Help with MOSFET circuit to control Muscle Wire

Hi all,

I am using a muscle wire based actuator that I am trying to control with an Arduino. The Arduino has 3.3V output from its I/O port.

The manufacturer told me that the muscle wire needs 3V - 3.6V and 600-700mA for 0.5 secs to actuate.

I am using lithium ion battery to power up the microcontroller. I am also using a NMOS to drive the muscle wire. One end of the muscle wire is connected to Vbatt (about 3.7V) and the other end is connected to the Drain of the NMOS. The Source of the NMOS is connect to ground and the Gate of the NMOS is connected to the MCU output pin. There is also a 10K pull down resistor connecting from Gate to Ground.

I have checked that the muscle wire is working properly when I connect it directly to the lipo battery terminals. I am having issues with selecting a suitable MOSFET for this application. I don't really know how to choose one. Right now I only have a MGSF1N02 and I tried connecting it up, but the muscle wire is not actuating. I replaced the muscle wire with a piezo buzzer and there is sound coming from the buzzer.

Anyone can advise? Thanks in advance!

If current is flowing trough the mosfet you got also voltage drop Vdrain-to-source on it,on datasheet i read 0.8 Volt, so your wuscle wire gets 2.9 Volt...maybe thats the reason?

I just tried using 3xAA batteries (4.5V) connected to one end of the muscle wire. But it is still not actuating.

I am sure the MOSFET is switching because I used a piezo buzzer to test and its giving a very loud sound. Muscle wire is still working too when connect directly to 3.3V supply.

boro3 said:

If current is flowing trough the mosfet you got also voltage drop Vdrain-to-source on it,on datasheet i read 0.8 Volt, so your wuscle wire gets 2.9 Volt...maybe thats the reason?

Click to expand...

Kian said:

I just tried using 3xAA batteries (4.5V) connected to one end of the muscle wire. But it is still not actuating.

I am sure the MOSFET is switching because I used a piezo buzzer to test and its giving a very loud sound. Muscle wire is still working too when connect directly to 3.3V supply.

Click to expand...

A piezo buzzer takes hardly any current at all, if the MOSFET isn't been turned hard ON it would still work the piezo, and not the muscle wire when it needs a LOT more current.

Measure the voltage across source and drain of the FET when it's turned ON, see how much you're losing.

The battery is 3.8V.

With the buzzer connected, when the FET is ON, the voltage across the source and drain is 0V. When it is OFF, the voltage across is 3.4V.

With the muscle wire connected, when FET is ON, voltage across the source and drain is 0V. When FET is OFF, the voltage across the source and drain is 0.6V.

Nigel Goodwin said:

A piezo buzzer takes hardly any current at all, if the MOSFET isn't been turned hard ON it would still work the piezo, and not the muscle wire when it needs a LOT more current.

Measure the voltage across source and drain of the FET when it's turned ON, see how much you're losing.

Click to expand...

Kian said:

The battery is 3.8V.

With the buzzer connected, when the FET is ON, the voltage across the source and drain is 0V. When it is OFF, the voltage across is 3.4V.

With the muscle wire connected, when FET is ON, voltage across the source and drain is 0V. When FET is OFF, the voltage across the source and drain is 0.6V.

Click to expand...

As it's only 0.6V when the FET is OFF, then there's no power going through the muscle wire - possibly you're not making proper contact with it?.

Try measuring the voltage on the muscle wire (measure from source on the FET - which is chassis or 0V) - there should be 3.4V everywhere on the wire when it's OFF, and when it's ON itshould vary from 3.4V at the top end of the wire, to near zero at the bottom end.

Test with a stronger mosfet.
That MOSFET isn’t suitable.

When the FET is off, I am measuring 3.8V (battery supply voltage) everywhere on the muscle wire.

Should I go for a 'stronger' MOSFET as suggested by Sobhan? Stronger means a higher source drain current? What other parameters should I lookout for when choosing the MOSFET? A very low Rds? How much is considered low? What about Vgs? The Vgs threshold in this case should be less than 3.3V? Since I am driving with MCU with 3.3V ouput

Nigel Goodwin said:

As it's only 0.6V when the FET is OFF, then there's no power going through the muscle wire - possibly you're not making proper contact with it?.

Try measuring the voltage on the muscle wire (measure from source on the FET - which is chassis or 0V) - there should be 3.4V everywhere on the wire when it's OFF, and when it's ON itshould vary from 3.4V at the top end of the wire, to near zero at the bottom end.

Click to expand...

Kian said:

When the FET is off, I am measuring 3.8V (battery supply voltage) everywhere on the muscle wire.

Click to expand...

Your problem is obvious then - you're got no connection between the bottom end of the muscle wire and the drain of the FET. 3.8V one end of the connection and 0.6V the other end of the connection = no connection.

A lot of mosfets need 5v at least to turn on, you can get low threshold voltage ones, for this circuit you need one, 2v ought to do it.

dr pepper said:

A lot of mosfets need 5v at least to turn on, you can get low threshold voltage ones, for this circuit you need one, 2v ought to do it.

Click to expand...

Not going to help when the wire and FET aren't connected

Ooops nope.

dr pepper said:

A lot of mosfets need 5v at least to turn on, you can get low threshold voltage ones, for this circuit you need one, 2v ought to do it.

Click to expand...

That was my thought too. But looking at the data sheet it has a threshold of 1V. And fully on at 5V.

I fixed the connection. The muscle wire definitely connected between the Drain of the FET and VBat.

I have made the measurements again:

FET is OFF, Vbatt is 3.77 V and VDS is 3.77 V
FET is ON, Vbatt is 3.77 V and VDS is about 1V.

Still the muscle wire is not actuated.

Here is what I have tried:

1. Connect muscle wire actuator directly to battery terminals (3.77 V). Muscle wire is actuated.
2. Connect the Gate (with pull down 10K resistor to ground) to a 3.3 V source. Not actuated.
3. Connect the Gate (with pull down 10K resistor to ground) to a 3.77 V source. Not actuated

Nigel Goodwin said:

Not going to help when the wire and FET aren't connected

Click to expand...

Hi,

What information should I be looking for in the datasheet for the VGS threshold?

The MGSF1N02 datasheet says the following:



I presume the minimum VGS I need is 1 V but the typical voltage is 1.7 V and the max voltage is 2.4 V?

Or should I look at this graph:



I also don't understand this graphs above. The one on the left shows the VGS and ID. Does it mean that if I apply 3V to VGS, I can get 1.5A flowing from drain to source? But the maximum drain current is only 750mA. The graph on the right is also confusing for me. I am applying 3.3V to VGS, so I am looking at the 3.25V curve. And its giving 2.3A for ID. And what is the x axis VDS for?

At the end of the day, I just really need 3V - 3.6V and 600-700mA for 0.5 secs to actuate the muscle wire. Could someone recommend me a part number a suitable MOSFET?

Thanks in advance!

shortbus= said:

That was my thought too. But looking at the data sheet it has a threshold of 1V. And fully on at 5V.

Click to expand...

Hi all,

I just like to add another observation. Using the same MGSF1N02 MOSFET, with Vbatt of 3.77V and VGS of 3.77V also, the muscle wire actually gets actuated after a long time if I keep the voltage at VGS there for about 10-20 seconds (its very variable).

Please correct me if I am wrong. The muscle wire decreases it length by a few percent when current passes through it. Basically the current generates heat and contracts the wire. Maybe what is happening now is that somehow the drain current is quite low and its causing the wire to heat up slowly, thus it taking a long time for the wire to heat up to the right temperature and contract. On the other hand when I connect the muscle wire directly between the battery terminals, a larger current flows through the wire, hence it generates the heat fast and contracts almost instantly.


I have another MOSFET lying around (DMN2075U) and tried using it. The muscle wire actuates slightly faster, within 5 seconds with VGS of 3.77 V (same as battery voltage). However, using VGS of 3.3 V does nothing. I guess my problem is with choosing the right MOSFET for this application (and I don't really know how to choose a right one).

I also wonder if the thickness of the wires can cause a problem too since I am expecting about 700mA to pass through the connections. I am doing the prototpying on a breadboard now with jumper wires. The MOSFET is connected via a SOT-23 to DIP breakout board.

Hope someone can help. Thanks!

Kian said:

I fixed the connection. The muscle wire definitely connected between the Drain of the FET and VBat.

I have made the measurements again:

FET is OFF, Vbatt is 3.77 V and VDS is 3.77 V
FET is ON, Vbatt is 3.77 V and VDS is about 1V.

Still the muscle wire is not actuated.

Here is what I have tried:

1. Connect muscle wire actuator directly to battery terminals (3.77 V). Muscle wire is actuated.
2. Connect the Gate (with pull down 10K resistor to ground) to a 3.3 V source. Not actuated.
3. Connect the Gate (with pull down 10K resistor to ground) to a 3.77 V source. Not actuated

Click to expand...

Your issue is that the FET isn't been fully switched ON, hence the 1V across it, you need an FET that will switch ON fully with the low drive voltage.

To test everything else, simply short the FET out, across drain and source, does it work then?.

Using a breadboard probably isn't helping at all, as they have fairly poor contacts that aren't good at passing decent currents..

A MOSFET drive running from the 4.2V supply might work.

Mike.

I would prefer to work with a lower supply voltage since I am using a Lipo. When its fully charged its about 4.2V but this voltage will drop as its drained. I don't want to muscle wire to stop working when the battery drops to something less than that.

If I need an FET that will switch ON fully with low drive voltage, what should I look out for in the datasheet? Please see my questions in post #15. I am still confuse how to read the datasheet properly and what to look out for when choosing the right FET.

Thanks!

The threshold voltage is when the mosfet is barely tuned on, that is what the graphs are showing. The graph, Fig.2, shows how much each gate voltage level alows the mosfet to conduct current wise. Look for what is called a "logic level" mosfet. But it seems what you have is one, you just need one with a higher pass through current rating.