Author Topic: How Power Consumption Works  (Read 6111 times)

Offline Clickbeetle

  • *
  • Posts: 3323
  • Reputation: 21
  • In Soviet Russia, bugs stomp YOU!
    • View Profile
    • Beetle Bros site
How Power Consumption Works
« on: October 17, 2009, 11:14:59 PM »
So, we all know that power consumption in RA2 is really weird.  There's the old adage: "Battery output == motor input for burst motors, battery output == half motor input for spin motors."  But is that really true?  How exactly does power consumption work?

I did some research so I could balance the batteries better for DSL3, and I'm not sure I like what I found.  No, wait, I'm 100% certain I don't like what I found.

First, a list of abbreviations I'll be using here.

MEMIOR == motor elecMaxInOutRate
BEMIOR == battery elecmaxinoutrate
PU == power units

The first thing I did is find out how using multiple batteries vs. a single battery works.  Everyone has always assumed that total power and BEMIOR are additive, but no one has really tested it.

I found that total power and BEMIOR are indeed additive.  So 1 battery with 1000 total power and 100 BEMIOR is the same as 2 batteries with 500 total power and 50 BEMIOR.

Then I moved on to some power consumption tests.

The most important thing that I found is this:  Spin motor MEMIOR has no effect whatsoever on battery drain.  That's right, folks.  Doesn't matter if it's a reverse angle motor, an HP Z-tek, a Dual Perm, or even a custom "gasoline" motor with -0.0001 MEMIOR.  They all drain the same amount of electricity.

With infinite BEMIOR, the battery drain is equal to 100 PU/second per spin motor.  So if you have a bot with 2 motors and a battery with 2000 total power, your battery drain will be 200 PU/second and your batteries will last exactly 10 seconds.

With finite BEMIOR, things get more complicated.  Perhaps it is easiest if I just show a graph here.



As you can see, with a lower BEMIOR, the battery drains more slowly, in a more exponential manner.  The bigger BEMIOR gets, the closer battery drain gets to a constant 100 PU/second.  This data is with a battery with 1000 total power.

Now you might look at this and say, "Whoa!  I should use batteries with as little BEMIOR as possible so my batteries last longer!"  While it's true that less BEMIOR means longer lasting batteries, there are advantages to having high BEMIOR as well.

The higher BEMIOR is, the longer your spin motors stay at full power.  Also, you need 100 BEMIOR per motor in order to achieve full power in the first place.  So if you have two spin motors, you need a minimum of 200 BEMIOR to get full power out of them.  However, at this minimum value, they will quickly start slowing down as your batteries drain.  Using overkill BEMIOR keeps your motors running at max speed for longer.

Practical example:  A typical HW HS with 6 HP Z-teks needs 600 BEMIOR for full power, and a typical LW HS with 2 angle motor drive and 1 HP Z-tek needs 300 BEMIOR for full power.  So in theory, you can power Absolute Chaos with 6 small battery packs.  The only problem is that they have low total power, so it's going to slow down very fast.

So what about burst motors?  We already know they're different from spin motors, but how exactly do they work?

For burst motors, I found that MEMIOR does have an effect on power consumption.  So unlike spin motors, a DDT is going to drain more power than a Snapper2.

I started off testing on a battery with infinite BEMIOR again.  With infinite BEMIOR, a burst motor will drain a number of power units equal to its MEMIOR/1.6 each time it is fired.  Why divide the MEMIOR by 1.6?  I have no clue, that's just how it works.  So a DDT with -500 MEMIOR will actually drain 312.5 PU each time it is fired.

With finite BEMIOR, burst motors behave much like spin motors.  The smaller BEMIOR is, the slower your batteries will drain, but the burst motors will stay at full power for less time.  Also, you need BEMIOR equal to MEMIOR/1.6 per burst motor to achieve full power.  Like spin motors, using more BEMIOR than necessary keeps the motors at full power for longer.

Practical example:  For a typical MW popup with 2 HP Z-tek drive and 3 DDT's, you need 100 (x2) + 312.5 (x3) = 1137.5 total BEMIOR for max power, which is about 1 Supervolt and 1 Nifty.  For a typical HW popup with 4 HP Z-tek drive and 4 DDT's, you need 100 (x4) + 312.5 (x4) = 1650 total BEMIOR, which is about 2 Supervolts.  Remember that these are minimum values, however, and your motors may start slowing down depending on how much total power you have.

So what does all this mean for DSL3?  It means it is impossible to balance spin motor battery drain fairly.  The Dual Perm will always drain the same amount of power as a Piglet.  This explains why the ant batteries are so effective in DSL2.

I don't know what I'm going to do as far as battery/motor balancing goes at this point.  It's going to take some thought.  I would have never expected power consumption to work like this.

To lack feeling is to be dead, but to act on every feeling is to be a child.
-Brandon Sanderson, The Way of Kings

Offline JoeBlo

  • *
  • Posts: 14539
  • Reputation: 13
    • View Profile
How Power Consumption Works
« Reply #1 on: October 17, 2009, 11:27:29 PM »
wow great work click, this may prove useful info for my upcoming mod too
14x BOTM Winner - Robot Arena for Mac

Offline Sage

  • *
  • Posts: 6031
  • Reputation: 10
  • RA2 Wizard & GTM's Favorite Stock Builder 2015
    • View Profile
How Power Consumption Works
« Reply #2 on: October 17, 2009, 11:37:49 PM »
Very nice work Click.


As for your problem with balancing batteries and motors, i think making a larger sized battery more powerful than its weight it smaller batteries would do the trick (like it is in stock. I mean, it works in stock. Even though small battery packs would last in a 36HS, they are not powerful enough to spin the motors fast enough.)
You got my vote for RA2 Wizard. Always and forever.

Offline 123savethewhales

  • *
  • Posts: 2771
  • Reputation: 30
  • Friendship is Magic
    • View Profile
How Power Consumption Works
« Reply #3 on: October 17, 2009, 11:50:59 PM »
Thanks a lot.  This explains why I seem to run my HW just fine with 5 ant batteries but nearly run out on a LW with just 3.  I know something was wrong but I can't pinpoint it exactly.

Quick question, so this works as some kind of breakpoint?  If I have 15 amp per motor it follows the 10 amp chart and use reduced consumption at 750?  Or would the line be somewhere between 10 and 20?
« Last Edit: October 18, 2009, 12:38:12 AM by 123savethewhales »

Offline Clickbeetle

  • *
  • Posts: 3323
  • Reputation: 21
  • In Soviet Russia, bugs stomp YOU!
    • View Profile
    • Beetle Bros site
How Power Consumption Works
« Reply #4 on: October 18, 2009, 12:22:32 AM »
Quote from: Sage;73543
As for your problem with balancing batteries and motors, i think making a larger sized battery more powerful than its weight it smaller batteries would do the trick (like it is in stock. I mean, it works in stock. Even though small battery packs would last in a 36HS, they are not powerful enough to spin the motors fast enough.)

Good idea.  Not a perfect solution, but it will do.

I'm thinking about incorporating size factors into some other component stats as well.

Quote from: 123savethewhales;73547
Now I just have to find the formula to solve for the least amount of batteries use per motor to remain at full power for 3 minutes.

Good luck with that... I spent maybe an hour or two in Excel playing with formulas, trying to make a model for power consumption.  I never did get one that works for all values of BEMIOR.

Quote from: 123savethewhales;73547
Quick question, so this works as some kind of breakpoint?  If I have 15 amp per motor it follows the 10 amp chart?  Or would the line be somewhere between 10 and 20?

The graph for battery consumption depends on total BEMIOR.  If you are supplying 15 amps per motor and you have 2 motors, then you have 30 total amps and your graph will be between the 200 BEMIOR line and the infinite BEMIOR line.

Of course, you'll have much more than 1000 total power on a real bot, so your graph might look more like the 50 BEMIOR line on a greater electotal scale.  I just used 1000 total power for my tests because I didn't want to sit for 3 minutes waiting for a regular battery to run out.

To lack feeling is to be dead, but to act on every feeling is to be a child.
-Brandon Sanderson, The Way of Kings

Offline trumpetguy

  • Super Heavyweight
  • Posts: 1194
  • Reputation: 0
  • Forum Ameri-asian
    • http://www.facebook.com/DC.Scinerd
    • http://www.youtube.com/user/PokeDenny
    • View Profile
  • Skype: dennycao202
  • Xfire: scinerd
How Power Consumption Works
« Reply #5 on: October 18, 2009, 12:46:12 AM »
How do you find out how much battery do you need for a SNS, since
A. They don't move around a lot. and
B. Their motors constantly go the same way.

"Two things are infinite:the universe and human stupidity;and I\'m not sure about the universe" -Albert Einstein
*IF YOU ARE NEUTRAL IN THIS SIG BUSINESS PUT THIS IN YOUR SIG*

Offline Clickbeetle

  • *
  • Posts: 3323
  • Reputation: 21
  • In Soviet Russia, bugs stomp YOU!
    • View Profile
    • Beetle Bros site
How Power Consumption Works
« Reply #6 on: October 18, 2009, 12:49:50 AM »
Quote from: trumpetguy;73559
How do you find out how much battery do you need for a SNS, since
A. They don't move around a lot. and
B. Their motors constantly go the same way.

Same way as any other bot.  SnS's are actually easy because of those things you mentioned.

Each drive motor needs 100 power.  So a 4WD SnS needs 400 power, which is 1 Nifty.  2WD needs 200, or 2 small battery packs, and 6WD needs 600, or 1 Nifty plus 2 small battery packs (might as well use 1 Supervolt).

To lack feeling is to be dead, but to act on every feeling is to be a child.
-Brandon Sanderson, The Way of Kings

Offline 123savethewhales

  • *
  • Posts: 2771
  • Reputation: 30
  • Friendship is Magic
    • View Profile
How Power Consumption Works
« Reply #7 on: October 18, 2009, 02:16:12 AM »
Quote from: Clickbeetle;73555
Of course, you'll have much more than 1000 total power on a real bot, so your graph might look more like the 50 BEMIOR line on a greater electotal scale.  I just used 1000 total power for my tests because I didn't want to sit for 3 minutes waiting for a regular battery to run out.

Total bemior?  Okay for a moment I thought it was 20 amp per spin motor.

edit:  NM you are right.  It can slow down anywhere and has nothing to do with the scales on top.  As for exactly how much it is much more difficult to say.

I think this finding will really affect LW, and to some extend MW.  HW probably has way more than enough batteries as is.
« Last Edit: October 18, 2009, 02:47:19 AM by 123savethewhales »

Offline Naryar

  • *
  • Posts: 21983
  • Reputation: 19
  • You are already defeated.
    • http://www.youtube.com/us
    • View Profile
  • Skype: TheMightyNaryar
How Power Consumption Works
« Reply #8 on: October 18, 2009, 03:00:18 AM »
Hey kids ! Time to spam ant batts on DSL popups as well !

Offline Avalanche

  • Current DSL HW Champion.
  • Ultra Heavyweight
  • Posts: 3085
  • Reputation: -35
  • Better than the mods
    • View Profile
How Power Consumption Works
« Reply #9 on: October 18, 2009, 03:22:19 AM »
Good idea.
Reigning DSL HW Champion, with 2 victories so far: One over Ty4er and one over DSC.
Runner-Up in The Cybernetic Colosseum HW division.
Semi-finalist in RNP: Sword of Glory.

You don't have to like me. You don't have to respect me fully. But you do have to acknowledge what I've achieved.

I, for one, welcome our new Apanx overlord.

Offline 123savethewhales

  • *
  • Posts: 2771
  • Reputation: 30
  • Friendship is Magic
    • View Profile
How Power Consumption Works
« Reply #10 on: October 18, 2009, 03:55:51 AM »
Quote from: Naryar;73570
Hey kids ! Time to spam ant batts on DSL popups as well !


You are right, say goodbye to those silly redpacks.

Offline HurricaneAndrew

  • Ultra Heavyweight
  • Posts: 2978
  • Reputation: -13
  • Admon is not pleased with your shenanigans.
    • HurricaneAndrew92
    • HurricaneAndrewGTM
    • View Profile
  • Skype: hurricaneandrew92
How Power Consumption Works
« Reply #11 on: October 18, 2009, 04:30:33 AM »
Woah... So, basically what this proves is that the battery function in RA2 is completely unrealistsic and that everyones previous hypotheses about it were wrong?

#BlackNamesMatter  |  #HurricaneAndrewDidNothingWrong  |  #DemodNaryar  |  #OldfagAlliance

Offline Pwnator

  • *
  • Posts: 6617
  • Reputation: 14
    • View Profile
    • http://pwnator.tumblr.com
  • See profile for gamer tags: 1
How Power Consumption Works
« Reply #12 on: October 18, 2009, 04:36:37 AM »
Not really. Though misinterpreted, it does have its advantages. Our previous rules of thumb still applies when you want your bot to be fully powered.
Clash Cubes 1 - Grey Matter (Runner-Up)
King of Karnage - Sideshow Freak (Runner-Up, Best Engineered)
Rust In Pieces - Paper Cut 3 (Grand Champion, Most Dangerous Bot)
Wheely Tag Tournament - Ion Thruster (Grand Champion, along with Ounces' DiSemboweLment)
UK vs USA - Dark Striker (Grand Champion)
Rust In Pieces 2 - Claymore (Runner-Up, Favourite Bot)
BBEANS 6 - Infection 4 (Runner-Up)
RA2 Team Championships - Serious Business, Skeksis (Runner-Up, along with Scrappy, S_M, and Badnik)
RA2 Team Championships 2 - The Other Stig (Runner-Up, along with Scrappy, S_M, Badnik, 090901, and R1885)
Replica Wars 3 - Abaddon (Runner-Up, Luckiest Bot)
BroBots - wheebot & yaybot (Runner-Up)
Robo Zone 2 - Dipper (4th place, Survival Champion, & Best Axle Bot)
ARBBC - The Covenant (3rd place, BW Rumble Winner, Most Feared BW)

Offline 123savethewhales

  • *
  • Posts: 2771
  • Reputation: 30
  • Friendship is Magic
    • View Profile
How Power Consumption Works
« Reply #13 on: October 18, 2009, 04:55:10 AM »
Well, the rules of thumb seem to have changed from "ant only" for everything except those with burst motor, to simply "ant only" for everything.  Unless you are talking about some other rules of thumb that I am unaware of.

| Edit: Lol thanks for clarifying.
V

Offline Pwnator

  • *
  • Posts: 6617
  • Reputation: 14
    • View Profile
    • http://pwnator.tumblr.com
  • See profile for gamer tags: 1
How Power Consumption Works
« Reply #14 on: October 18, 2009, 04:59:04 AM »
I'm talking about Stock. The '1 black for 4 HP-Zteks' thingy. XD
Clash Cubes 1 - Grey Matter (Runner-Up)
King of Karnage - Sideshow Freak (Runner-Up, Best Engineered)
Rust In Pieces - Paper Cut 3 (Grand Champion, Most Dangerous Bot)
Wheely Tag Tournament - Ion Thruster (Grand Champion, along with Ounces' DiSemboweLment)
UK vs USA - Dark Striker (Grand Champion)
Rust In Pieces 2 - Claymore (Runner-Up, Favourite Bot)
BBEANS 6 - Infection 4 (Runner-Up)
RA2 Team Championships - Serious Business, Skeksis (Runner-Up, along with Scrappy, S_M, and Badnik)
RA2 Team Championships 2 - The Other Stig (Runner-Up, along with Scrappy, S_M, Badnik, 090901, and R1885)
Replica Wars 3 - Abaddon (Runner-Up, Luckiest Bot)
BroBots - wheebot & yaybot (Runner-Up)
Robo Zone 2 - Dipper (4th place, Survival Champion, & Best Axle Bot)
ARBBC - The Covenant (3rd place, BW Rumble Winner, Most Feared BW)

Offline Naryar

  • *
  • Posts: 21983
  • Reputation: 19
  • You are already defeated.
    • http://www.youtube.com/us
    • View Profile
  • Skype: TheMightyNaryar
How Power Consumption Works
« Reply #15 on: October 18, 2009, 05:41:51 AM »
Quote from: 123savethewhales;73588
Well, the rules of thumb seem to have changed from "ant only" for everything except those with burst motor, to simply "ant only" for everything.


Not exactly. Ant batts would be better for bursts if the bursts are not too powerful, but you can bet PC645's are still a good choice on Beta burst bots.

Offline roboman2444

  • Ultra Heavyweight
  • Posts: 1225
  • Reputation: 0
  • Linux, Nexuiz, Quake, and Darkplaces lover.
    • View Profile
    • http://www.freewebs.com/teamrckm
How Power Consumption Works
« Reply #16 on: October 18, 2009, 09:40:58 AM »
wow, so thats why 8 rad weped bots suck.
Real life robotics team www.teamrckm.tk
Real life game studio www.v2games.tk


Offline Scrap Daddy

  • *
  • Posts: 2654
  • Reputation: 18
    • scraaaaaap
    • View Profile
How Power Consumption Works
« Reply #17 on: October 18, 2009, 10:30:04 AM »
Quote from: roboman2444;73612
wow, so thats why 8 rad weped bots suck.


Maybe for YOU.

Well done, Click.

Offline Madiaba

  • *
  • Posts: 1438
  • Reputation: 8
    • View Profile
How Power Consumption Works
« Reply #18 on: October 18, 2009, 11:36:03 AM »
Because of the evidences I suspected something like this, but not to this magnitude.
Good research, CB. This info changes a few things....
« Last Edit: October 18, 2009, 12:05:48 PM by Madiaba »
Input is appreciated. :)
-Arrogance is a quantity devoid of quality...
-As a client once told me "This is my story, and it's sticking to me!"
-Relationships these days are like the 'Arrival' section of the airport: a lot of baggage is being revealed in one place, and not a lot of it is being correlated to its real owners...

Offline 123savethewhales

  • *
  • Posts: 2771
  • Reputation: 30
  • Friendship is Magic
    • View Profile
How Power Consumption Works
« Reply #19 on: October 18, 2009, 12:21:55 PM »
Quote from: Naryar;73590
Not exactly. Ant batts would be better for bursts if the bursts are not too powerful, but you can bet PC645's are still a good choice on Beta burst bots.


If I am interpreting Click's graph correctly, and if my field test is correct, 2 spring loaded beta with 2 drive will cost 282.5 amps, or 29 ant batteries.  Perhaps red can be use to deal with rounding issues as a replacement for 3 ants, but for every red you put in in replace of ants, it's going to hurt your max power duration.  I don't think it's worth it to go beyond 20 ants + 3 red.