Low Voltage Shutoff

[LukeKnepp]

I have this circuit built, and the first batch of 6 pieces all worked.

Now the second time, I made 4, and only 1 works, the others all do the following.

When the voltage goes below 16.2V, the output drops to 7.8V and something buzzes softly. The voltage then after a bit, does go down to zero, but it cannot be that low for my application... Is it a problem with the Zener Diode, the latch mechanism, the MOSFET, or what?

Thank you in advance!

 

[AnalogKid]

all worked.

Perhaps you should tell us what the intended function is, what the power source is, and the characteristics of the load.

Check very carefully the component placements between the good and bad boards. a consistent problem across a batch of boards usually comes down to assembly issues.

There are no buzz-able components in the schematic. Try to determine the location of the noise.

ak

 

[LukeKnepp]

Very sorry about the few details given!!

what the intended function is, what the power source is, and the characteristics of the load

This schematic is added in front of a small capacity buck converter circuit, using the MP2357 buck regulator.
The Converter output voltage is adjustable between 5V and 10V to run a small fan.
The power supply is a lithium cordless tool battery, which needs a low voltage protection at about 16V.

component placements between the good and bad boards

I checked this as much as was possible, but I cleaned these boards with isophyrol alchohol, and could no longer read the markings very clearly.

There are no buzz-able components in the schematic

The buck converter is probably doing the buzzing because of the undervoltage when this shutoff drops to 7.8V.

 

[AnalogKid]

You are correct about the converter buzzing; it probably is the inductor windings grumbling about very narrow current pulses.

First candidate - zener diode is in backwards.

ak

 

[LukeKnepp]

My zener diode is in a SOT-23-3 package, so it cannot really be backwards, and I tested the forward and reverse-biased voltage, and it is the same, even on the ones that work and do not work.

 

[AnalogKid]

Component selection and placement: A 10 K resistor in the place for R8 would keep Vgs = Vin/2 even when the latch has Q1 turned off.

Possible swap of R8 and R6.

If the board is placed by a robot, then any of several combinations of misloading the reels could cause your symptom.

Photo of the board?

ak

 

[LukeKnepp]

I checked almost all of the components to make sure they are in the correct places...

I ordered some parts that should arrive soon to try swapping out the Zener diode among other things...

Would it have to do with static in assembly, because the first ones I made sure to always be at the static-free station, and the second batch, I did not use as much caution as I should have...

 

[AnalogKid]

Would it have to do with static in assembly, because the first ones I made sure to always be at the static-free station, and the second batch, I did not use as much caution as I should have...

Always a possibility, especially this time of year. AND, especially for M1.

For a surprisingly good, cheap, anti-static coating to apply to a work surface such as Formica, use a diluted solution of unscented Downy.

ak

 

[LukeKnepp]

Okay, I will wait and see if my new pieces fix the problem, and I will be back with you...

Thanks for your help!

 

[LukeKnepp]

How static is a solid wooden workbench for this type of work?

Also, my case enclosing this board is like a rubber/plastic type material that is injection molded.

 

[LukeKnepp]

Okay, so I switched out M1, D2, and Q3, and it still does the same thing... Should I continue replacing parts, or what?

 

[Diver300]

C2 looks like a mistake to me. You could try putting it in parallel with R1 instead.

The possible issue is that in real life, the battery will have resistance, so when the MOSFET turns off, the battery voltage will increase. Having a capacitor in parallel with R8 will tend to make the MOSFET turn on again as the battery voltage increases.

The MOSFET has a gate threshold voltage between 1 and 3 volts. https://www.onsemi.com/pub/Collateral/FDS4685-D.PDF and it could be that the second batch has a lower gate threshold voltage, making turning back on more likely.

You could try simulating with some internal resistance in the battery.