I have been working on a circuit board for an Arduino project. It contains an LCD screen, linear regulators and several relays. In an attempt to remove voltage spikes and dips, I have added several capacitors with several different values all around the circuits. I may have over done it.

I am wondering if having lots of capacitance on my power lines will cause an unsafe draw of power from my transformer when I turn it on? How can I prevent this? Also, will having lots of capacitance cause problems for the Arduino and the LCD circuit at start-up (because of a lack of power availability at first)?

I am powering an:

  • One Arduino Mega
  • One 32 character LCD screen
  • Two 9 volt to 5 volt linear regulators
  • 16 relays
  • Two large 8 segment LED displays
  • About 20 LED strip segments

with a 2 amp, 9 volt power supply.

Near the power supply: 100 uF and 10 uF
Near the Arduino Mega: 100 uF and 10 uF
Near both linear regulators: 10 uF and 1 uF on both sides
Near Relay power lines: 1 uF and a zener diode
Near LCD Power lines: 1 nF and 100 nF
Near all 9 digital in/out lines: 100 pF and a zener diode
Near all 20 switches: 100 nF

Total is getting close to 300 uF

  • 3
    \$\begingroup\$ "Inrush" can be a problem and it totally depends on the power supply and capacitors added. Given the info you have given, that's all that can be said. \$\endgroup\$
    – Andy aka
    Sep 4, 2015 at 22:11
  • \$\begingroup\$ Is this a theoretical problem or is something actually wrong? Charging a few caps is unlikely to cause any problems as far as "transformer power" is concerned, but it's more of an issue if you're talking about some regulators in said "transformer" \$\endgroup\$
    – Nick T
    Sep 4, 2015 at 23:07
  • \$\begingroup\$ I would suggest that you total up the capacitance on the adaptor, and also let us know how much capacitance you put on the two 5V rails, then provide a link to the adaptor. Seldom is there a reason to put more than 50 or 100 uf on a micro-processor board unless there is motor control or heavy RF transmit activity or something like that. But this is just a rough guide. If you have special circumstances, it could be justified. \$\endgroup\$
    – user57037
    Sep 4, 2015 at 23:08
  • \$\begingroup\$ By the way, the problem I have usually had with inrush is that the inrush current causes a fuse to blow. \$\endgroup\$
    – user57037
    Sep 4, 2015 at 23:09
  • 3
    \$\begingroup\$ At 10,000 uF I'd worry about this; at 300uF or 1000uF if it's a problem, then your supply must be extremely fragile... \$\endgroup\$
    – user16324
    Sep 5, 2015 at 10:24

2 Answers 2


You will have no problem with only 300uF of capacitance. Although the comments by other posters are true you won't see any problems at this level.

300uF from a 2Amp supply will meant it will only take ~2ms to charge up to 5V. The amount of energy involved will not cause any trouble.

The comment about conduction angle for the rectifier is true but there is invariably enough resistance and leakage inductance in the transformer and series resistance in the capacitor so that it is not a problem in low power supplies - if you were designing a supply for many tens of Amps then other configurations would need to be considered.

  • \$\begingroup\$ At what level of capacitance would I start to have problems? \$\endgroup\$
    – Hoytman
    Sep 5, 2015 at 16:05
  • 1
    \$\begingroup\$ Up to a few thousand microfarad would probably not be any issues. I would expect the first thing you would notice is that the supply took a second or two to power-up that could affect how the Arduino started up. I would not expect any physical damage. \$\endgroup\$ Sep 5, 2015 at 18:05
  • 1
    \$\begingroup\$ Heck an rc delay to hold the arduino in reset until the caps charge would help. \$\endgroup\$
    – Passerby
    Sep 5, 2015 at 18:16
  • \$\begingroup\$ That could work. The ATMEGA328 in the Arduino has a fairly good power-on reset and I have never had a problem with it but it may not work for very slow rise times. \$\endgroup\$ Sep 5, 2015 at 18:24

A large capacitor behaves as a short to a power supply.

A dc power supply of 2A at 5V would charge a 300uF capacitor in 0.75ms only if the capacitor equivalent series resistance ESR was 2.5 ohms. However, since the ESR is within 100mohms, the charging current rises to 50A, which is way beyound the capability of the power supply. In effect, the power supply will burn.

To prevent the damage to power supply, you need to insert a series resistor of 2.5ohms in order to limit the charging current to 2A.

Now to answer the question: adding extra filtering capacitance in parallel to a power supply could burn it.

  • 2
    \$\begingroup\$ " the charging current rises to 50A, which is way beyound the capability of the power supply. In effect, the power supply will burn." - a 2 amp rectifier diode might be rated for 80 amps surge, and (as Kevin White pointed out) a small transformer will probably have enough resistance and leakage inductance to limit the current. I measured the resistance of a 9V 2A transformer and it was 0.5 ohms. When combined with capacitor ESR (and ignoring diode resistance) that would limit charging current to 15 amps - well below the diode's surge current rating. \$\endgroup\$ Feb 28, 2020 at 6:33
  • \$\begingroup\$ you present a case of almost-matching impedance. You measured 0.5 ohms and your capacitor must have had an ESR of more than 100 milliohms (or 0.1 ohms). Even so, the current you mention is 15 A, more than 7 times the rated 2A. This proves my point that capacitor is a short-circuit for power source and if capacitance increases too much will burn the source at charge time. \$\endgroup\$
    – WindSoul
    Feb 28, 2020 at 15:25
  • \$\begingroup\$ "if capacitance increases too much will burn the source at charge time" - but if it's not too much it won't. 300uF is unlikely to be too much for a 9V 2A power supply (the OP's scenario). The power supply I tested already has a 6800uF capacitor in it (and a 5A slow-blow fuse). \$\endgroup\$ Feb 28, 2020 at 20:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.