2 amps of ripple
handling capacity compared to the 15.6 amps of the 22,000 uF
capacitor...
This extra capacity of ripple current = longer capacitor life by less
internal heating of the capacitor...
Next question...How much ripple current do the capacitors have to be
sized for...They must handle the total current of the power
supply...A 20 amp power supply has 20 amps of ripple...
I did a quick search on google for the following "capacitor ripple
current life" and a zillion links came up.
The following link says...
http://industrial.panasonic.com/ww/news_e/nr200304MC002_e/nr200304MC00
2_e.html
"Allowed Ripple Current
When a pulsating flow (AC component) is added to voltage applied to
the capacitor, the resulting current is referred to as "ripple
current." This ripple current affects the component life of the
capacitor. "Allowed ripple current" is a specification value
indicating how much ripple current may be applied to the capacitor.
The higher this value, the higher the component's reliability may be
said to be."
The following pdf explains how capacitors are one of the main source
of failures in VFD drives...And says
"All the degradation mechanisms are exacerbated by ripple current
heating."
http://www.wempec.wisc.edu/reports/2002/2002_35.pdf
Wally
Discussion Thread
cnc_4_me
2005-01-26 14:19:08 UTC
Capacitor selection for power supply design
Andy Wander
2005-01-26 19:02:36 UTC
RE: [CAD_CAM_EDM_DRO] Capacitor selection for power supply design