Effect of mechanical shock on Tantalum capacitors

Discussion in 'General Electronics Discussion' started by Emam, Jul 7, 2014.

1. Emam

58
2
Jul 7, 2014
Hello,
I have a circuit with a tantalum capacitor Inside.
The capacitor is connected in parallel to a DC generator (a coil and a magnet). The generator charges the capacitor correctly.
For some reasons I should provide high linear acceleration to my circuits (something like 80'000 m/S2).
However, after this acceleration, the capacitor discharges rapidly.
Do you know why? Is there any relation between mechanical acceleration and tantalum capacitors?
Bests regards

2. shumifan50

548
56
Jan 16, 2014
Wow, that is some acceleration 0-80kph in one second. About three times that of a quick motorcycle. That is quick, but it would surprise me if it has any affect on the cap, unless the acceleration is not linear.

1. How long does this acceleration last and what is the top speed?

2. How big is the capacitor? And what current do you draw from it?

3. If you replace the cap, does the circuit behave as expected again?

4. How do you stop the projectile at the end of the test? The deceleration might be a bigger problem.

Last edited: Jul 8, 2014
3. KrisBlueNZSadly passed away in 2015

8,393
1,269
Nov 28, 2011
It should be easy to accelerate something small at 80,000 m/s2 - just hit it with something heavy! The acceleration isn't necessarily sustained like it is with a vehicle. Or he might be talking about deceleration; AFAIK they're the same thing from the point of view of the thing being accelerated/decelerated.

I can't answer the question. But I suggest that you post a schematic and some photos. Also, you know that if you connect a capacitor directly across a coil, it will discharge into the coil, right?

If you have some way to stop it from discharging back into the coil, then I would try to find out whether it's actually the acceleration/deceleration of the capacitor itself that's causing the discharge, or whether it's due to something else. If possible, move the capacitor off the part being accelerated/decelerated and connect it with long flexible wires, so the capacitor itself doesn't move, and see whether that fixes the discharge problem. If it does, clearly the capacitor doesn't like being accelerated/decelerated. You might be able to fix this (permanently, or as a test) by shock-mounting the capacitor.

shumifan50 likes this.
4. shumifan50

548
56
Jan 16, 2014
It depends what the OP means by linear; the acceleration is linear or it is accelerated in a straight straight line. That is why the questions.
The coil discharge is a good point - I assumed the circuit behaved as expected before launch and failed after launch. Maybe that is a bad assumption.

I am not sure but from memory tantalum is a clay and I would guess if it is subjected to an impact, it might crack/shatter. The OP states the acceleration, but if it is decelerated by hitting a solid surface the deceleration will be huge compared to the acceleration as it will slow down from terminal speed to 0 in a fraction of a second. That is why it would be nice to know how the object is stopped.

5. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,404
2,777
Jan 21, 2010
Let's see, 80,000 m/s/s is over 8000 G

Yep, if the tantalum cap us undergoing this sort of acceleration it would probably fail.

Imagine it's held in place by its leads. Imagine it weights 1 gram. Now hang an 8 kilo weight onto the capacitor.

Depending on the rate of change of the acceleration (it's called jerk, and is measured in m/s/s/s) fractures might be even more likely.

If the cap is not being accelerated, but is rapidly discharging into an inductor I agree with Kris' suggestion (did he make it, or did I just read it into his comments??) that it might be being destroyed by reverse voltage.

6. Emam

58
2
Jul 7, 2014
Dear all,
In fact, the capacitor is not directly conntected to the coil. There is a switch between them which we can open (turn off) once the capacitor is charged.
The accerleration is not only linear. It is also rotational.
What we do is that we want to produce new intelligent systems for producing Fireworks.
So we use projectile launcher (with linear 80'000 m/s2 and rotational acceleration of 50 Hz). Then we put our circuits inside.
After few seconds, the circuits should be able to produce light in high altitudes. So there is no mechanical impact or decceleration.
The problem is that when we recovered failed circuits after launching (and off couse after falling down to the earth), we saw that the capacitors could not save energy and discharges easily.

- Do you think that this is the problem of high acceleration of the departure?
- or something else?

Thank you very much for your help.

7. shumifan50

548
56
Jan 16, 2014
It seems to me you answered your own question: 'falling down to earth'. I think you should be investigating the impact of your landings.

8. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,404
2,777
Jan 21, 2010
You're accelerating the electronics at 8000G. Are they potted?

Do you know what would happen to you if I launched you at 100G? (the answer is "jam").

I'm not at all sure you understand what 80000m/s² actually means. What it means is that the projectile, if accelerated for just over 1/8 of a second would have sufficient velocity to escape the earth's gravitational field (of course that assumes constant acceleration).

After 1.6 seconds it is officially in space, and after 5 seconds it is 1000km from the earth's surface.

In under 2 minutes it passes the moon at a speed of 400 km per second.

In a little over 30 minutes it would reach the sun going at 152 thousand kilometers per second

In just over an hour it starts overtaking photons.

You say it does this for a couple of seconds then falls back to earth. Your figures are wrong because after 1.6 seconds it's above the bulk of the atmosphere traveling way above the earth's escape velocity, which means it's not coming back.

How about you get some realistic figures?

As I said, even if it starts at 8000G, it's probably going to mash the electronics right there.

9. shumifan50

548
56
Jan 16, 2014
@Steve:
You're right. I converted 80k/s to 80kph at a ratio of 1:1 - wish we could do that, travelling would be quick if we survive.