Author Topic: tantalum (tantalex) solid capacitor surge rating? choosing inrush resistor size?  (Read 738 times)

0 Members and 1 Guest are viewing this topic.

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 9733
  • Country: us
  • $
I have a tantalex capacitor that I am curious about. A M39003.

I read a few articles on failure modes of tantalum capacitors, but I have trouble understanding the data sheet in regards to surge current rating

https://www.mouser.com/datasheet/2/427/150d-1764082.pdf

So assuming you are feeding it from a regulated power supply (5V), and the capacitor I have is 18V rated at 150uF, is there a concern from a low source impedance charging the capacitor on start up?

The capacitor ESR measures 0.18 ohms at 1khz, 0.2 ohms at 120Hz. Assuming a perfect supply, the inrush to a bare capacitor should be ~27A. I don't see a surge rating on the capacitor. Its big also. But I did read some articles that claim that tantalum capacitors have a surge rating, but it seems mostly to be in regards to voltage.

How do you find out what is acceptable for this capacitor to take?  I feel like a infrequent event is not going to heat up such a large capacitor very much.

Quote
"Derating is a means for designers of space systems to further reduce the probability of failures by limiting the level of stresses to capacitors during application. Typical derating requirements for solid tantalum capacitors limit the maximum applied voltage to 50% of the rated voltage (VR) and the inrush currents are bounded by additional resistors used in series with the capacitors. First, in the 1960s, the requirement for resistors was 3 Ω per each volt of operating voltage, but by the 1980s, due to improvements in the reliability of the parts and a strong need to increase the efficiency of power supply systems, this requirement was reduced to 1 Ω, and in the 1990s even to 0.1 Ω per volt [6, 15] or 1 Ω, whichever is greater."

https://escies.org/download/webDocumentFile?id=60981


Do each of these capacitor series just have a empirical number associated with required resistance? Where does a tantalex capacitor stand in regards to these numbers? It seems that the answer is between 0.1-3 ohms per volt, or in this case with a 5V supply, 0.5 to 15 ohms. Thats quite a range.
« Last Edit: August 26, 2022, 03:48:16 pm by coppercone2 »
 

Offline srb1954

  • Super Contributor
  • ***
  • Posts: 1104
  • Country: nz
  • Retired Electronics Design Engineer
I have a tantalex capacitor that I am curious about. A M39003.

I read a few articles on failure modes of tantalum capacitors, but I have trouble understanding the data sheet in regards to surge current rating

https://www.mouser.com/datasheet/2/427/150d-1764082.pdf

So assuming you are feeding it from a regulated power supply (5V), and the capacitor I have is 18V rated at 150uF, is there a concern from a low source impedance charging the capacitor on start up?

The capacitor ESR measures 0.18 ohms at 1khz, 0.2 ohms at 120Hz. Assuming a perfect supply, the inrush to a bare capacitor should be ~27A. I don't see a surge rating on the capacitor. Its big also. But I did read some articles that claim that tantalum capacitors have a surge rating, but it seems mostly to be in regards to voltage.

How do you find out what is acceptable for this capacitor to take?  I feel like a infrequent event is not going to heat up such a large capacitor very much.

Quote
"Derating is a means for designers of space systems to further reduce the probability of failures by limiting the level of stresses to capacitors during application. Typical derating requirements for solid tantalum capacitors limit the maximum applied voltage to 50% of the rated voltage (VR) and the inrush currents are bounded by additional resistors used in series with the capacitors. First, in the 1960s, the requirement for resistors was 3 Ω per each volt of operating voltage, but by the 1980s, due to improvements in the reliability of the parts and a strong need to increase the efficiency of power supply systems, this requirement was reduced to 1 Ω, and in the 1990s even to 0.1 Ω per volt [6, 15] or 1 Ω, whichever is greater."

https://escies.org/download/webDocumentFile?id=60981


Do each of these capacitor series just have a empirical number associated with required resistance? Where does a tantalex capacitor stand in regards to these numbers? It seems that the answer is between 0.1-3 ohms per volt, or in this case with a 5V supply, 0.5 to 15 ohms. Thats quite a range.
If the cap is down stream of a regulator such as a 7805 then you can dispense with any series resistance as the regulator current limiting is usually sufficient to protect the tantalum cap. However, if the tantalum cap is subject to major surges such as when you directly switch or connect a supply on to the circuit or plug a PCB into a powered-up back plane then some series resistance is essential to obtain acceptable reliability from the tantalum cap.

You may need to look into Vishay/Sprague application notes or reliability reports to determine their recommended series resistance for your application.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 9733
  • Country: us
  • $
well the first problem is that the capacitor I have is designated J in the category relevant to surges, that is not in the data sheet. I wanted it for a post regulator on a 10A linear regulator (this might be trouble territory) for a 10A 5V meanwell supply. Maybe I will just buy aluminum polymer.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16778
  • Country: us
  • DavidH
If the cap is down stream of a regulator such as a 7805 then you can dispense with any series resistance as the regulator current limiting is usually sufficient to protect the tantalum cap. However, if the tantalum cap is subject to major surges such as when you directly switch or connect a supply on to the circuit or plug a PCB into a powered-up back plane then some series resistance is essential to obtain acceptable reliability from the tantalum cap.

That is the way I understand it.  Tantalum failures are usually on the input side of the regulator where surge current is uncontrolled.  Note that this only applies to surge current during charging; discharging has no limitation.

Quote
You may need to look into Vishay/Sprague application notes or reliability reports to determine their recommended series resistance for your application.

There are a lot of good application notes available on the subject.  Voltage derating has a huge effect on reliability.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 9733
  • Country: us
  • $
I should test the meanwell PSU I have to see what happens on turn on to a test capacitor with a p60something clamp I think
 

Offline srb1954

  • Super Contributor
  • ***
  • Posts: 1104
  • Country: nz
  • Retired Electronics Design Engineer
I should test the meanwell PSU I have to see what happens on turn on to a test capacitor with a p60something clamp I think
Or you could just measure the maximum slew rate dV/dt of the Meanwell O/P voltage at switch-on and calculate the capacitor current from that.

You are probably fairly safe putting tantalums on the output of a switching PSU as they tend to turn on relatively slowly at power-up: the O/P voltage rise-time is slowed by the charging of its own O/P capacitors.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16778
  • Country: us
  • DavidH
I should test the meanwell PSU I have to see what happens on turn on to a test capacitor with a p60something clamp I think

The test circuit for surge turn on failures just uses a MOSFET to connect the tantalum capacitor to a low impedance source.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf