power a 3.3v circuit from 5v using a 2N7000 eMosfet?

R

rmeldo

Senior Member
Hi,

I have searched the forum But I haven't found an answer. I need a sanity check, as I am not confident in my limited skills.

I need to control a 3v3, 200mA load (ESP8266) with a Picaxe running at 5v using 3AA batteries and to make the batteries last for days I need to turn off the 3v3 load most of the time, turning it on only when needed.

I am trying to avoid using an always on 3v3 voltage regulator because it draws current even when idle, in region of 3mA (not sure whether it is just the on board LED). This is three times the current drawn by the rest of the circuit, excluding the ESP8266.

So I had the thought to instead use a eMOSFET 2N7000. I could drive it with the Picaxe and it would be able to cope with the required 200 mA. I tried it out and, when on I get a voltage drop of about 1.6V, so I wouldn't need the 3v3 voltage regulator.

Is this a good approach or am I going to run into trouble?

Many Thanks
Riccardo
 
S

sghioto

Senior Member
Dissipation is 320mw with a 1.6 volt drop at 200ma, getting near the max 400mw the 2N7000 is rated but probably workable. Another approach is to connect three diodes in series with the ESP8266 working as a zener that will drop the voltage appx 1.8 volts.

Steve G
 
A

AllyCat

Senior Member
Hi,

If you're just taking the gate to "5 volts" (via a PICaxe pin) and relying on the Gate-Source voltage drop, then no, that's not likely to be very stable or predictable. Better to use a bipolar Darlington (single package or "DIY" with two NPNs) which will drop close to 1.2 volts (two forward diodes) quite reliably. Also, the PICaxe output will (or can be arranged to) drop more than zero volts!

But unless you're using expensive Lithiumm AA cells, you're not going to get 5 volts for long (if at all); 3 x Alkalines will soon be down to 4.5 or 4 volts, with 3.3 volts being a realistic "end of life". One solution would be to find an ultra-low quiescent current/LDO 3.3v regulator, but I cant suggest one (and it very probably would be SMD).

The (rather more advanced) way that I would do it, is a "DIY" PNP (or P-channel Logic-Level FET if you can find one) "buck" switch-mode converter (basically just a Transistor, Inductor, Diode and Capacitor), driven from a PICaxe PWM pin (via a resistor). That should actually reduce the current drawn from the battery. If there is always some load current (when active), it doesn't have to work "closed loop" (i.e. with analogue voltage feedback into the PICaxe), you could predict/control the voltage drop by calculating the duty cycle. Or monitor the PICaxe supply voltage (Vdd) from the batteries using CALIBADC (no pins required) and control the PWM to deliver a "stable" ~3.3 volts.

Cheers, Alan.
 
fernando_g

fernando_g

Senior Member
A big favorite of mine is the LP2951.

Its guaranteed quiescent current is about 120uA all the way to within 0.5v of dropout, in which case it only increases to 170uA.

Best of all, it has a shutdown pin, and the quiescent current drops to less than 10uA.

It is available in 5v, 3v3 and 3v0 versions, which are accurate within 1%

Best of all, it is not expensive, and comes in easy-to-solder SOIC8 packages.

I've used it in many battery project over the years, and it works great.
 
R

rmeldo

Senior Member
thanks.

I like the idea of a shutdown pin.

The one you recommend has only 100mA current capability however. I will have to look for one which can handle more.

Thanks
Riccardo
 
A

AllyCat

Senior Member
Hi,

I've never used one, but the Microchip MCP1700-3302E/TO that you linked above looks highly suitable. And it appears you could still buy just one from CPC for 28p with free shipping! But they do have lots of other goodies as well. ;)

Cheers, Alan.
 
You must log in or register to reply here.
Top