So I am using a 5V Adaptor to power my arduino via the 5V pin. To protect the Arduino against sudden voltage spikes when the adaptor switches on, I was thinking of attaching a zener diode 5.1V to GND. I was also thinking that maybe a capacitor would also be a good idea. I need to know whether the capacitor should be between the Adaptor +ve and GND, or parallel to just the +ve. Also, how to calculate the required capacitance. What's the maths? Thanks.
See my link:-
http://www.thebox.myzen.co.uk/Tutorial/Protection.html
Or see :-
Protecting Inputs in Digital Electronics | DigiKey.
Between power output and ground, but you also need a resistor before the capacitor to have something to drop the voltage excess across.
What is the width of these spikes? Use 3 times the time constant of your circuit RC.
@Grumpy_Mike Is this what you are proposing?
I have no safe way of exactly knowing the spike widths, but they should be not more than a tenth of a second. If I use a 100 Ohm resistor and a 1000uF capacitor, that should do it?
Do you even know whether these spikes exist?
You might be trying to protect from something that isn't even a problem.
I am sure there are spikes, I have just no way to find their amplitude or width. And as @Grumpy_Mike writes in his article its a good habit, and I like adding such habits into my SOPs just to be safe, they hardly cost anything.
And @Grumpy_Mike maybe 100 ohm is too small? How about 1K?
No.
C1 should go to to D1 and R1 should go between the power supply output and the junction of C1 & D1.
You should measure them using an oscilloscope, assuming that as @JohnLincoln says you have any.
The time constant of those values is 0.1 seconds. I said three times the time constant.
However, a series resistor of 100Ω is way too high, because at 500mA current draw you will drop the voltage input voltage to zero (well by 50V to be exact, which is next to nothing)
Did I get you right this time?
Is R absolutely necessary? What value of R would make it?
P.S. I don't have an oscilloscope.
Yes that is right.
Absolutely because without it any spike would go straight onto the capacitor and charge it up the the full spike voltage. This would probably destroy the capacitor.
I would make so that it drops half the voltage that you would get from the current rating of D1.
Then you can't design it properly. Unless you know:-
- The duration of any spike
- The height of any spike
- The current you want to draw from the power supply.
This just leaves you with what i think is unrealistic guesses.
2 Likes
When you say at 500mA current draw, you mean 500mA through the Zener diode, right?
One more thing, if we have a resistor is series, would the arduino get full 5V? Wouldn't it experience a voltage drop?
Are you only concerned about spikes when it turns on or at any time?
Only when it switches on.
A turn on delay circuit would work but it'a a little more complicated than just a zener and resistor.
In the circuit suggestion given by @Grumpy_Mike , is there a risk of voltage drop happening for the arduino? The same power supply would be powering other stuff also, so the more current is drawn the more the drop? Perhaps we could add individual resistors to both the zener and the capacitor in series to each ??
Does this thing make any sense?
No.
1 Like
Come on budd, elaborate on my "voltage drop to Arduino" fears. I'm not an expert in electronics else I won't be here, but you guys are.
Yes there would be a voltage drop. Use ohms law to determine what it would be
E (voltage) = I (current) x R (resistance)
Say you have a 2 ohm resistor in your power line and everything you have connected draws 500mA.
So E = 500mA x 2 ohms = 1V. Yes, your 5 V supply would now drop to 4V
No.
I mean that if you had the series resistor you suggested in the circuit and the circuit was supplying 500mA then you would have no voltage left.
Yes it would, you need a voltage drop to limit the current, without it you will get unlimited current through the zener diode, and fry that.
You could replace that resistor with an inductor and that will limit the speed of the current rise without adding much pure resistance. It all depends on the value of the inductor and the resistance of its wire.
And we are trying to help, but you don't seem to get it.
You are asking for things you can not characterise. Yes you can get a voltage spike on switch on of power supplies. However, not all power supplies show this, it tends to happen with lower quality power supplies. As I have said before taking guesses at possibly surges in a power supply is not the way to go.
Its possible (in fact likely considering we have this problem) that the power supply is bad quality), this adaptor used to be an old phone charger. Anyways, if I provide the supply through a LM7805 linear regulator, would that do the trick?
P.S. I've noticed that when I have my soldering iron connected to the same multi point socket board as the adaptor, things get quiet noisy. Does that happen often? Would a linear regulator help with that also?
No, the 7805 should have at least a 7V input in order to regulate.
But in principle it would work? If I provide a 7V input? Or do I need to go for the ultra low noise LDOs?
And if I am looking for some high quality 5V supply, how should I look for it? Some specific brand?