How to calculate battery requirements based on current draw of robot hardware?

[Marvin]

 
How can the battery energy storage capacity requirements(in mAh) to effectively run my robot for say half an hour non-stop be calculated from the current draws of the actuators and other electronic components (including the extra current drawn from having a heavy robot )? Is there one formula or a series of calculations I can use to calculate battery ?

 for example:
-I have 2 servos wired into one digital port and 2 of the other digital port (not sure if this is the right way to connect more servos?)
-a servo for infrared rangefinding connected to the third digital port
-a rangefinder to adc03
-at present i only have a 1600 mah 6v battery regulated to the vcc
-atmega8 of course

further to this:
Can I connect another battery of the exact same specs. in parallel to ameliorate my situation?

[waltr]

This does that some measuring and a few guesses.
In theory a 1600 mA-Hr battery will deliver 1.6A for 1 Hour. But due to internal losses and other factors this will be less, more like 1.6A for 50 minutes or less.

So the first step is adding up the average current draw from all of the devices the battery powers. For some of these you can use the current on the data sheet for others you will need the measure. The processor and sensors normally have a fairly constant current draw but motor and servos do not. Also the greater the load (mass they must move) the greater the current draw. These do need to be measured or take a guess from their specs.
There is a battery and an energy calculator in the SoR tutorials:
http://www.societyofrobots.com/calculator.shtml
These will give a good first estimate.

Batteries can be paralleled to increase the capacity but be careful. The batteries should be matched exactly for charge and discharge characteristics. Many times it is better to power different circuits from separate batteries. Like the processor and sensors from one battery and the motors from a different battery. Another way is to parallel two batteries for powering the bot and charge each separately.

[Marvin]

thanx waltr,
So i can use identical nimh 1600mAh batteries in parallel?
 They need the same full charge at the start?
 Do they're discharge rate the same in most instances?

[Soeren]

Hi,

So i can use identical nimh 1600mAh batteries in parallel?
 They need the same full charge at the start?
 Do they're discharge rate the same in most instances?

Only if they're exactly matched they will discharge at the same rate. 10 cycles from that balanced state, one battery will pull more and this will only worsen over time, ruining the weakest battery.

But... Time for a reality check: How long do you need it to go on a single charge?
(To compare, most robot races are over in a matter of minutes, so you don't wanna drag around a battery heavier than needed).
It's better to have a spare battery, so you can charge while running.

[nottoooily]

I thought it was pretty safe to parallel batteries as long as they're the same chemistry (ie nominal voltage), regardless of capacity. Well that's for lead acid, maybe it somehow doesn't work here.

Series is where you really have to be careful, that's where a single bad cell can get dragged down by the others. But in parallel if one battery is weaker, the other will pull it up because they're forced to be at the same voltage. Again, maybe only applies to lead-acid ??

Another caution is the amp-hour rating is usually defined for a 20 hour discharge. The faster you discharge, the less capacity you get.

[Soeren]

Hi,

I thought it was pretty safe to parallel batteries as long as they're the same chemistry (ie nominal voltage), regardless of capacity. Well that's for lead acid, maybe it somehow doesn't work here.

Each battery is build by a number of cells (re. your description of problems with serial connections) and while new will work well, but each charge/discharge cycle will emphasize what was initially minute differences and this accelerates, since the weakest cells will weaken more than the strong cells.

But in parallel if one battery is weaker, the other will pull it up because they're forced to be at the same voltage.

Depends on the viewing angle, you might say that the weaker cell will pull the other one down.

Another caution is the amp-hour rating is usually defined for a 20 hour discharge. The faster you discharge, the less capacity you get.

Yes, that applies to all known chemistries.

What's even worse, many rechargeables doesn't hold their nominal capacity (neither on first charge, third, fifth or any other). To be absolutely sure of the capacity, test them yourself and if possible, do it with a current that mimics the intended use, to see how well they hold up.

[nottoooily]

Each battery is build by a number of cells (re. your description of problems with serial connections) and while new will work well, but each charge/discharge cycle will emphasize what was initially minute differences and this accelerates, since the weakest cells will weaken more than the strong cells.

Yep that's what I mean. When they're made, they match them up pretty well. But I've found that when SLA's do fail, it's just one cell that's gone totally dead or into reverse voltage, while the others are fine.

You could consider a cell to be two different capacity cells paralleled. They just happen to be joined at the hip. The higher capacity one will deliver more current, so they both get loaded in a way that suits them.

But in parallel if one battery is weaker, the other will pull it up because they're forced to be at the same voltage.

Depends on the viewing angle, you might say that the weaker cell will pull the other one down.

Yep but either way they the bad one doesn't get made worse, so it's stable. Unlike series which unstable.