• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Robotics Application Needs Battery or Heat Source

Status
Not open for further replies.

gutsandgears

New Member
Hey guys, I'm building a small robot. I'm printing pieces from a 3D printer and using small pieces of nitinol wire as the actuators. The robot will have everything self-contained- actuators, power supply, microprocessor and pcb, wifi transciever, accelerometer, etc. It is planned to be around the size of an action figure: 6-7" tall.

Obviously the most important part is getting the thing to move. Nitinol allows me to create a robot that is smaller, more compact, and with more realistic movements than using any kind of DC motor (please don't say servo motor to me lol). The nitinol pieces are either .75 mm or 1 mm thick and typically a little less than an inch long. The nitinol seems to be able to reliably move the light plastic pieces I'm getting from the 3D printer, but I have one large problem: POWER!

It's like this:
For those of you who are unfamiliar with nitinol, it is called a shape-memory alloy. When heated (in this case 80 degrees F), it will change shape. That doesn't sound like a lot of heat, but when you factor in how quickly I want this to move (it can move almost instantaneously if you immerse it in enough heat- i.e. more than the transition temperature of 80F), and the fact that I'm not heating the nitinol directly, I really probably need around 300-400F.

You can run current through the nitinol itself, it's natural resistance will heat it up, and it will change shape. However, unless you are using pieces with diameters the size of a human hair, it is difficult to get enough current to heat it quickly (especially with a battery- the power supply this project calls for). So, what I'm currently using is nichrome wire- the stuff they put in old heating elements. It seems like I can get a huge amount of energy savings by using 40 gauge nichrome wire. And indeed, when wrapped around the nitinol and heated with some current, it does cause the nitinol to actuate.

However, I cannot get the nichrome to heat up fast enough for the robot to have crisp, quick movements (like I know nitinol is capable of).

Initially I was using a 9v battery just to get my feet wet. When I was 'ready' I got two lithium ion batteries- both 3.7v, one 1500 mA, the other 2500 mA. I figured they would absolutely fry the nichrome wire. Surprisingly, they are not. The thing moved faster with the 9v attached. It seems that the 3.7v just won't push the current fast enough (however, I admittedly don't know all that much about Electrical Engineering and this may not be the case at all), and I can't find any other voltage for lithium ion or lithium polymer batteries that aren't huge in size.

I need help! I need other ideas of how to generate a high amount of heat (flicking a lighter on underneath the nitinol makes it move fast enough; just to give you an idea of the kind of burst of heat I need), in a small space, very quickly.

Or

I need someone to point me in the direction of a power source/battery that is approximately the same size as a lithium ion battery (i.e. I can't use the kind of battery an RC car uses, or use a wall-mounted power supply, it has to fit inside the robot, which is approximately the size of an action figure, so that it can be wireless).

So basically, is there anything better than nichrome? Is there some sort of insulation I could use to amplify the heat I already have? Am I missing something with the batteries? Would using a wire wound resistor instead of a small amount of bare nichrome generate more heat faster? Maybe there is a simple chemical reaction that would solve this problem? I've blown up a lot of electronic parts just tooling around, I never thought actually making heat would be a problem.

Thank you in advance for your replies, this project means a lot to me,
Brad
 

Pommie

Well-Known Member
Most Helpful Member
It looks like you've found the reason all other builders use DC motors.

Mike.
 

MikeMl

Well-Known Member
Most Helpful Member
Yeah, this is all dictated by the laws of thermodynamics. (hint: you cant fool Mother Nature).
 

alec_t

Well-Known Member
Most Helpful Member
Welcome to ETO.
As Mike says, thermodynamics are the key.
For a given mass of Nitinol wire you can calculate how much energy is needed to raise its temperature to some target value in a target time.
You can't escape the need for that energy, however you heat the wire, and that energy must come from the battery. Power = energy/time, so the quicker the heating the greater the power requirement. Would you risk melting your 3D-printed parts with the heat?
Something else that would need to be considered is cooling. How would you achieve rapid cooling if you wanted rapid contraction of the Nitinol?
 

ClydeCrashKop

Well-Known Member
Most Helpful Member
Let your robot pull a battery in a wagon behind him. It will give him something to do.
 

cowboybob

Well-Known Member
Most Helpful Member
From a Wikipedia article:
"... The fastest nitinol actuation recorded was carried by a high voltage capacitor discharge which heated an SMA wire in a manner of microseconds, and resulted in a complete phase transformation (and high velocities) in a few milliseconds. ...". (My empasis)

This got me to wondering if you could use an SMPS (Switched Mode Power Supply) boost IC (such as one of these) circuit to charge a so called "super" cap bank (2 - 5.5VDC caps in series (examples).

The idea is to get the Li battery's 3.7VDC voltage up to 11VDC (or so) across the cap bank that can unload a lot of current quickly into the SMA referenced in the above quote OR, perhaps, as a direct shot to the nitinol wire itself.

You'd have to (obviously) "play" with the appropriate voltage and current levels (as well as discharge times) to arrive at the movement(s) that suit your needs.

Keeping in mind Post #4 above (MikeMI's), this won't reduce the total power (I times E) needed, but it should give you better reaction results from the power source(s) you have available.
 
Last edited:

MikeMl

Well-Known Member
Most Helpful Member
...to charge a so called "super" cap bank (2 - 5.5VDC caps in series (examples).
....
The Chinglish Data sheet is hard to interpret regarding ESR and what is the peak discharge current available from these types of caps?
 

cowboybob

Well-Known Member
Most Helpful Member
... The Chinglish Data sheet is hard to interpret regarding ESR and what is the peak discharge current available from these types of caps? ...
Hey, Mike. Here's one example of a supercap explanation. ESR is addressed. There are also example circuits that utilize these devices.

Brad, I've used Super caps (typically 1F and up) and they are very good for short (depending, up to 500mS), high current (up to 1+ amp) bursts, at or near supply Vcc when backed by a continuously operating boost charger (like a "joule thief") powered by "small" energy sources. This process does deplete the battery quicker but optimum Vcc levels are maintained for a longer period of time.

And, a self heating tininol wire "muscle" circuit concept could be adapted to your application.

As shown, power would be fed directly to the 40 gauge nichrome wire through an appropriate switching MOSFET transistor, that could be PWM controlled by your μC, to actuate the "muscle".

My suggestion is this arrangement, i.e., battery, boost SMPS and Supercap(s), providing power to the MOSFET (under control of the PWM circuit) feeding power to the nichrome wire. I think there would be sufficient voltage and current to create a more vigorous, natural, and controllable movement`AND allow proportional and/or maintained movement position from the action of the tininol wire. This could then mimic a servo.

It would be helpful if you would let us know a little more about how much current you need per contraction at the speeds you hope to achieve. Empirical data is preferred, i.e., "well, at 9VDC, 1A for 1/4Sec suits me".
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top