5 volt 1000 mA regulator or reduction to

Thread Starter

Schidtztorm

Joined Oct 1, 2019
72
I'm trying to figure out a way to design a simple 5 volt 1000 mA regulator, or at least a way to drop the 12 volt supply down to 5 volts with 1 Amp. I don't have access to IC's or zeners, so I'll have to use voltage dividers to set voltage. And here's the real kicker! I have to make this with a FET. Enhancement/switcher type. But it's good for many amps, and up to 60 volts.
This is a type of contest between a couple of colleagues of mine and I'm setting them up for it. Lol. I just want to make sure I have a realistic design before I dish it out. They don't think it can be done without risks of frying the gadgets plugged into it.
Can anyone help me with this. If you need the FET info, I'll give it to you, or send you the datasheet. I'm hoping for a general design that will work on most FETs of this caliber.
 

crutschow

Joined Mar 14, 2008
23,378
Yes, we need the FET info.

If you use a resistive divider from the 12V to generate the 5V, then the 5V value will change with the 12V.
Is the 12V supply stable enough for that?
 

AnalogKid

Joined Aug 1, 2013
8,109
Without a zener, output regulation will be very poor with a FET. A 7 V differential is more than enough to use the FET as a source follower. But - unlike a bipolar NPN transistor in an emitter follower, a FET has a much less steady gate-source voltage. It is larger (2-3 V instead of 0.7 V -ish); for the same change in load current (for example, 0.25 A to 1.0 A), Vgs varies more than Vbe, creating a larger percentage change in output voltage; it varies way more with temperature, and the device temperature will vary significantly with output current. At 1 A without a heatsink, the transistor will get hot enough to melt your fingerprints.

Without a voltage reference and some feedback, two resistors and a FET will get you sorta-kinda what you want. Two resistors and a bipolar NPN will be better. Two resistors and a bipolar NPN darlington will be best, but with some trade-offs.

ak
 

crutschow

Joined Mar 14, 2008
23,378
You could use a 2-transistor NPN differential (long-tailed) pair to control a series P-MOSFET.
One differential input is the 5V from the voltage divider and the other is the output 5V.
For proper negative feedback, the MOSFET gate is connected to the NPN collector of the transistor controlled by the 5V from the divider.
That will give good regulation as long as the 12V is stable.
 

Alec_t

Joined Sep 17, 2013
10,384
If you are allowed or have access to BJTs, normal silicon diodes and passive components you could make a simple switch-mode buck converter.
 

crutschow

Joined Mar 14, 2008
23,378
Below is the LTspice simulation of a circuit using a P-MOSFET and two NPN's in a differential stage.
The output voltage is proportional to the input voltage, with 5V out for a 12V input.
upload_2019-10-1_13-37-40.png
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
72
If you are allowed or have access to BJTs, normal silicon diodes and passive components you could make a simple switch-mode buck converter.
I'm not familiar with those
I have standard rectifier diodes. I might be able to find a bit, bit nothing big enough to withstand 7 watts. The big FETs are all I have that can hold it.
I posted the get info above. Could you send me a schematic?
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
72
Below is the LTspice simulation of a circuit using a P-MOSFET and two NPN's in a differential stage.
The output voltage is proportional to the input voltage, with 5V out for a 12V input.
View attachment 187171
Thank you for that. Though your schematic isn't clear (it's distorted) but I'm familiar with that type of regulator. And you are absolutely correct. I just don't have access to any big BJTs. Only the FETs. Is this hopeless? The 12 volt supply is extremely stable. (switcher type)
Is there a way to make the above power FETs work with resistive voltage dividers?
 

crutschow

Joined Mar 14, 2008
23,378
your schematic isn't clear (it's distorted)
Looks fine in my post.
Is there a way to make the above power FETs work with resistive voltage dividers?
You could use the N-MOSFET as a source-follower with the gate connected to a resistive divider (one resistor being a pot so you can adjust the voltage to the desired value.
The voltage output will be somewhat sensitive to load and temperature.

LTspice simulation of example circuit below for load varying from about 1A to 0.1A:

upload_2019-10-4_0-45-28.png
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
72
Thank you so much for that. And your attachments no longer look distorted. Must have been a bad loggin or something.
Can I use a voltage divider instead of a potentiometer. If so, what voltage do you think is best. I'm not sure of the voltage drop from the gate source. Is it going to vary a lot? What do you think the tolerance is on a cell phone voltage wise?
 
Top