I'm doing a project that will use a variable voltage output from an Arduino Micro. It varies slowly over time, so I filter the pwm a lot, resulting in little ripple in the outgoing signals. Here's my very simple circuit:
This circuit needs to be quite small, but I want to stick with through-hole technology. I went to Fry's to buy the caps, and found 'NTE brand, tantalum capacitors'. There were three different varieties: 10v, 25v, and 35v.
But they were all the same size! (and the same price, about $2 each.)
So my question is: what advantage is there to staying with the lowest voltage that would work? My circuit uses regulated 5 volts throughout, so I believe I would be fine with the 10v cap. But I bought 35volt caps instead, same price, same size ... What do ya think?
Make sure you don't reverse the polarity of tantalum caps. They don't like that.
Electrolytic caps are more forgiving.
That is not filtering a lot. That's just enough (~6mVpp according to LTspice).
Replacing the 4k7 resistor with a 10k resistor halves the ripple, and using PWM pin 5 or 6 halves the ripple again.
This circuit might be ok to produce a control voltage, but you can't "load" it.
An opamp buffer could fix that.
Leo..
It is always better to use the highest voltage rating on a capacitor for an application like this because you will be running it at a lower debating and therefore it wil be more reliable. That is it is less likely to fail.
The 10V cap should give you plenty of safety margin and be quite reliable.
I'd usually choose the 35V part for a different reason... I usually buy at least one extra part, and when I need a 10uF cap in the future I may need one with a higher voltage rating.
Running caps of all types well below the rated voltage improves linearity and probably reduces
dielectric absorption - only important if precision is wanted. Many ceramic capacitors are notably
non linear at the top of their voltage range, but this doesn't matter for decoupling which is their
main role. This is because they rely on the ferro-electric effect.
Thanks guys! Your opinions are well taken. Leo, This is actually on the output of an opamp. Should it be before the amp? Mike, good. Reliability is what I shoot for - why engineer when you can over-engineer? Doug, yay! that is exactly what I did - I needed 3 but bought 4 cause I KNOW I'll need another one at some point! Thanks Mark - I rarely think about the physics of something like a capacitor, but actually, I like thinking about that stuff. Joe, I guess I plead guilty. The convenience of stopping by my Fry's is well worth the extra few bucks to me. I really don't do well with ordering then waiting ... waiting ... waiting ... until it arrives. I had to have dinner with my family before I went to Frys, and wow. I am just plain impatient.
So I guess, different breakdown voltage does not affect the size of tantalum caps. It's interesting.
ChrisTenone:
This is actually on the output of an opamp. Should it be before the amp?
I assume you want to convert PWM into a variable voltage.
Then you use this circuit directly on the PWM pin.
0-5volt is available on the capacitor. You can measure that with a digital volt meter.
But you can't "drive" anything with it.
Connect the voltage to the +input of an opamp. And the output of the opamp to the -input.
Then the opamp buffers the voltage (more current available).
Add two resistors to the -input, one to ground and one to the output, and you amplify the voltage. e.g. 0-10volt.
Leo..
Wawa:
I assume you want to convert PWM into a variable voltage.
Then you use this circuit directly on the PWM pin.
0-5volt is available on the capacitor. You can measure that with a digital volt meter.
But you can't "drive" anything with it.
Connect the voltage to the +input of an opamp. And the output of the opamp to the -input.
Then the opamp buffers the voltage (more current available).
Add two resistors to the -input, one to ground and one to the output, and you amplify the voltage. e.g. 0-10volt.
Leo..
Thanks. So, I'm doing it backwards? Here is the circuit for my analog-ish (and atomic-ish) clock driver:
ChrisTenone:
Thanks. So, I'm doing it backwards? Here is the circuit for my analog-ish (and atomic-ish) clock driver: ...
Ok, I get it now I think: 1) filtering before the opamp puts less strain on the amp since it does not to follow each rise and fall of the pwm signal. And 2) this also puts less strain on the resistor (and produces less heat) when the load is high. A pre-filter will only see about a milliamp, regardless of how much current is drawn at the H, M, S outputs.
Thanks again.
The breakdown voltage does affect the size of a capacitor. In this case however because of the actual capacitance they have managed to fit diffrent voltages in the same package. If they had wanted to, these lower voltages could have been smaller.
Not so much a strain but an impedance question. The output of the filter is a high impedance, this is fine feeding it into the input of an op amp, but useless if you want to feed it into a low impedance input like your load. An opamp's output is much better to feed into a load.
Well not empty but filled with non functioning plastic encapsulation. They do that so the range of capacitors are in the same size package and this makes production easer. It applies to all types of capacitor except ceramic. For any given technology the volume of a capacitor is proportional to the capacitance and working voltage.
Grumpy_Mike:
Well not empty but filled with non functioning plastic encapsulation. They do that so the range of capacitors are in the same size package and this makes production easer. It applies to all types of capacitor except ceramic. For any given technology the volume of a capacitor is proportional to the capacitance and working voltage.
Yep, that's what I've always thought - bigger = more capacitance and higher breakdown voltage. To a certain extent, size in inversely proportional to the newness of the manufacturing technology. (Many of my electrolytic caps are 1970s vintage. In general, they are up to 1.5 times as big as current electrolytics. Ceramic discs are about the same though.)
Doug, yay! that is exactly what I did - I needed 3 but bought 4 cause I KNOW I'll need another one at some point! Thanks Mark - I rarely think about the physics of something like a capacitor, but actually, I like thinking about that stuff. Joe, I guess I plead guilty. The convenience of stopping by my Fry's is well worth the extra few bucks to me. I really don't do well with ordering then waiting ... waiting ... waiting ... until it arrives. I had to have dinner with my family before I went to Frys, and wow. I am just plain impatient.
