Bus board decoupling capacitors voltage question

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, Joe., luketeaford, lisa

Post Reply
User avatar
sutekina bipu-on
Ultra Wiggler
Posts: 883
Joined: Fri Nov 17, 2017 9:24 pm
Location: USA

Bus board decoupling capacitors voltage question

Post by sutekina bipu-on » Fri Nov 01, 2019 2:41 am

I didnt order any new capacitors for my covfefe eurorack bus boards. They have two spaces for 100uf-470uf caps. I have some 16v 470uf's and just wanted to be sure that was an acceptable voltage rating for the caps. I figured each one is for each 12v line but i would rather not find out the hard way if i should for sure use 25v or higher :hihi:

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Fri Nov 01, 2019 8:33 am

16V should be OK.

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3318
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Post by Graham Hinton » Fri Nov 01, 2019 8:39 am

Why do you think you need them at all? Have you bothered to work out what they are actually doing?

User avatar
latigid on
Ultra Wiggler
Posts: 967
Joined: Fri Oct 25, 2013 11:05 am

Post by latigid on » Fri Nov 01, 2019 9:27 am

I typically overrate caps by 2-3x the expected input voltage, so for 12V I'd maybe go for 36V, but to my knowledge Graham is correct here and the caps do basically nothing in this situation.

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Fri Nov 01, 2019 11:07 am

I've just reviewed literature from a number of makers of aluminum electrolytic capacitors. They all state that voltage derating is unnecessary. They do say that a capacitor operated below its max temperature will double its life for every 10 degree C reduction. There is a slight advantage to voltage derating, but:

"Reducing the amount of voltage applied to the capacitor can also increase the expected life of the capacitor. The expected life can be increased by the ratio of applied voltage to rated voltage for the capacitor. The expected life by voltage derating is limited to a 2:1 ratio even if the ratio is more than 50% of the rated voltage of the capacitor."
https://www.illinoiscapacitor.com/pdf/P ... eneral.pdf

You'll get a lot more reliability benefit from keeping the temperature and ripple current down. 25% voltage derating is very reasonable. Do keep in mind that the PEAK voltage (e.g. in a ripple filter) is what counts.

User avatar
sutekina bipu-on
Ultra Wiggler
Posts: 883
Joined: Fri Nov 17, 2017 9:24 pm
Location: USA

Post by sutekina bipu-on » Fri Nov 01, 2019 11:53 am

Graham Hinton wrote:Why do you think you need them at all? Have you bothered to work out what they are actually doing?
No, i'm trying to teach myself EE stuff in my spare time but there is a lot i don't know so my approach has been to ask when i can't google my way to an answer.

I have absolutely zero ability to read a schematic or PCB and work out what parts may or may not be doing anything useful so when i see a component not strictly labeled for what it needs to be i go :hmm:

Honestly i would really appreciate it if you let me know what fundamentals im missing to not have known that. "All of them" is a valid answer but im trying to learn as i go :despair:

Anyway, i'll slap the 16v's on, they're nice and low ESR :) Thanks everyone.

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3318
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Post by Graham Hinton » Fri Nov 01, 2019 12:40 pm

sutekina bipu-on wrote: Anyway, i'll slap the 16v's on, they're nice and low ESR :)
So you don't know why, but you'll put some on anyway?
If they make any difference at all it means there is something wrong with your PSU or your modules. Either way, this isn't the fix. You may as well sacrifice a chicken.

Work it out. You need an R or L with a C to make a filter and with the values of R involved on a busboard the cutoff frequency is going to be above the audio range. f = 1/2πRC is all you need to know at this stage. If you want to learn EE, don't copy people who don't know what they are doing.

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Fri Nov 01, 2019 1:03 pm

sutekina,

Welcome to our community.

Don't let cranky people wear you down. A very small percentage of us engage in that behavior.

User avatar
Kevin Mitchell
Wiggling with Experience
Posts: 297
Joined: Tue Jun 27, 2017 10:55 am
Location: Springfield, Massachusetts

Post by Kevin Mitchell » Fri Nov 01, 2019 1:38 pm

You don't need to be an EE to follow a build guide. If the caps aren't needed it's certainly not the consumer's fault they're in the design :despair:

Build on and prosper. And maybe get that guy some TLC
-KM

User avatar
Mungo
Super Deluxe Wiggler
Posts: 1186
Joined: Fri Jan 04, 2013 8:12 pm

Post by Mungo » Fri Nov 01, 2019 6:42 pm

Kevin Mitchell wrote:You don't need to be an EE to follow a build guide. If the caps aren't needed it's certainly not the consumer's fault they're in the design :despair:
Exactly, just following the instructions to the best of their ability.

A more complete answer includes the usual "it depends" as there are many things interacting. Many small power supplies are improved by adding some/more/larger capacitors to the bus board:
http://mungo.com.au/euroPowerZ.html
But at the same time its those same power supplies that often have limits to the amount of capacitance before they will fail to start up, so there are bus boards that have switches for the end user to remove the capacitors!

Without knowing the specific power supplies, cabling, and bus boards being used any answer could be completely right or wrong, it depends.

As far as the original question from the OP, different people and different manufacturers have different ideas about derating, and it can change with different capacitor types. But the vast majority of modern electrolytic parts have no voltage derating recommended by their manufacturer, for longest life running closer to the rated voltage can be an improvement (forming). Since we dont know the busboard we can't comment if there are any capacitors from -12 to +12 that would need a 25V rating.

User avatar
sutekina bipu-on
Ultra Wiggler
Posts: 883
Joined: Fri Nov 17, 2017 9:24 pm
Location: USA

Post by sutekina bipu-on » Sat Nov 02, 2019 7:41 pm

Thanks everyone, and thanks Mungo for that writeup. My plan is a 2A AC wall wart to a NLC small PSU board to a NLC covfefe bus board, thats all. I may add a 2nd bus board down the line. I got some big ol heatsinks to mount to the voltage regulators.

http://nonlinearcircuits.blogspot.com/2 ... u-pcb.html

I found some 35v 470uf's today... I tried calculating f = 1/2πRC but I don't know what cutoff frequency i care about or resistance coming from the PSU power terminals as my small power pcb isn't complete yet (awaiting my voltage regulators from Tayda). I thought the point of decoupling caps were to provide a constant voltage though, not act as a LPF?

Am i mixing myself up? :lol:

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Mon Nov 04, 2019 12:07 pm

sutekina bipu-on wrote:I thought the point of decoupling caps were to provide a constant voltage though, not act as a LPF?
That's correct - the decoupling capacitor is there to supply energy when the load circuit pulls current spikes. By supplying the energy close to the load, we reduce the amount of current spike that passes through wiring between buss board and power supply. Any resistance or inductance in that wiring will result in the current spike becoming a voltage spike, so keeping some of the spike away from some of the wiring contributes to cleaner voltage in the system.

User avatar
mskala
Super Deluxe Wiggler
Posts: 2714
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Mon Nov 04, 2019 8:34 pm

sutekina bipu-on wrote:I thought the point of decoupling caps were to provide a constant voltage though, not act as a LPF?
Those are kind of the same thing. A constant voltage is the lowest frequency you can get, zero Hertz. Changes in the voltage are higher-frequency transients, or continuous higher-frequency signals in the case of stuff like power line or switching-regulator ripple; keeping the voltage constant is equivalent to filtering those higher frequencies out.
North Coast Synthesis Ltd.
Latest Web log posting: Vexations

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Tue Nov 05, 2019 8:50 am

mskala wrote:
sutekina bipu-on wrote:I thought the point of decoupling caps were to provide a constant voltage though, not act as a LPF?
Those are kind of the same thing... keeping the voltage constant is equivalent to filtering those higher frequencies out.
Not exactly. Decoupling caps aren't there to filter noise from the PSU and keep it away from the load. They are there to prevent current spikes/ripple in the load from becoming voltage spikes (due to resistance or inductance in the wiring from PSU to load), and to keep those current spikes away from other loads (modules) and the PSU.

User avatar
mskala
Super Deluxe Wiggler
Posts: 2714
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Post by mskala » Tue Nov 05, 2019 9:14 am

jorg wrote:
mskala wrote:Those are kind of the same thing... keeping the voltage constant is equivalent to filtering those higher frequencies out.
Not exactly. Decoupling caps aren't there to filter noise from the PSU and keep it away from the load. They are there to prevent current spikes/ripple in the load from becoming voltage spikes (due to resistance or inductance in the wiring from PSU to load), and to keep those current spikes away from other loads (modules) and the PSU.
A filter between the power supply and the load, and a filter between the load and the power supply, are both filters.
North Coast Synthesis Ltd.
Latest Web log posting: Vexations

jorg
Ultra Wiggler
Posts: 820
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Tue Nov 05, 2019 2:32 pm

It's a matter of perspective. Are you concerned with "downstream" or "upstream" filtering?

If "downstream" (PSU to load) then as Mr. Hinton suggests, you want to calculate the performance in terms of ripple entering the filter from the PSU, which will depend on the source resistance from the PSU's components (e.g. diodes).

If "upstream" (load to PSU) then you are concerned with how many Coulombs of charge are injected by each spike from the load, and how much that will change the voltage on the decoupling capacitor.


Post Reply

Return to “Music Tech DIY”