micro power switch with variable resistance

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
I'm searching for a micro power switch with variable resistance (from 0 to infinite).
Does anyone know where I can find one?
 
I'm searching for a micro power switch with variable resistance (from 0 to infinite).
Does anyone know where I can find one?

What current do you plan to run through it?
What voltage do you plan to run through it?
What size?
What lifetime (number of cycles)?

You appear to be after a linear potentiometer. What you are after (micro size) VERY likely does not exist.

cheers,
george.
 
Thanks georges80 for your help!

> What current do you plan to run through it?
less than 5A

> What voltage do you plan to run through it?
3-4V

> What size?
the smallest

> What lifetime (number of cycles)?
? number of cycles? does it really works with limited number of cycles?

> You appear to be after a linear potentiometer.

Thanks for the name!

> What you are after (micro size) VERY likely does not exist.

Ah. Isn't it a common need for flashlights?
I need it to donate variable intensity to my custom cell.phone/flashlight hybrid: http://www.candlepowerforums.com/vb/showthread.php?p=2210021
 
check out force sensors. very small, but no current capability. to be used for controlling a converter. digikey has some, lusense is another brand.
 
check out force sensors. very small, but no current capability. to be used for controlling a converter. digikey has some, lusense is another brand.

Hi AilSnail,
Can I ask you what do you mean with "no current capability"? can they be used for controlling the brightness intensity of LEDs?
 
Hi AilSnail,
Can I ask you what do you mean with "no current capability"? can they be used for controlling the brightness intensity of LEDs?

Very simple - you will NOT find a variable resistor that is linear in action or rotary that is both miniature AND able to carry any current more than a few mA.

That's why there's something called ELECTRONICS :)

You use a sensor/switch etc and some electronics like big transistors or FETs to carry the current.

Variable intensity is done with electronics and uses things like PWM and/or variable current. There's a reason a driver like nFlex or bFlex or maxFlex are the size they are to provide variable brightness... If it was as easy as using a simple variable resistor...

Any YES, potentiometers have limited lifetimes as do switches. They are mechanical and will wear out and even faster with higher currents.

cheers,
george.
 
About current capability: What georges80 said - they will not stand the current directly and besides, the ones I know of has way too high resistance for that.

I don't know what is the size of nFlex or bFlex or maxFlex, but many standard IC converters, with only a very few and small external components, adjust voltage* by two, or even one resistor. In the place of that resistor you can put a force sensor, they can be really small, less than a mm thick.
So the cirquit can be made really small, I am putting one into an arc AAA now, and found that even with a thick copper sink and a huge inductor I had more space, so I am waiting for some tiny smd 0402/0603 battery indicator LEDs. It is a bit of a mechanical challenge though and I am not sure yet if it will work inside the light - but the test cirquit does.
I don't recall the cycle life but it wasn't horrible I think. The lusense ones had a few cm leads that I could not cut off since I was unable to connect reliably to the stumps. No problem with the ones digikey carries (solderable leads), but those are a bit more fragile. The black antistatic foam used when packing IC's also has the property of reducing resistance when compressed, but it has a sort of shape memory (won't expand fully after severely compressed) so I think it is not suitable.

For my application I could otherwise use some tiny reed switches (reacts on/off to magnetism), or maybe a linear hall effect sensor (reacts continously to magnetism), though I didn't find a suitable one yet.

*I don't need current control, current will be limited by the converter itself or by a PTC thermal resistor in series with the force sensor. Besides, I size the resistors for the specific LED's Vf.
 
Last edited:
About current capability: What georges80 said - they will not stand the current directly and besides, the ones I know of has way too high resistance for that.

I don't know what is the size of nFlex or bFlex or maxFlex, but many standard IC converters, with only a very few and small external components, adjust voltage* by two, or even one resistor. In the place of that resistor you can put a force sensor, they can be really small, less than a mm thick.
So the cirquit can be made really small, I am putting one into an arc AAA now, and found that even with a thick copper sink and a huge inductor I had more space, so I am waiting for some tiny smd 0402/0603 battery indicator LEDs. It is a bit of a mechanical challenge though and I am not sure yet if it will work inside the light - but the test cirquit does.
I don't recall the cycle life but it wasn't horrible I think. The lusense ones had a few cm leads that I could not cut off since I was unable to connect reliably to the stumps. No problem with the ones digikey carries (solderable leads), but those are a bit more fragile. The black antistatic foam used when packing IC's also has the property of reducing resistance when compressed, but it has a sort of shape memory (won't expand fully after severely compressed) so I think it is not suitable.

For my application I could otherwise use some tiny reed switches (reacts on/off to magnetism), or maybe a linear hall effect sensor (reacts continously to magnetism), though I didn't find a suitable one yet.

*I don't need current control, current will be limited by the converter itself or by a PTC thermal resistor in series with the force sensor. Besides, I size the resistors for the specific LED's Vf.

Could you please post/link to some pictures of the components you mentioned?
 
Remember resistors waste energy. A better solution is to use a driver circuit with a variable duty cycle to control brightness. You can find examples of simple driver circuits in Radio Shack handbooks and electronic handbooks from various manufacturers. Just Google and you will find many examples also.
 
Last edited:
All of the above is true, but there is a little known 'device' that might do just what you want! its called a "Quantum Tunnelling Compound Pill" there cheap its basically a small bit of foam that changes resistance as you squash it, they can handle a fair amount of current also, Ive known of them for a while but not had chance to 'play' with them yet. heres a couple of links:
http://www.maplin.co.uk/module.aspx?ModuleNo=44202&doy=20m11#overview
http://www.peratech.co.uk/pills.htm

it may be that it does not give the desired 'control' of the LED due the the sharp VI curve of LEDs (i.e a small change in voltage causes a large change in current, i.e. without current control it can become very on off) but I would definitely suggest trying it -its cheap and easy even for a beginner, where as the circuits described above can be very tricky!
 
Those would make for a real "hot finger" if used more than a second or two. Plus they inherently waste energy just like any regular resistor. All that energy is turned to heat at the point of force application.
 
I'm searching for a micro power switch with variable resistance (from 0 to infinite).
Does anyone know where I can find one?

If by "micro power switch with variable resistance" you mean pot, then I can't help you. I've been looking for the same thing for quite a while now.
 
All of the above is true, but there is a little known 'device' that might do just what you want! its called a "Quantum Tunnelling Compound Pill" there cheap its basically a small bit of foam that changes resistance as you squash it, they can handle a fair amount of current also, Ive known of them for a while but not had chance to 'play' with them yet. heres a couple of links:
http://www.maplin.co.uk/module.aspx?ModuleNo=44202&doy=20m11#overview
http://www.peratech.co.uk/pills.htm

it may be that it does not give the desired 'control' of the LED due the the sharp VI curve of LEDs (i.e a small change in voltage causes a large change in current, i.e. without current control it can become very on off) but I would definitely suggest trying it -its cheap and easy even for a beginner, where as the circuits described above can be very tricky!

Ah, wonderful!! I want to try it!

"Typical Resistance Range> 10 12 Ohms to < 1 OhmOperating Voltage0 to 40 VMax Current10 A"


If I only find where to buy them from Italy!...
As soon as I find them I buy them and will try and report to you my experience with them.
 
You do know the Vf of a LED changes slowly over time, right? :D

oO

Yes, but all it will do is change the physical relative position of the head and tail where maximum light is given. And possibly move the minimum point a few mA up. I think the problem with these sensors is more that the movement input is linear, and must be "converted" to force by a cushion/spring in top of the sensor. It seems like adjustment can become inaccurate with jumps, and probably it will flicker when the light is shaken. Got the 0402 leds: I can't believe so much light comes from a device so small I can barely see it!
 
"Could you please post/link to some pictures of the components you mentioned?"

sorry, they are too small for my mobile cam!
Do a search on digikey for the sensors, resistors, leds, ceramic caps,
www.ti.com for tps61020 (the cirquit is in the datasheet, works with or without the loop compensation capacitors described there - I am not able to measure efficiency with my poor old multimeter)
The circular sensor with a rubber cap I don't know where to get, because digikey has a large minimum order for those. I have only one, and it is JUST the right dia for the AAA.
check here for newbie/jarhead's maglite solitair with a 61020 inside - he required a delay cirquit since the 61020 does not give full current until the set voltage out is reached - and it can not reach that voltage without full current! so he made an attack curve on the voltage setting by putting a cap and a transistor in the feedback resistor network, IIRC.

The twist-adjustment makes the delay unnecessary.
There are many other little chips can be used, there was a thread called "fighting with efficiency - eureka!" that listed many. Jarheads pages also shows many. http://www.molalla.com/~leeper/led.htm

thanks for the pill link.
 
Top