Negative Voltage Output from Logic Level Input

Thread Starter

AI1999

Joined Feb 17, 2023
9
Hey guys,

I'm looking for a circuit that generates -5V from a +5V signal from a microcontroller to open and close a depletion mode n channel-mosfet. I figured this could be done with transistors/solid-state devices but I'm stuck and would appreciate some help with this

I'm looking for a circuit with the following input/output/behavior:

Input: +5V --> Output --> -5V --> FET OPEN
Input: 0V --> Output --> 0V --> FET CLOSED
 

Ian0

Joined Aug 7, 2020
10,277
Hey guys,

I'm looking for a circuit that generates -5V from a +5V signal from a microcontroller to open and close a depletion mode n channel-mosfet. I figured this could be done with transistors/solid-state devices but I'm stuck and would appreciate some help with this

I'm looking for a circuit with the following input/output/behavior:

Input: +5V --> Output --> -5V --> FET OPEN
Input: 0V --> Output --> 0V --> FET CLOSED
Why not simply use an enhancement mode MOSFET? If you don't have a negative supply, it's not exactly easy.
 

sarahMCML

Joined May 11, 2019
423
If you really want to do it that way, look up "Negative charge pump circuits". You can either drive it with a spare oscillating output pin from your micro, or something like a CMOS 555 timer.
 

ronsimpson

Joined Oct 7, 2019
3,220
I have a bad idea, but it might work. lol
This IC has a 5V to +/- 8V power supply built inside. It also has a driver that takes a TTL signal and outputs a +/- 6V inverted signal.
1716381717090.png
 

LadySpark

Joined Feb 7, 2024
194
Hey guys,

I'm looking for a circuit that generates -5V from a +5V signal from a microcontroller to open and close a depletion mode n channel-mosfet. I figured this could be done with transistors/solid-state devices but I'm stuck and would appreciate some help with this

I'm looking for a circuit with the following input/output/behavior:

Input: +5V --> Output --> -5V --> FET OPEN
Input: 0V --> Output --> 0V --> FET CLOSED
Normally, you use a resistor on the gate tied to +Vcc and the low logic turns it off.
 

Ian0

Joined Aug 7, 2020
10,277
I have a bad idea, but it might work. lol
This IC has a 5V to +/- 8V power supply built inside. It also has a driver that takes a TTL signal and outputs a +/- 6V inverted signal.
View attachment 322905
Not such a bad idea at all, but you’d need a diode to block the positive voltage so that the gate can only go negative of zero. And MAX232 is dead cheap these days.
I used one to give a bit extra voltage to a piezo buzzer!
 

Ian0

Joined Aug 7, 2020
10,277
Normally, you use a resistor on the gate tied to +Vcc and the low logic turns it off.
You would if it were an enhancement mode MOSFET, but the TS wants to use a depletion mode MOSFET, which needs a voltage lower than -3V to turn it off. With the gate at 0V, a depletion mode MOSFET will still be ON.

One day he might tell us why he wants to use a depletion MOSFET because it’s a mystery so far.
 

LadySpark

Joined Feb 7, 2024
194
One day he might tell us why he wants to use a depletion MOSFET because it’s a mystery so far.
some reset circuits are made that way, but with 0V threshold depletion mode gate. An example would be the raspberry Pi4b reset circuit that I completed the onboard power switch circuit before by completing the global_reset gate circuit.

Screenshot_2024-05-22_07-10-55.jpg
 

Ian0

Joined Aug 7, 2020
10,277
some reset circuits are made that way, but with 0V threshold depletion mode gate. An example would be the raspberry Pi4b reset circuit that I completed the onboard power switch circuit before by completing the global_reset gate circuit.

View attachment 322906
DMG1012T is enhancement mode, with a threshold as high as +1V.
(And I presume that the GLOBAL_RESET isn‘t left floating like it is shown on the diagram)
 

LadySpark

Joined Feb 7, 2024
194
(And I presume that the GLOBAL_RESET isn‘t left floating like it is shown on the diagram)
The gate circuit is left uncompleted, that is why they are instantly on when you plug them in.
Pick one up and check it out, because its the first time I ever seen someone make something without completing a gate circuit and leaving it hanging in outer space like that.

here is my power swtch circuit I came up for a 4b with a case fan control (with generic low speed for pwm fans)
Debug-Off switch is for killing the Pi when prototyping. I used an (on)-off-(on) toggle for this, the 4B power reg is one used in smart phones and tablets, so you hold the power button 'on' for 5 seconds. In the final build, I used a 10K resistor IC array and tied two resistors in parallel to get 5K for the 4.7K resistors in the circuit.
rasp4bpwr-fan.jpeg
 
Last edited:

Thread Starter

AI1999

Joined Feb 17, 2023
9
Do you have -5V available?

Why is The circuit using a very rare depletion-mode MOSFET?
Or is it a JFET?
I have an application that needs a normally closed switch (for formula student/FSAE racing team). Most people use NC relays but we wanted to reduce the size of our system as much as possible so we're opting for a solid state solution
 

Thread Starter

AI1999

Joined Feb 17, 2023
9
Why not simply use an enhancement mode MOSFET? If you don't have a negative supply, it's not exactly easy.
My specific application calls for a normally closed switch. I don't want to use a relay for size reasons and thought it would also be a good opportunity to learn more about electronics. Maybe I'm trying to reinvent the wheel here but I came up (well saw it D_fet_drive.pngonline and changed some values) with something that seems to fit my needs (attached). It doesn't exactly go to -5V but to open my Depletion fet, datasheet says I need -2.5V to -4.5V Vgs
 

Ian0

Joined Aug 7, 2020
10,277
My specific application calls for a normally closed switch. I don't want to use a relay for size reasons and thought it would also be a good opportunity to learn more about electronics. Maybe I'm trying to reinvent the wheel here but I came up (well saw it View attachment 322940online and changed some values) with something that seems to fit my needs (attached). It doesn't exactly go to -5V but to open my Depletion fet, datasheet says I need -2.5V to -4.5V Vgs
Does it matter that one terminal of the "switch" is connected to the negative supply? If that isn't acceptable, then it will have to be the relay. When you have incorporated the ICL7660, and its capacitors, and a level translation circuit, it will be the size of a small relay.
The only reason to employ the depletion mode device is to be certain that when the driving circuit loses power the output will be connected to 0V. If you are not worried about what happens when the driving circuit loses power and the switched circuit is still powered up, an enhancement mode MOSFET driven by something to invert the logic will do the job in less space.
 

Thread Starter

AI1999

Joined Feb 17, 2023
9
Does it matter that one terminal of the "switch" is connected to the negative supply? If that isn't acceptable, then it will have to be the relay. When you have incorporated the ICL7660, and its capacitors, and a level translation circuit, it will be the size of a small relay.
The only reason to employ the depletion mode device is to be certain that when the driving circuit loses power the output will be connected to 0V. If you are not worried about what happens when the driving circuit loses power and the switched circuit is still powered up, an enhancement mode MOSFET driven by something to invert the logic will do the job in less space.
My idea was to have the source pin of the FET connected to GND with a resistor, the drain connected to a positive supply rail, and the negative signal from my LTSpice schematic driving the gate (Discharge_en in the below screenshot/Altium schematic).

Discharge_en is a +5V logic signal from some other circuit that is generated based on how charged or discharged our motor controller is, and in case of failure (i.e. Discharge_en is 0V), the safe state of the circuit is closed (discharging using bleed resistors).

If I'm understanding correctly, the only pin of the FET connected to my generated negative signal should be the gate, which is perfectly acceptable

The motivation for using a depletion FET is to ensure that the switch is normally closed in the case of the drive circuit failure, so it's as you say.

I will look into the ICL7660 IC and supporting components, seems like a possible option as well. Thank you for your detailed replies.
 

Attachments

AnalogKid

Joined Aug 1, 2013
11,202
Here is a first-pass at a signal level translator that meets the requirements in post #1. It might need adjustments as more information becomes available.

The output changes state when the input crosses approx. +2.4 V. Input current required is less than 1 mA. There is no hysteresis.

UPDATE:


3B is a re-draw of the schematic with a more conventional power flow direction.

ak


-5-V-Signal-Conv-3A-c.gif -5-V-Signal-Conv-3B-c.gif
 
Last edited:

Thread Starter

AI1999

Joined Feb 17, 2023
9
Here is a first-pass at a signal level translator that meets the requirements in post #1. It might need adjustments as more information becomes available.

The output changes state when the input crosses approx. +2.4 V. Input current required is less than 1 mA. There is no hysteresis.

ak


View attachment 322944
Wow this is much more elegant than what I did, I'll try to play around with it and understand it properly, thanks! Also, do you mind sharing how you made this diagram?
 
Top