PDA

View Full Version : [Question(s)] Undervolting motors and current-draw?



Adrenalynn
05-07-2008, 01:53 PM
You ever start whiteboarding a robot and realize that you're in faaar over your head?

Generally I run some quick numbers, triple them, then double that "for safety". That starts to become a challenge when one starts thinking about 250lb robots and you have to start accounting for gravity, rolling friction, static friction, tire displacement, ...

Anywho - on to the question:

If you undervolt a motor do you increase the current draw inherently?

I find myself wondering. On the one hand, the answer seems an obvious "duh!" and "Ohm hates you for even asking"... On the other hand, we're decreasing the inductance which decreases the current draw and also decreases the resistance (due to heating) which also decreases the current draw. So we get an "obvious" "No" and "Ohm hates you for even asking"...

I can compute all my horsepower and torque numbers, and backend load my final drive ratio and solve my motor sizing and gearing questions, but undervolting does raise questions I can't answer today regarding motor controller selection as well as tradeoffs of battery voltage versus battery weight.

I think I'm leaning towards "yes, it increases the current draw" - but these things are expensive. Anyone out there with experience in ~300lb robots on the "cheap"?

sthmck
05-07-2008, 02:33 PM
Hey,
I dont know the answer off hand but I am pretty sure you could find it out if you went to a combat robot forum. I'm not sure if it is a sin to suggest something like that but oh well I did it. Lol

Adrenalynn
05-07-2008, 02:47 PM
Worth a shot, thanks.

I keep thinking I know the answer, arguing with myself, getting fed up, and walking away from myself.

If this keeps up it's gonna be another one of those times I get ticked off at myself and not talk to myself for a week...

Those combat'bot guys always like overvolting - their heads would probably explode if I asked about undervolting...

You know... What if I just tested the darned theory by undervolting a 12v motor at 6v and measuring current? ;)

Matt
05-07-2008, 02:55 PM
This so reminds me of those philosophical questions where both answers could be right. Like the one with a plane trying to take off on a treadmill... Why am I not surprised that Jodie is the one who has managed to find a robot version of one to twist herself into a mental knot over? :)

LinuxGuy
05-07-2008, 03:10 PM
This so reminds me of those philosophical questions where both answers could be right. Like the one with a plane trying to take off on a treadmill... Why am I not surprised that Jodie is the one who has managed to find a robot version of one to twist herself into a mental knot over? :)
Too funny! It sounds like something I would come up with.. :veryhappy::veryhappy: I knew there was a reason I like her. :happy:

8-Dale

Adrenalynn
05-07-2008, 03:24 PM
This so reminds me of those philosophical questions where both answers could be right. Like the one with a plane trying to take off on a treadmill... Why am I not surprised that Jodie is the one who has managed to find a robot version of one to twist herself into a mental knot over? :)


What, are you accusing me of overthinking problems? :mad: Dang - know me better than I thought. :p

Seriously though - four 24v 30A motor controllers vs four 36v 30A motor controllers vs four 24v 60A motor controllers - it does make a rather substantial difference. Then add in another 60lbs of batteries costing another couple hundred dollars, and frame to support 'em, as well as the torque needed to move 'em, and it really does start to add up...

[edit to add: If Alex gets the knowledgebase online at some point, I'll do a write-up [after our white-paper] on how to calculate torque/hp requirements for robots. ;) ]

Adrenalynn
05-07-2008, 03:44 PM
Thanks, Robo'guy - I think...:p

You might just be a [accidental] genius, Matt.:p

Just thinking logically for a moment against what you posted:

What if the operating current comes out about the same (HP = Volts, Torque = Amps basically), but the start-up (force-moment) and stall current are higher? How do I figure, you ask? Glad you did[n't]:

The windings on a 36v motor would be heavier - thicker gauge. Resistance across the windings should be consistant at about one ohm (ignoring impedance here). A stalled motor is effectively a short-circuit. A modern high output gell battery should deliver >250A into a short circuit (generally around 320A rated for an 18Ah). If the windings are heavier gauge then it'd pull more current through the wire before the magic smoke escaped.

Ok - so now it's about burst current ratings instead of operating ratings - and about throwing some massive fuses in there to deliver enough juice to get it rolling (overcoming static friction, drive-train loss, etc) but still pop reliably if the motor is stalled. And, of course, about choosing a motor controller that can deliver short bursts in the 100A range but run at around 35A...

I'm sure I'll disagree with myself again in a minute. And shouldn't I really be working instead of worrying about this stuff?!?!

LinuxGuy
05-07-2008, 04:17 PM
I'm sure I'll disagree with myself again in a minute. And shouldn't I really be working instead of worrying about this stuff?!?!
You sound like a Gemini.. :happy:

8-Dale

Adrenalynn
05-07-2008, 04:29 PM
Capricorn, actually...

Droid Works
05-07-2008, 04:42 PM
It seems to me it all depends on what you are using. What kind of motors batteries and controller are you using and how many of each?

Adrenalynn
05-07-2008, 04:47 PM
Pretend there is no controller. Pretend there is no battery.

What a motor draws is what a motor draws. Controllers and batteries are sized from the motor, and the motor is sized from the environment. (weight/speed requirement/worst-cases)

Droid Works
05-07-2008, 04:56 PM
But not all motors are crated equal..lol Different motors behave differently under the same conditions, much like a guitar pickup. For example: 2 startocaster pickups both with the same design and same amount of wire wraped arround the coil, but one is hand wound and one is machine wound. though they are the same they are 2 entirely different pickups with entirely different behavior. I know guitar pickups are not motors but the principle is the same. A motor may or may not behave the the way you expect it to depending on the construction and design. also if I know what you are using the may be a cheaper alternative.

Adrenalynn
05-07-2008, 05:09 PM
See - that's the question. I have an environment, and I have a weight/velocity requirement.

Still - if you put a 100A motor on a 5A motor controller, it still draws 100A - until the magic smoke escapes from the motor controller... ;)

Droid Works
05-07-2008, 05:12 PM
See - that's the question. I have an environment, and I have a weight/velocity requirement.

Still - if you put a 100A motor on a 5A motor controller, it still draws 100A - until the magic smoke escapes from the motor controller... ;)
But why would you do that? It wouldn't cost much to resolve the problem.

Adrenalynn
05-07-2008, 05:37 PM
Yes.

But I have to source my motors, then source my controllers, then source my batteries.

The question was related to sourcing the controllers. I found 36v motors that will do the job, but I want to drive them at 24v. So now I need to know if decreasing the voltage will increase the current _through a motor_. Normal Ohm's Law application would say "yes". It would also say "no". If it increases the current 33%, it makes a lot of difference when we're talking about $750 x 4 motor controllers...

When I wrote my previous reply, I was on the phone with a physcist friend of mine. After 20mins of debate, he admited that: "that's a real head-scratcher"... :)

I hate questions I can't calculate. The next step has to be putting a smaller motor on a variable bench supply and measuring current draw - otherwise this question, it appears, will remain unanswered...

LinuxGuy
05-07-2008, 05:47 PM
The question was related to sourcing the controllers. I found 36v motors that will do the job, but I want to drive them at 24v. So now I need to know if decreasing the voltage will increase the current _through a motor_.
A motor should never draw more current than it needs, unless there is some kind of problem. A motor draws less when it is not stalled than it does when stalled. It shouldn't draw more than required.

8-Dale

Adrenalynn
05-07-2008, 06:18 PM
Almost...

A motor is stalled every time you attempt to take off from a stand-still. That's the force-moment. It's also known as "start-up current". "start-up current" is for zero load. Overcoming static friction, drive-train loss, mass-moment, etc requires current approaching stall-current. How close depends upon those factors multiplied by the efficiency of the motor.

metaform3d
05-07-2008, 06:35 PM
Try here (http://www.motortech.com/BULL_E-1.htm). The money quote is at the end:

"Apply twice the voltage and you get twice the stall current - half the voltage and you get half the stall current. Why? Because when it is not rotating (e.g. stalled) the armature appears in the circuit as a resistor."

Adrenalynn
05-07-2008, 06:42 PM
That whole TB is flat-out GOLD, Meta. I agree it appears as a resistor - something less than 1ohm. I need to ingest that TB and follow the numbers [after a little nap] - but regardless - it sounds totally believable [and matches one side of my argument with myself ;) ].

You just made my entire evening. :o

metaform3d
05-07-2008, 07:21 PM
Glad I could help!

Adrenalynn
05-07-2008, 09:51 PM
Now that I've had a couple hours nap and a chance to ingest that document, Meta, I can reconfirm: It's money from the first word to the last.

I'm not one for blindly bolting sh..tuff together and not understanding what I'm doing it for. That kinda mini-whitepaper is priceless to me. I owe ya!

Adrenalynn
05-07-2008, 09:57 PM
And so it begins. 4 x ~1.60922651 HP ( ;) ) motors purchased...

Wingzero01w
05-07-2008, 10:19 PM
Nice start Adrenalynn, but if you don't mind me asking- what exactly are you making? :)

Adrenalynn
05-08-2008, 11:01 AM
"Oscar"...

He's a 96 gallon trashcan. (A fully autonomous, people-interacting, self-emptying, bohemoth with an estimated payload capacity of around 200lbs on a 'bot platform closing on 200lbs, with about two hours of runtime or two days of loiter. He'll have about 4.8Ghz of onboard processing power, with 2GB of RAM, and 500GB of triple shock-insulated storage.)

Or at least that what's the 6'x4' of equations on my design whiteboard says...

Now I need to go source my 6+ motor controllers... ;)