0
\$\begingroup\$

I am building a circuit that stops giving power to the load when the voltage drops below 3.08V. I am using MAX809T (3.08V version) and a SI2302 mosfet (\$R_{ds}\$ on 0.056Ω@2.5V)

my circuit

When there is no load, the voltage at reset pin and mosfet's source is VCC unless the voltage is below 3.08V giving 0V at reset pin and mosfet's source. But when there is load, the reset pin is always 0V whatever voltage is at VCC.

What is wrong with my circuit?

There was a similar question supply-voltage supervision circuit (follow up) with no answers.

Update:

I 've updated the circuit based on your suggestions but the problem remains.

I 've used also pull up and pull down resistors at mosfet gate with same results.

updated circuit

I have this electronic load. When I start my power supply with 5V and the load adjusted to 0 mA, it works normally: Output is almost 5V (4.95V).

Then I adjust my load to 100mA and start decreasing the dc power to 3.0V. When voltage is about 3.1V output (load) is turned off (0V 0mA) as it should.

Then I start increasing the VCC voltage to 5V but the output (at load) is never recovered. It remains 0V. It recovers only if I adjust my load current to 0mA.

MAX809T reset pin remains grounded while increasing VCC. I can not understand why. When I test it without the mosfet if VCC<3.08 RESET = GND and if VCC>3.15 RESET = VCC.

I 've tried with AO3400 mosfet with same results.

Update 2:

After changing my load to a simple resistor of 22Ω my circuit works as expected. I don't know how my electronic load works but testing my circuit with it caused unexpected results. Thank you all for the help.

Update 3:

The circuit does not work with a real load. I 've modified it like the following image, but still does not work.

undervoltage protection

When mosfet gate is not connected (floating) I can measure the voltage at transistor collector = VCC if reset is LOW and 0V if RESET = HIGH = VCC. But when I connect the mosfet gate to transistor collector, the voltage at collector is always equal to VCC.

I 've tried also a 2N7000 mosfet in place of the transistor with same results.

\$\endgroup\$
  • \$\begingroup\$ Generally to use an N-MOS as a switch it needs to go on the low side. You can create a high side switch using a P-MOS. \$\endgroup\$ – hekete Jul 18 '19 at 14:52
  • \$\begingroup\$ @hekete I edited my port with the updated the circuit but I have the same problem. \$\endgroup\$ – Dimitris Jul 19 '19 at 7:54
  • \$\begingroup\$ So is it true that the only situation in which your circuit works is when you gradually adjust the load current from 0 to 100mA? For example a cold start at 5V with full load doesn't work either? \$\endgroup\$ – hekete Jul 19 '19 at 9:12
  • \$\begingroup\$ @hekete Correct. Cold start with load at 100mA does not work too. \$\endgroup\$ – Dimitris Jul 19 '19 at 9:18
0
\$\begingroup\$

Your updated circuit looks okay. The MAX809 has push-pull output on the RESET pin and can source 20mA. I believe 20mA is enough to drive your MOSFET, but double check that.

Since you have a push-pull output, no pull-up/down resistors are needed (though RESET will become floating when Vcc drops below 1V).

You had unexpected readings in both your circuit configurations, so really at this point either one of the components are damaged or something is going on with your load. So, what exactly is your load?

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ I write what my load is, below the updated circuit. I will test again tomorrow with a single resistor as a load to see if there is a difference. \$\endgroup\$ – Dimitris Jul 21 '19 at 14:18
  • \$\begingroup\$ @Dimitris interesting load, how are you wiring it up? Just to the battery screw terminals? The other variable is your input, how much current is it rated to deliver at 5V? \$\endgroup\$ – hekete Jul 22 '19 at 5:53
1
\$\begingroup\$

The effect of a load seems like a wiring problem of some kind, but in any case this circuit will not work very well. You have the load wired on a source follower so the MOSFET will always drop Vgs at the load current, which will vary from unit-to-unit and with the load current but typically it's more than 1V at a few mA.

If you look at your circuit you can see that the voltage that turns the N-channel MOSFET "on" (Vgs) is the Vcc (on the output of the MAX809) minus the voltage across the load. Since it takes a couple volts to turn the MOSFET fully on you can never get anywhere near the full Vcc across the load (doesn't explain the "zero" you say you are observing though).

What you can use is a voltage detector with active HIGH push-pull output (eg. MAX810) and a P-channel logic-level MOSFET.

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

The state of the MOSFET depends of the voltage drop between its gate and source terminals (Vgs). You should put the load before the MOSFET (on the drain terminal), and tie its source to GND.

You should also put a pull-down resistor in the gate. 10k is a good value to start with.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.