Which value of decoupling capacitor to use for power rail

Thread Starter

Yami

Joined Jan 18, 2016
354
I have got a faulty Dyson bladeless fan control board with a shorted out decoupling capacitor. I have removed the capacitor which have cleared out the short. I don't know which value of capacitor to use as there are no schematic available and its a 0603 package capacitor.
The capacitor is placed on the main board on the trace which goes to the small board which has got the IR sensor, buttons etc.

Thanks in advance for the help.
 

Thread Starter

Yami

Joined Jan 18, 2016
354
decoupling what chip? anything between 10nF and 100nF would be a common value?
Could I be using the term 'decoupling' wrong? The capacitor is not connected to the micro controller or any IC for that matter, but its placed close to the micro. One end is connected to the 5V pin which goes to the small board and the other end connected to the ground.

Can you post a pic of the cap?
Max.
The cap is annihilated hehehehe, perhaps I could post a photo of where I have removed it from, would that be of help?
 

Irving

Joined Jan 30, 2016
3,843
That's exactly the right term, it decouples (or decoupled) the power rail from the component. It provides a local energy store to handle transient demand from the component and thus protect others from the switching spikes generated in the component.
 

Thread Starter

Yami

Joined Jan 18, 2016
354
Thanks for the help :) Just out of curiosity what would be the worst case scenario if I were to use a larger value cap or smaller value cap from the original?

Once again thanks for the help.
 

MrChips

Joined Oct 2, 2009
30,707
Thanks for the help :) Just out of curiosity what would be the worst case scenario if I were to use a larger value cap or smaller value cap from the original?

Once again thanks for the help.
Nothing. 100nF is typical.
Post a photo of the entire board.
 

MrChips

Joined Oct 2, 2009
30,707
One question,

Decoupling capacitors dont blow for no reason,
have you identified and removed the reason ?
Not entirely true. The decoupling capacitor is across the power rails. Possible causes of failure are:

1) power surge
2) capacitor voltage rating exceeded
3) capacitor voltage rating was too low
4) faulty capacitor

I have replaced many power rail decoupling capacitors that had fail shorted. They were never 100nF ceramics.
They are typically 10μF tantalum electrolytic capacitors.
 

Irving

Joined Jan 30, 2016
3,843
Very strange, 100 nf Ceramics are normally like 110 % reliable. :->
You'd think so... but go and watch Louis Rossmann's (Apple MacBook repair) YouTube channel, amongst others, and see how many times its a flaky power rail capacitor . MLCC, at higher values 100nF & up have incredibly thin dielectric, esp in the smaller SMD sizes. Typically a 100nF small SMD will be 10 or 16v rated and they don't tolerate much above that.


Tantalums used to be a thing, now no one in their right minds would/should use one!
 

Thread Starter

Yami

Joined Jan 18, 2016
354
Here is a photo of the PCB, I have marked the cap which was shorted out (its removed in the photo). Its on the 5V rail. I have also marked a point where it looks like it has burned, its also on the same trace. Actually there was some kind of 'gooey' substance stuck there - possibly some kind of grease or something that has caused the tracks to corrode. I have cleaned the area and touched it with some solder and have applied some epoxy as solder mask. I am not sure that it has caused the cap to fail or not.
I have replaced the cap with a 100nF cap and have run the fan - its working. I have checked for any heating, no bad signals yet.
The corroded tracks goes to a connector which goes into the fan assembly, I am assuming its connected to some kind of sensors. The connector below that (near the dyson logo) is the U V W motor connection.
Thanks for the help
IMG_2061.JPG
 

Irving

Joined Jan 30, 2016
3,843
The corroded tracks goes to a connector which goes into the fan assembly, I am assuming its connected to some kind of sensors.
If there are 3 it will probably be the hall effect sensors to feed back fan rotor position (sinusoidal, 3 @ 120deg phase shift) . Alternatively, and less common for this application, it will be a quadrature a/b square wave on 2 and an index pulse on the third.
 

MrChips

Joined Oct 2, 2009
30,707
FYI only.
If the bad capacitor was across the power rail then you could use an ohmmeter (with power OFF) and find all the other decoupling capacitors that are across the power rail. You will notice some are physically larger than the one you had to replace.
0603 package is relatively small. Are you certain on the size? 0603 would be more like 10nF.
If you replaced it with 10nF or 100nF you would be fine.
If it is working, let it be. It will be OK.

Congrats on a good repair job done!
 
Top