Author Topic: Discharge current for capacitors  (Read 2045 times)

0 Members and 1 Guest are viewing this topic.

Offline aheidTopic starter

  • Regular Contributor
  • *
  • Posts: 245
  • Country: no
Discharge current for capacitors
« on: November 18, 2018, 03:02:44 pm »
Noob q here. Was fooling around in LTSpice with a buck converter like design, and I noticed that the smoothing capacitor was charging and discharging at rates >300A, albeit just for microsecond or so.

I had selected "real" parts for everything but voltage sources and resistors, so the capacitor had a (hopefully) realistic ESR.

This made me wonder about maximum safe discharge current from a capacitor, and also how much current you realistically can discharge from a capacitor.

For example, take a 2200uF aluminum electrolytic capacitor charged to 12V. Would I even be able to discharge it at >100 amps? If so, would it be safe to do so repeatedly?

I tried looking at some datasheets, but not a lot specified beyond temperature. But without thermal resistance, I'm not sure how I can estimate temperature rise from the charging and discharging.
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Discharge current for capacitors
« Reply #1 on: November 19, 2018, 05:58:42 am »
Hi aheid,

You have touched on an interesting area that does not seem to be well covered, going by my recent search on the net for a tutorial.

Just a bit of basic stuff to set the scene: A capacitor has a maximum working voltage, and a maximum ripple current. The first parameter is always given on datasheets, but the second is normally only given for relatively large capacitors intended mainly for use as reservoir capacitors.

In the real world there is no such thing as a perfect component, and that applies especially to capacitors. At a very rudimentary level a real-world capacitor looks like a perfect capacitor in series with a resistor, called the equivalent series resistance (ESR).

In terms of how much average current a capacitor can pass, that is stated on relevant data sheets as the maximum ripple current.

Between the ESR and the ripple current you get heat according to the formula: I squared R, and it is this heat that is the ultimate limit on the ripple current that a capacitor can tolerate.

On to your question, how much peak current can a capacitor take. The answer is quite a lot, but the value does not tend to be included on data sheets.

To give an example of how much current reservoir capacitors do supply, consider the case of a typical half wave rectifier circuit comprising a rectifier diode and a reservoir capacitor. Say that 5 Amps is being drawn from the reservoir capacitor by the load.

As you no doubt  know, the diode charges up the  reservoir capacitor only at the peak of the input voltage sine wave, and it does this with huge gulps of current. The period during which the diode conducts is called the conduction angle. Working out the conduction angle is complex, so just consider a typical conduction angle of 7 degrees.

So, the situation is that the average current drawn from the reservoir capacitor is 5 A, but the diode has only got 7 degrees, out of 360 degrees, to supply that current to the reservoir capacitor. From this you can calculate the average current that the diode must supply during the conduction angle, by Iaca = Id * 360/Ca. And in the example you get an average current during the conduction angle (Iaca) of 257 amps. With a typical mains supply of 50Hz, these high current pulses occur every 20milli seconds (for a half wave rectifier).

I hope this illustrates that reservoir capacitors are capable of handling some pretty high currents, but what is the numeric limit. The answer is I do not know, but there are millions of rectifier circuits in the world operating without any problem.

Finally, here are a few words about the effect of a high current on the innards of a capacitor: The obvious effect is the instantaneous local heating of the internal conductors, joints and plates. Also the electrolyte can boil and give off gasses. And the other thing is the mechanical shock to the plates due to the electrostatic forces. To cater for al this, reservoir capacitors are sturdy.

I hope this very much simplified lecture goes some way to answering your question. :)
 

« Last Edit: November 19, 2018, 08:28:52 pm by spec »
 
The following users thanked this post: ggchab, aheid

Offline aheidTopic starter

  • Regular Contributor
  • *
  • Posts: 245
  • Country: no
Re: Discharge current for capacitors
« Reply #2 on: November 19, 2018, 07:42:40 pm »
Thank you for the very illuminating reply.

I had noticed the ripple current in the datasheets, but I hadn't realized it's importance. Your post lead me to discover how to find the RMS current in LTSpice, as well as this nice article. In my simulated case, a mere 250mA load generated 5A RMS current in the capacitor  :o

So clearly something to keep an eye on, although I'm assuming the true value would be somewhat lower as I haven't included resistances of PCB traces etc that a real-world circuit would have in my sim yet.

Thanks again!
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Discharge current for capacitors
« Reply #3 on: November 19, 2018, 08:22:28 pm »
No sweat,

It is very gratifying to know that my lecture helped. You seem to be a quick learner and have a solid engineering approach. :)
 

Offline RES

  • Regular Contributor
  • *
  • Posts: 109
  • Country: 00

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Discharge current for capacitors
« Reply #5 on: November 20, 2018, 09:03:09 pm »
Can you post the waveform?

While a discussion on capacitor ratings is fruitful, equally valuable is a discussion of numerical stability and running a SPICE engine.  It is by no means a set-it-and-forget-it process.  Most especially not with the rough compromises that have been made in LTSpice specifically.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Discharge current for capacitors
« Reply #6 on: November 21, 2018, 07:00:31 am »
Can you post the waveform?
If that question is aimed at me, no I do not have a lab at the moment and my wife is fairly unreasonable: she objects when I put my solder station, cutters etc on our nice new oak table.

But it is a great idea. I wonder if someone with a high-current rectifier (5A to 10A say) could do the honors. I would very much like to see the actual ripple voltage waveform against the rectifier diode current waveform.
« Last Edit: November 21, 2018, 07:03:12 am by spec »
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Discharge current for capacitors
« Reply #7 on: November 21, 2018, 07:12:15 am »
In my simulated case, a mere 250mA load generated 5A RMS current in the capacitor  :o
I only just noticed that you wrote 5A. For some reason I read that as 0.5A. 5A just seems way too high. Feel like posting the rectifier circuit so we can check?

Simulators are great tools, but you have to be careful of them- I will say no more in fear of triggering a simulator war. :palm:
« Last Edit: November 22, 2018, 05:07:23 am by spec »
 

Offline aheidTopic starter

  • Regular Contributor
  • *
  • Posts: 245
  • Country: no
Re: Discharge current for capacitors
« Reply #8 on: November 21, 2018, 02:10:36 pm »
I only just noticed that you wrote 5A. For some reason I read that as 0.5A. 5A just seems way to high. Fell like posting the rectifier circuit so we can check?

As I wrote, I was fooling around (sims are fun!) and while pursuing an idea I didn't notice I effectively had the capacitor shorted to ground via the switching MOSFET (I effectively had + side connected between L1 and M1, and - to GND in the attached circuit). So only thing holding it back was the ESR... :palm: :-DD

This goof however is what triggered the question I raised in this thread.

Moving the cap back in it's proper place, the RMS ripple current is now down to "just" 1.2A, which is within the limits of beefier caps from what I can gather, see attached image.

Simulators are great tools, but you have to be careful of them- I will say no more in fear of triggering a simulator war. :palm:

Indeed. I'm a relative noob, I feel I have a decent grasp of basic Ohms law and stuff, but I have limited experience with real circuits so far. So when I for example see hundreds of amps in the sim I have no idea if this is a sim artifact due to missing cable losses, parasitic elements or the numerical integrator etc, or if it's a decent reflection of reality.

I recently got a decent lab space tho so looking forward to building some of my sim experiments in order to get some intuition of how the real world compares to the sim world.
« Last Edit: November 21, 2018, 02:12:57 pm by aheid »
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Discharge current for capacitors
« Reply #9 on: November 22, 2018, 05:11:31 am »
In my simulated case, a mere 250mA load generated 5A RMS current in the capacitor  :o
I only just noticed that you wrote 5A. For some reason I read that as 0.5A. 5A just seems way too high. Feel like posting the rectifier circuit so we can check?

Simulators are great tools, but you have to be careful of them- I will say no more in fear of triggering a simulator war. :palm:

Ah, that clears the mystery :)

An interesting exercise, comparing sim with actual- let us know how it goes!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf