Author Topic: What size capacitor for power circuit?  (Read 957 times)

0 Members and 1 Guest are viewing this topic.

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 364
  • Helpful? 10
What size capacitor for power circuit?
« on: March 04, 2012, 03:29:49 PM »
So here's the scenario: we have a golf cart with 48V total battery power, made up of 8 6V batteries.  We want to take one of the 6V's and use it to power the on board electronics.  In theory, the voltage will spike like hell, since the batteries are driving a big old DC motor.  We will, I assume, need a capacitor or something to filter it a bit.  But I'm not sure how people arrive at the values for their filter caps. 

Is there an easy way to figure that out, or do I just toss a 10uF onto it and see what happens? :-P

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline definitionofis

  • Jr. Member
  • **
  • Posts: 41
  • Helpful? 1
Re: What size capacitor for power circuit?
« Reply #1 on: March 04, 2012, 04:36:51 PM »
The spikes might not be so bad if the circuit is kept far from the battery/motor power wires since the battery is holding the voltage constant and there won't be much inductance. I think twisting together the positive and negative power supply wires going to the circuit might help reduce inductance.

I'll take a stab at it for fun though in case some inductance leaks across.
Compare my post with others before taking it as the ultimate answer.

You calculate R x C time constant such that time equals the period of the DC motor spikes.
I will guess they are 1/200 of a second wide (0.005s).

Then calculate R as the resistance of the circuit. It is 6v divided by the biggest amount of current you think the circuit draws.
I'll guess 0.030A (30mA).

6/0.030 = 200 ohms.

So RC = 0.005s (time width from above).

200C = 0.005
C = 0.005/200 farads = 5,000/200 microfarads = 25 mfd

and then you double it or more to be sure. You locate the capacitor as close as possible to the circuit, of course.
You could also locate smaller capacitors close to the sensitive inputs to ICs for example, like 0.01 mfd or something appropriate
to not interfere with the speed of their signals.
« Last Edit: March 04, 2012, 04:42:49 PM by definitionofis »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: What size capacitor for power circuit?
« Reply #2 on: March 04, 2012, 08:11:30 PM »
Hi,

So here's the scenario: we have a golf cart with 48V total battery power, made up of 8 6V batteries.  We want to take one of the 6V's and use it to power the on board electronics. 
If you mean using a 6V tap on the 48V battery pile, I'll advice against it, as that will create an unbalanced drain that, even if small, over time may develop into a chain with a weak link, which will matter for the entire pile.
Much better to go with a separate battery (or switching down from the entire pile, but going 48V to 6V, you need every percentage of efficiency that you can get).


In theory, the voltage will spike like hell, since the batteries are driving a big old DC motor. 
Theory is good, but nothing replaces a 'scope on the working motor under somewhat average conditions.
Getting specifics on both voltage, current. rise- and fall times and max spike time is crucial to calculate your needed filter.

Will the motor go on the total pile of 48V?
If so, you have both the division rate of 8 and the low impedance of the battery to help, but just in case you want to go with a tap... Always attack noise where it emanates, not at the other end of the wire.

Imagine a rope that two persons hold between them and one start shaking it up and down to create oscillations. Where do you think it will take the least amount of strength from you, grabbing the rope to kill the oscillations? (Poking the eye of one of them, while probably quite efficient, doesn't count here ;))


We will, I assume, need a capacitor or something to filter it a bit.  But I'm not sure how people arrive at the values for their filter caps. 

Is there an easy way to figure that out, or do I just toss a 10uF onto it and see what happens? :-P
Don't underestimate the value (and speed) of empirical selection.

In your case, I'd first concentrate on the motor and dampen the spikes it generates with caps and a transil/tranzorb/VDR of a suitable size. This could easily get the spikes to less than 60V and even if the battery didn't kill any of it, your circuit would only see 60/8=7.5V max at the bottom 6V battery, hardly a challenge for an LDO, or a Pi (CLC) filter or even a simple RC filter followed by a zener.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 364
  • Helpful? 10
Re: What size capacitor for power circuit?
« Reply #3 on: March 05, 2012, 03:19:40 PM »
Hm, thanks a lot for the info.  What I'm gathering is:

1) it is entirely possible to calculate a filter capacitance by the method (or something similar) posted by definitionofis
2) It is not ENTIRELY wise to tap a 6V battery from a 48V pile due to drain problems.

Fair points on the scope to figure out the actual info, instead of just guessing.

Awesome, thanks a lot.

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: What size capacitor for power circuit?
« Reply #4 on: March 05, 2012, 04:55:34 PM »
Hi,

1) it is entirely possible to calculate a filter capacitance by the method (or something similar) posted by definitionofis
The formula for sizing the capacitor is: C = A * s / V
Where...
C = capacity in Farad
A = current
s = time in seconds
V = voltage


2) It is not ENTIRELY wise to tap a 6V battery from a 48V pile due to drain problems.
Understatement alert!  ;D
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

 


Get Your Ad Here

data_list