MOSFET instead of Diode for Solar panel?

[panoss]

Can I use a MOSFET instead of a diode for preventing discharge of the battery through the solar panel?
In the picture you can see how I was thinking of connecting it with a diode and how with a MOSFET.
Is it going to work? (with the MOSFET)

(the charger module is this one)
(the MOSFET is an AO3401, P-Channel MOSFET.)

[Sowlaki]

In the picture you connected the MOSFET so that during normal operation the Gate is low and will be able to draw current from the module to the solar panel which is what you wanted to prevent (?). The sink is connected to the + pole of the panel. Don't know what you wanted to accomplish with this. Did you draw it wrongly?

[panoss]

In the picture you connected the MOSFET so that during normal operation the Gate is low and will be able to draw current from the module to the solar panel which is what you wanted to prevent (?). The sink is connected to the + pole of the panel. Don't know what you wanted to accomplish with this. Did you draw it wrongly?

(it 's a P-Channel MOSFET )

normal operation = panel delivers 5V.
MOSFET 'S VGS < 0 (Gate has 0V, Source has more than 0V)
MOSFET is conducting, current is flowing from Drain to Source (left to right), charging normally, everything ok.

NOT normal operation = panel delivers 0V(at night).
MOSFET 'S VGS < 0 (Gate has 0V, Source again has more than 0V (voltage coming from the battery))

So the MOSFET 's conduction does not stop and current is flowing from Source to Drain (right to left ).
Current is flowing from the battery to the panel?
So it 's not working?


(I want to use a MOSFET because it has a lower voltage drop. So I should go for a diode with a low voltage drop?)

[langwadt]

In the picture you connected the MOSFET so that during normal operation the Gate is low and will be able to draw current from the module to the solar panel which is what you wanted to prevent (?). The sink is connected to the + pole of the panel. Don't know what you wanted to accomplish with this. Did you draw it wrongly?

(it 's a P-Channel MOSFET )

normal operation = panel delivers 5V.
MOSFET 'S VGS < 0 (Gate has 0V, Source has more than 0V)
MOSFET is conducting, current is flowing from Drain to Source (left to right), charging normally, everything ok.

NOT normal operation = panel delivers 0V(at night).
MOSFET 'S VGS < 0 (Gate has 0V, Source again has more than 0V (voltage coming from the battery))

So the MOSFET 's conduction does not stop and current is flowing from Source to Drain (right to left ).
Current is flowing from the battery to the panel?
So it 's not working?


(I want to use a MOSFET because it has a lower voltage drop. So I should go for a diode with a low voltage drop?)

it won't work because once the fet is on it won't turn off again


use something like what was used on raspberry pi

https://raw.githubusercontent.com/raspberrypi/hats/master/zvd-circuit.png

[panoss]

I 'll just use a low voltage drop diode because your circuit, langwadt, is too complex.
But I 'll have it in mind for some future use.

[NiHaoMike]

There are "ideal diodes" like the SM74611 specifically designed for that purpose, albeit for much larger solar panels. I used one in my setup and was surprised how it barely got warm when the Schottky diode I was using before got pretty hot. (I was even more surprised how much fame the setup got me, but that's another story...)

For your application, take a look at "energy harvesting" chips.

[langwadt]

I just looked at that charger module, it seems to use a TP4056 charge controller and the datasheet specifically mentions that no diode is required

[panoss]

I just looked at that charger module, it seems to use a TP4056 charge controller and the datasheet specifically mentions that no diode is required

It 's actually the TC4056a.
I found it 's datasheet here, but to tell the truth I did not understand much.

So I found LT 's LTC4056, I suppose it 's an identical device(??).
At least I understand what it says in the datasheet!
"No Sense Resistor or Blocking Diode Required", so I 'll try it without a diode! I hope it 'll be ok.

[Kleinstein]

The MOSFET instead of a diode for reverse polarity protection is Ok. It may need extra protection for the gate of the voltage can gat higher than some 15 V (max. gate voltage) - with suitable fets this is just a resistor at the gate pin. There is no need for an extra ideal diode chip in this application. 

A diode with very low forward drop (e.g. larger low voltage schottky diode) also tends to have high leakage current.

[panoss]

The MOSFET instead of a diode for reverse polarity protection is Ok

It was not meant to be for reverse polarity protection, but for preventing battery 's discharging through the solar panel.

[Zero999]

The MOSFET instead of a diode for reverse polarity protection is Ok

It was not meant to be for reverse polarity protection, but for preventing battery 's discharging through the solar panel.

Yes, a MOSFET is fine for reverse polarity protection, but will not prevent the battery from discharging back into the solar panel.

[tunk]

I'm no expert on this, but there may be two potential problems with this setup:
- most TP4056 boards are setup to charge with 1A. With your solar panel this may not work at all.
  It may be better to use a (schottky) diode and a BMS.
- this board outputs the 18650 voltage: 3V may be too low for a 5V arduino, and 4.2V may be too
  high for a 3V arduino.

[Seekonk]

A solar panel is a current source and having a diode drop or not makes no difference. The current leakage is likely so small that having a diode doesn't gain anything.

[Eldi4]

If you really want to minimize the voltage drop because of diode forward voltage on solar panel circuitry, you should take a look at 'Ideal Diode' IC, like LTC4376, or you can build yourself with some p-channel mosfet, some tranny and resistor, just search 'Ideal Diode Circuit' at google,

[panoss]

I
- most TP4056 boards are setup to charge with 1A. With your solar panel this may not work at all.
  It may be better to use a (schottky) diode and a BMS.

I 'm in the process of testing, I 'll post results.
(till now I 'd say 'so and so')

I
- this board outputs the 18650 voltage: 3V may be too low for a 5V arduino, and 4.2V may be too
  high for a 3V arduino.

I 've tested my 3.3V Arduino with 4.2V, it has no problem. (I think it can tolerate up to 6V (datasheet, Maximum operating voltage= 6V) at it's power supply pin(the 3V3 pin))

If you really want to minimize the voltage drop because of diode forward voltage on solar panel circuitry...

According to Seekonk, the voltage drop doesn't matter:

A solar panel is a current source and having a diode drop or not makes no difference. The current leakage is likely so small that having a diode doesn't gain anything.

@Seekonk

The current leakage is likely so small that having a diode doesn't gain anything.

When you 're saying 'current leakage' you are refering to the current that returns from the battery to the solar panel when the solar panel ouputs no voltage, right?

[Ikthyoid]

In order to do this, you need to use two "back-to-back" P-channel MOSFETs (PFETs) with their sources connected together and sharing a common gate connection. It's common to see a pull-up resistor from the common gate to the common source in order to ensure stability when the gate is not actively being driven.

Just like using any PFET as a high-side load switch, you close this switch by lowering their common gate voltage to something below the "gate threshold voltage", often 0V or logic LOW; bringing the gate voltage higher than the "gate-source threshold voltage" will open (or break continuity of) the switch.

This back-to-back PFET topology is necessary (vs. using only one PFET) because the body diode of just a single PFET will allow reverse current flow.

You can also use NFETs instead of PFETs, but this can require voltage boosting to drive the gate high enough. There are also "high-side load switch" ICs available that encapsulate all of this functionality into a single IC.

[Seekonk]

Everyone loves an experiment. Why don't you measure the leakage from a fully charged battery to the solar panel.Just see if this is worth the over thinking.

[boB]

Seems to me that if this does work and turn on during the day, and the PV terminals of that PCB has battery voltage appearing at those PV terminals, the that voltage will still appear there at night, turning on the FET.

Back-to-back FETs as was mentioned might work if you operate it off of the PV voltage itself which will only be there during the day.

But you will also want to make sure that the FETs are not partially on due to low G-S voltage.

But maybe this PV module is so dang small that it cannot hurt anything ?  No idea on that part.

[panoss]

Ok guys, I have some data which I think answer to my question:



In the picture you can see the battery voltage from 07/08/2019 (Day/Month/Year) up to 09/08/2019 (today).
07/08/2019
10:33   3.38V
20:45: 3.53V

08/08/2019
09:54  3.52V
20:30  3.61V

09/08/2019
09:07  3.59V
10:13  3.63V

So we see that when no sun (e.g from 20:45 of 7th of August till next day in the morning, 09:54) practically, the voltage remains stable.
So, I conclude, the solar panel consumes no (or VERY little) energy from the battery when outputs no current! (that is when there is no sun)

(the circuit I use is the one in the picture posted in my first post without the diode D1)