Low power micro switch MOSFET transistor tutorial

[Doctor_Code]

In my mods I've always had lots of problems with switches. To get a switch small enough I would usually have to run the switch far past what it was rated for. This would burn the switches out pretty quickly and lead to complete failure, or at least 'flaky' operation.

So far I've seen some talk about transistors, MOSFETs, and things of the like, but nothing that showed 'hey! this is how to do it' - so here it goes.

What you will need:

• 1 MOSFET Transistor - I used one from Radio Shack (IRF510) Max 60V and 4A
• 1 1k Ohm resistor
• 1 100k Ohm resistor
• 1 Micro switch (you could also use a Darlington pair and just have two contacts that would trigger the atty like a touch switch - but I digress, and that is a bit beyond this first post)
• As this is only the switch, you will need the rest of your mod, including power source, atty, LED and whatever other additions.

The MOSFET transistor itself is pretty simple in operation. There is a Gate, Drain, and a Source. Drain is where current flows into, Source is where current flows out of, and the Gate is what controls if the MOSFET is on or off. I also found this video (not mine) to be somewhat helpful YouTube - Transistor / MOSFET tutorial

The quick explanation of the circuit is as follows:

The center pin of the atty gets connected to the positive battery terminal.

The outer part of the atty gets connected to the Drain.

The Source pin gets connected to the negative battery terminal.

This is the part where I hope I am explaining things correctly. You can't just pump the full current into the Gate pin - as it would most likely blow the transistor (or at least your low current micro switch). So you will need some way to limit this.

One leg of the micro switch gets connected to the positive battery terminal. The other leg of the micro switch gets connected to a 1k Ohm resistor. The other end of the resistor gets connected to the Gate pin.

The gate pin also gets connected to the negative battery terminal with a 100k Ohm resistor. You may ask well, 'Why is this needed?' I'm not really sure, but it won't work if it's not there.

Here's the schematic:

And here it is in all it's breadboard test circuit glory!

I know it looks like I'm dumping about 8V into the atty (which I am) but this was only a test, and I was firing the atty for a very short time - just to make sure things work. I plan to use a 5V regulator.

The left side of the breadboard is the MOSFET circuit. The right side is not completely unrelated, but it is not used for the circut. I was playing around with the TI PTR08100WVD I just got in. I'm using pretty much the same setup here as http://www.e-cigarette-forum.com/forum/battery-mods/114901-evercool-variable-voltage-mod.html except I am using a regular 220 Ohm resistor instead of the variable one.

Now I just have to get everything soldered together and figure out what I am going to put it all in. I'm thinking a nice wooden case this time

Kind of long winded, but, I hope this helps someone out there!

 

[BobFromChicago]

Being that i am somewhat of a ..... and still believe that electricity is really magic there is one question that i have, according to the video a seperate or reduced power supply of 5 volts is needed to operate the gate which then can allow the flow of a larger voltage to go through it to what ever you wish to operate. If small or mini switches are buring up because of the voltage wont they still burn up using them as the switch for the gate since it is still getting 5 volts through it? Or am i (as i believe i am) missing something here?

P.S. sorry for the stupid question but it's the only way i'll learn.

 

[Doctor_Code]

It's less of a voltage thing, and more of an amperage thing. The micro switches I am using are rated at 12V and 0.05A (50mA). It takes a lot more amperage to fire off the atty. I just put a multimeter on my atty, and got 3.7 Ohms of resistance. With the 5V mod the atty would pull about 1.35A.

Without the MOSFET transistor, I would be pushing the full 1.35A into the switch that is only rated at 0.05A

Back to the switch burning out. We can use Ohms Law. E = I * R ( I = Amps, E = Voltage, and R = Resistance)

Because I = E/R

Assuming the full, unregulated, 7.4 Volts from the two 3.7V Li-ion batteries that I am using. Knowing the resistor behind the switch is rated at 220 Ohms. We can now fill in E and R.

E = 7.4V

R = 220 Ohms.

I = 7.4 / 220 = .0336 Amps.

I am still well within the maximum ratings of the switch.

 

[Bubo]

Sweet... But where were you two weeks ago?!

 

[BobFromChicago]

ahhh ok, so it isnt the voltage, its the amperage thats messing up the switch, and the resistor is going to reduce some of that amperage so as not to burn out the switch and will still have enough voltage to trigger the gate in the MOSFET there by allowing the full power to go to the atty. Right?

 

[Doctor_Code]

Exactly!

 

[Sci]

Back to the switch burning out. We can use Ohms Law. E = I * R ( I = Amps, E = Voltage, and R = Resistance)
Because I = E/R

E=I*R... Thank you. I see V used instead of E so much anymore.

Also, thank you for this explanation. Great job on it.

 

[AttyPops]

This is really great Doctor_Code, thanks!!!!

I have a really dumb question tho.... When I wire in the switches (at least so far) they are between the positive terminal and the atty. Yours is between the negative terminal (DC Ground) and the atty.

Now, I know that disconnecting the negative terminal of a battery is recommended since electricity (an electron) is a negative thing and Ben Franklin messed us up. So.... Is it normal to wire a switch on the Negative side, or the positive side? I know atties don't care, but if the entire body of ... say... a metal mod was a ground (-) why wouldn't it just be grounded all the time????? and the circuit completed on the positive terminal? In most schematics, the positive is logically viewed as the emitter and everything ties to a common ground (the negative side) for DC even tho electrons don't work that way. So... ?????????

 

[bigblue30]

So far I've seen some talk about transistors, MOSFETs, and things of the like, but nothing that showed 'hey! this is how to do it' - so here it goes.

Great link to the video. I watched that before and it does have some very good info about MOSFETS

What you will need:

• 1 MOSFET Transistor - I used one from Radio Shack (IRF510) Max 60V and 4A

If you are going to actually use a MOSFET in a mod, there are better ones out there then the IRF510. Why…when you look for a MOSFET for our use there are a few very important numbers you want to look at on the spec sheet.

The first is the BV (or what is the max voltage this thing can handle)

BV = 100 volts

For the IRF510 that you are using the 100 volt rating is more than enough.


The second is the Continuous drain current Current (or how many amps can this thing take before the “smoke” inside gets released… All electronics run on smoke….If you let it out they quit working

Continuous drain current = 130 amps
Again for the IRF510, 4 amps should be good for what we are doing.


The third number you want to look at is the Gate Threshold Voltage (or how many volts do I have to put on the Gate to turn this thing on)

Gate Threshold Voltage 2 to 4 volts

The 4 volt number here is not good for us…Why…. I know you are looking to build the Evercool mod and this is not as big of a problem for a 2 battery mod, but let’s look at a 1 battery mod first.

With a 1 battery mod you are working with a 3.7 volt battery (4.2 when it comes off the charger). The GTV of the IRF510 is between 2 and 4 volts. This means that the voltage to turn on the IRF510 can be anywhere between 2 and 4 volts. Quality control is not the best for these… one might be 2 volts and another one might be 4.

So what you might say… A MOSFET “should” act like a switch… ether “on” or “off”, but with a high GTV it can act like a resistor. Not really on and not really off. This is not a good thing for a MOSFET. I guarantee you will let the “smoke” out if this happens.

So when choosing a MOSFET for our Mods always try to get one with a 1-2 volt GTV spec. Like this one:

http://search.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&site=ca&keywords=IRLU3114ZPBF-ND&x=0&y=0
VGS(th) Gate Threshold Voltage 1.0 ––– 2.5 V

Back to the evercool mod. When you first put in the batteries you have around 7.4 volts…..but when one dies, and one always dies before the other, you now have 3.7. If you keep hitting the switch when that happens….see above. Like I said before, this is not as big of a problem as a 1 battery mod.

The center pin of the atty gets connected to the positive battery terminal.

In your schematic you show the center pin going to ground. The long line is positive and the short one is negative. As this is a teaching “post”, I thought I would let you know.

This is the part where I hope I am explaining things correctly. You can't just pump the full current into the Gate pin - as it would most likely blow the transistor (or at least your low current micro switch). So you will need some way to limit this.

One leg of the micro switch gets connected to the positive battery terminal. The other leg of the micro switch gets connected to a 1k Ohm resistor. The other end of the resistor gets connected to the Gate pin.

You are correct here. The gate looks at voltage and you want to limit the current as much as possible to the gate and the switch. A better choice here would be a 47k resistor. No need to burn up millamps when you do not have to.

The gate pin also gets connected to the negative battery terminal with a 100k Ohm resistor. You may ask well, 'Why is this needed?' I'm not really sure, but it won't work if it's not there.

I do not see a need for this resistor. You are “shorting” your output to ground with a 100k resistor.

Here are 2 pictures of the same circuit that have worked for me.

 

[bigblue30]

E=I*R... Thank you. I see V used instead of E so much anymore.

Also, thank you for this explanation. Great job on it.

Sci,

The "E" is really old school.. and by old I mean...that is how I learned it 30 some years ago. It stood for "electromotive force"

 

[WillyB]

Great link to the video. I watched that before and it does have some very good info about MOSFETS


You are correct here. The gate looks at voltage and you want to limit the current as much as possible to the gate and the switch. A better choice here would be a 47k resistor. No need to burn up millamps when you do not have to.


I do not see a need for this resistor. You are “shorting” your output to ground with a 100k resistor.

Here are 2 pictures of the same circuit that have worked for me.

Slightly confused here. Isn't the 47k resistor a 'pull down resistor' and isn't it's purpose to keep the gate from floating in a semi-open state as opposed to being a current limiter?

FWIW following Nuck's example, he actually solders a 10k resistor between the source and the gate in his guide.

I had a hard time following your mosfet suggestions.

Does this one meet your various criteria (the GTV is fine). Couldn't find 'ON-State Drain Current'

http://www.jameco.com/Jameco/Products/ProdDS/669943-DS01.pdf

 

[bigblue30]

Slightly confused here. Isn't the 47k resistor a 'pull down resistor' and isn't it's purpose to keep the gate from floating in a semi-open state as opposed to being a current limiter?

Yes WillyB, That 47k is there to make sure the MOSFET does not fire when it is not suppose to, or go into a resistor mode (not on and not off) . I redrew his circuit to show the connection to ground.

The other thing the 47k does....... If you look at my circuit now.. On the left you have the positive side of the battery, the switch, then a 47k to ground.

In this configuration, the 47k limits the amount of current going through the switch. This is why I suggested to replace the 1k with the 47k, and redrew the one he had.



I had a hard time following your mosfet suggestions.

Does this one meet your various criteria (the GTV is fine). Couldn't find 'ON-State Drain Current'

http://www.jameco.com/Jameco/Products/ProdDS/669943-DS01.pdf[/QUOTE]

You are correct... I should of said Continuous drain current ( I will fix the previous post.... )and yourd is 30 amps... this should work great. and your GTV is 1-2. This is even better, and at 1/2 the cost, nice find.

 

[WillyB]

...

In this configuration, the 47k limits the amount of current going through the switch. This is why I suggested to replace the 1k with the 47k, and redrew the one he had.

Now I see. Kinda like putting a resistor on the negative leg of an LED.

Thanks

 

[AttyPops]

This is really great Doctor_Code, thanks!!!!

I have a really dumb question tho.... When I wire in the switches (at least so far) they are between the positive terminal and the atty. Yours is between the negative terminal (DC Ground) and the atty.

Now, I know that disconnecting the negative terminal of a battery is recommended since electricity (an electron) is a negative thing and Ben Franklin messed us up. So.... Is it normal to wire a switch on the Negative side, or the positive side? I know atties don't care, but if the entire body of ... say... a metal mod was a ground (-) why wouldn't it just be grounded all the time????? and the circuit completed on the positive terminal? In most schematics, the positive is logically viewed as the emitter and everything ties to a common ground (the negative side) for DC even tho electrons don't work that way. So... ?????????

NM....misread the text....

so.... in the original post the atty is shorted to ground via the pull-down resistor even when "off"; the voltage regulator(with gate) or power MOSFET when activated "shorts" the atty to ground with little resistance when "on"..... BigBlue30's schematic is a better way, assuming voltage regulators are near infinite resistance when off - no batt drain. True?

 

[BJ43]

2 x IRFZ44 IRFZ44N MOSFET N-Channel 49A 55V | eBay

Will this work with the 47K resistor??

 

[slimest]

This transistor will work, but I would not use it for voltages about 3.5 - 7 volts. Better to use with lower Gate to Source voltage.

P.S. 47k or 10k, or even 1 к and 100 k - does not matter. Every suitable MOSFET will work reliably independent of resistance. If you solder instead of this resistor 4-10 Mohms, you can replace the button with external contacts and switch on your atomizer by your fingertip.

 

[jhonutz]

yep will works, but only if you use switch.For touch need 1M to 10M ohm.


Edit: @slimest was faster then me .Try to use mosfet`s with lower RDS-ON and Vgs threshold 0.x - 1V

 

[BJ43]

yep will works, but only if you use switch.For touch need 1M to 10M ohm.


Edit: @slimes was faster then me .Try to use mosfet`s with lower RDS-ON and Vgs threshold 0.x - 1V

Thanks I plan to use a tac switch, I would forever be setting off a touch switch, not for me.

 

[BJ43]

Do I connect the mosfet to the (-) in or out of the LM2596 regulator? I normally use 12 v 5amp source in and 5 to 6v out.

 

[nicotime]

What about the RDS(on)...I thought that was important to look at also?

I have been using both the Z34 and the 3114 with the schematics BB30 posted for a couple years now and never any problems. I like the 3114 because its the perfect size for mods. I started out using 47K on the gate and then saw where people were using 10K..so I went with that...but I havent noticed any difference either way. I'll put a pic up at the bottom for size reference for everyone...they are sitting beside a tact switch and 3A diode.

Good info guys..thanks for putting it out there!