Fried Tantalum Capacitor 3s to 4s

Hello Everyone,

I am using the APM Plane for more than 2 years now. I started with the APM 1 (with the ATmega1280), then APM2.5/2.6 and more recently the Pixhawk. I’ve used the APM always in Planes.

More recently, I allocated an apm 2.6 to a DJI F450 with e300 power system . I started flying with it using a 3s Lipo. Yesterday I a changed to a 4s lipo. When I connected it I saw magic smoke coming form the PM. Please see the attached photo.

I was a bit astonished because a few minutes earlier I have flown with 3s battery with no problem… Certainly it couldn’t be reverse voltage! So I disconnected the PM and take a look. It was the big yellow component that produced the smoke… After a quick search I found it to be a tantalum capacitor 100uF 25V. I continued my investigation by understanding what this component was used for. It is a key component for the the PTH08080WAH voltage regulator module (http://www.ti.com/lit/ds/symlink/pth08080w.pdf). I read the datasheet and found a few things that made me a bit intrigued. First, looking to the bottom of the first page, the schematic clearly shows an electrolytic capacitor and the PM module used a tantalum capacitor… So I continued reading… On page 9 second paragraph I found this: “When specifying regular tantalum capacitors for use at the input, a minimum voltage rating of 2 × (maximum dc voltage + ac ripple) is highly recommended. This is standard practice to ensure reliability.” Ok So it is possible to use a tantalum capacitor. However, The the capacitor used is 25V capable and a 4s battery can output 16.8V. So, the 100% safety margin was not respected…

After this I removed the capacitor and cleaned the PCB with isopropyl alcohol. Everything is fine… It was just the capacitor that fried… The APM 2.6 is working properly.

I really would like to hear something from 3dr about what just happened to me. Thank you in advance.

Regards,

Pedro Santos

It appears the capacitor is polymer tantalum http://www.kemet.com/Lists/ProductCatalog/Attachments/252/KEM_T2016_T521.pdf. As stated in the datasheet polymer tantalum is not affected by the restriction I mentioned. Sorry about the confusion…

So I have no clue about what happened… Is it possible to have a large spike when charging the ESC capacitor that exceeds 25V?

Thank you again.

Regards,

Pedro Santos

The link you shared is to a similar capacitor but this is the specific one:
capacitoredge.kemet.com/capedge2 … X107K025AT
part number T491X107K025AT
I looked over your situation and looked into the specs on this part; the maximum voltage is rated at 25V at a maximum temperature of 85, when it’s pushed to it’s maximum operating temperature of 125, the maximum voltage drops down to 16.75V
I’m not assuming your previous flight got the component up to that temperature but 16.75V is easy to reach with a 4s LiPo battery.
Hope this info was helpful.

So does this mean we can’t safely use 4s LIPOs with the 3DR power module?

Read again and think… do you usually have >85°C in your copter?

[quote=“santiago”]the maximum voltage is rated at 25V at a maximum temperature of 85

when it’s pushed to it’s maximum operating temperature of 125, the maximum voltage drops down to 16.75V[/quote]

Hello all,

Thank you for your answers… When the capacitor blew the temperature was much lower than than 85ºC! I’ve read a bit about tantalum capacitor failure, and it appears that when the component fail it shorts the bus… That’s why it produced I lot of smoke! At least it didn’t fried anything!

I’ve replaced the original tantalum with a low ESR Panasonic Alluminium capacitor (120uF, 35V http://industrial.panasonic.com/www-data/pdf/AAB8000/AAB8000CE9.pdf ) . It is much better than the original (more endurance, more ripple tolerant and higher voltage) . I’ve made 3 flights of 20min each with no problem.

I think 3dr should revise the power module and change that tantalum capacitor with a different one, in order to increase reliability!

Regards,

Pedro Santos

Read again and think… do you usually have >85°C in your copter?

[quote=“santiago”]the maximum voltage is rated at 25V at a maximum temperature of 85

when it’s pushed to it’s maximum operating temperature of 125, the maximum voltage drops down to 16.75V[/quote][/quote]

I took that to mean temperature of the component. In a passively (or not at all) cooled computer/electronic system components can very easily reach this temperature, no?

A resistor - maybe. If it’s badly dimensioned.
But I don’t see how a (properly dimensioned) capacitor would get that hot in regular operations.
85°C is VERY hot! It’s 185°F!!!
Most of the electronics I have flying around on my tinkerdesk don’t get warmer than maybe 35-40°C in nonstop operation.
An example for bad abuse is my test-setup for the flight instruments - I use the 5V reg of an Arduino MEGA2560 to power the Arduino, an APM with GPS and radio and (currently) 2 gauges, which adds up to something like 1-1.2A. That’s borderline torturing the reg as it doesn’t have any cooling device and still it only goes to about 65°C.