Tantalum Cap as an Electrolytic Cap Replacement?

General project help for Adafruit customers

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
User avatar
jasonchildress
 
Posts: 53
Joined: Fri Apr 19, 2013 11:47 am

Tantalum Cap as an Electrolytic Cap Replacement?

Post by jasonchildress »

Hello all, I'm looking for some advice on shrinking down a component on my PCB project.

I have a custom ESP32 PCB running WLED and I currently have a 1000uf electrolytic capacitor on it for the LED strips. I would like to reduce the overall height of the PCB design and this cap is making that hard. I was wondering if I replaced the electrolytic cap with a SMD tantalum cap, it that was an equal replacement? If not a tantalum, are there any other SMD caps which can replace the electrolytic cap?

Follow up question: If I can replace with an SMD version but those short/small enough are not enough capacitance, can I simply chain multiple together to reach my desired amount?

Also, I do have some SMD electrolytic caps, but those are also too tall for my needs. I need this replacement cap to be no taller then the ESP32 module itself.

Thanks!

User avatar
westfw
 
Posts: 2006
Joined: Fri Apr 27, 2007 1:01 pm

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by westfw »

I have in fact added a 470uF tantalum cap to a neopixel ring, and it seems to work fine.
(Then again, neopixels sometimes work without any cap, and the “1000uF” value is just a suggestion…). It should work, and you can parallel multiple caps for higher capacitance.
Tantalums are somewhat infamous for failing in spectacular fashion if subject to relatively small overvoltages - you probably want at least a 10V cap for a 5V neopixel.

User avatar
adafruit_support_bill
 
Posts: 88041
Joined: Sat Feb 07, 2009 10:11 am

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by adafruit_support_bill »

(Then again, neopixels sometimes work without any cap, and the “1000uF” value is just a suggestion…)
We recommend the large capacitor as inexpensive 'insurance' against power supply spikes. Some power supplies can put out voltage spikes at power-up before their regulators stabilize. And we've seen whole strips destroyed by that.

The newer SK6812 based Neopixels are a little less sensitive than the older WS2812 chips. But a large-ish capacitor for protection is still good practice. As westfw notes above, you can parallel multiple smaller caps to increase capacitance.

User avatar
adafruit_support_mike
 
Posts: 67391
Joined: Thu Feb 11, 2010 2:51 pm

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by adafruit_support_mike »

If you have the option, check the supply rail with an oscilloscope in all of the system's expected operating conditions, and especially during the transitions from one to another.

Putting ideal capacitors in parallel increases the total capacitance like putting buckets in parallel increases the amount of water you can store.

Ideal capacitors don't exist physically though. What we have are about seven kinds of devices whose behavior is more capacitive than anything else: air, mica, ceramics (mostly barium titanate), plastic film, tantalum (solid or 'wet', with a liquid dielectric), electrolytics, and supercapacitors. Their approximation of an ideal capacitor ranges from excellent (air, mica, film) to BANNED (electrolytics), and they all have their own set of quirks.

For power filtering applications, the most important quirks are Equivalent Series Resistance (ESR: how quickly current can move into or out of the capacitor) and parasitic inductance. ESR limits the maximum frequency a capacitor can respond to, and parasitic inductance makes signals bypass the capacitor completely. At some frequency the capacitive impedance matches the inductive impedance and the capacitor resonates (its self-resonant frequency). Above the self-resonant frequency a real capacitor acts more like a resistor.

Because of all that, power filters usually need a set of capacitors.. small values to capture enough of the highest frequencies to let the larger, slower ones catch up, eventually leading up to the ones big enough to handle the largest spikes.

The interaction between quirks of different sized capacitors is hard to predict though, and can occasionally make a group of capacitors behave worse than any single capacitor in the set. Therefore it's always best to check a capacitor set's behavior for the specific input it will see in practice.

User avatar
jasonchildress
 
Posts: 53
Joined: Fri Apr 19, 2013 11:47 am

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by jasonchildress »

Thanks everyone for the Reponses! I've updated my PCB design to use a few tantalum caps in parallel.

I do have a follow up question regarding the amount of uf needed. I know that Adafruit recommends a 1000uf cap, but is that value assuming you're going to have certain number of LEDs in use? If my project only requires < 30 LEDs, do I need that much capacity? or can I go lower with 330uf?

User avatar
adafruit_support_mike
 
Posts: 67391
Joined: Thu Feb 11, 2010 2:51 pm

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by adafruit_support_mike »

The amount of capacitance you need depends on the amount of ripple rejection you want.

There are no hard rules, but 1uF per 1mA of expected current tends to be a good starting point. That should limit the slew rate of any voltage spikes to 1V/ms. That still sounds fast, but is slow enough for voltage regulators to handle gracefully.

User avatar
jasonchildress
 
Posts: 53
Joined: Fri Apr 19, 2013 11:47 am

Re: Tantalum Cap as an Electrolytic Cap Replacement?

Post by jasonchildress »

Thank you Mike!

Locked
Please be positive and constructive with your questions and comments.

Return to “General Project help”