Maker Pro
Maker Pro

Low power cutoff not working

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
Hi all,

I’m trying to perfect a circuit using a TL431 and a PNP MOSFET to cut off power to, in my drawing, shown as a LED. I’m using two 18650 cells in parallel (3.0-4.2V).

The problem is I can’t seem to get it to work. The TL431 is good from 2.5-36V, but it seems, when it gets down to the lower end of its voltage range it struggles. I tried all kinds of fixes including different resistances and transistors. I’m presently using a TP2104 MOSFET. If I adjust it to come on at say 3.5V it doesn’t turn off at 3V, but rather just gets dim.

I would most welcome any ideas or suggestions.

Thank you.
 

Attachments

  • TL431.png
    TL431.png
    5.1 KB · Views: 9

bertus

Moderator
Nov 8, 2019
3,303
Joined
Nov 8, 2019
Messages
3,303
Hello,

How is the led powered?
I do not see any power source for the led.

Bertus
 

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
The load is still not connected to Vcc.

There is not enough voltage to turn the MOSFET on and off, since the saturation voltage from the TL431 anode to ground is about 2V.
It should work if you use a P-BJT instead of a MOSFET.

What is the maximum output load current?
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
The load is still not connected to Vcc.

There is not enough voltage to turn the MOSFET on and off, since the saturation voltage from the TL431 anode to ground is about 2V.
It should work if you use a P-BJT instead of a MOSFET.

What is the maximum output load current?
Thanks, I'll give it a try
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
I asked about the maximum load current since that affects the transistor base biasing to get the needed base current for a given load.
Hi, the max load is 90ma.
One of the problems is if you set The reference to 2.54 it stays steady over all changes of input voltage however if you set it to something other than that it varies with the input voltage strange?
 

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
it stays steady over all changes of input voltage however if you set it to something other than that it varies with the input voltage strange?
Don't understand the reason for that unless the transistor bias is not correct, or you are still using a MOSFET.
Post the exact circuit you have.

Below is the LTspice simulation of the circuit using a P-BJT with the trip voltage set at 3.5V (green and red traces).
R4 provides a small amount of positive feedback to give about a 100mV hysteresis between the upper and lower trip voltages to prevent dithering/oscillation at that point.

1668449906048.png
 
Last edited:

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
TM1 will turn the load off, but not on.
 

Attachments

  • Low voltage disconnect.png
    Low voltage disconnect.png
    6.7 KB · Views: 3

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
Why are you operating at such a low voltage?
I though you wanted it to trip around 3 to 3.5V?

For such a low voltage the TL431 will not work, you might try the TLV431. which has a 1.24V reference.
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
The TL431 is rated 2.5 to 36V and I thought it would work at 3V. I'll try the TVL431.
Thanks!
 

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
Here's the sim with the TLV431 set for a trip point of around 3V:

1668532795301.png
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
Thank you, I ordered a TLV431, and it will be here in a week. I'll give it a try then.
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
I received the TLV431 and as predicted, it works very well with a led as a load, but when a 35mA load is placed, I doesn’t output. I supplied the circuit layout with voltages. The 2n2907 should easily carry the load, but the TLV431 does not allow the voltage to turn on the 2n2907.
 

Attachments

  • TLV431-voltage.png
    TLV431-voltage.png
    7.6 KB · Views: 4

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
My simulation of the circuit works.
What do you mean "the TLV431 does not allow the voltage to turn on the 2n2907"?
What are all the circuit node voltages when that happens.
 

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
Thank you for the help,

Here are all the readings, and as you can see the transistor base stays high, keeping the load disconnected.
What I want to happen is; as the voltage declines from 4.2V to 3.0V, and at that time, the load disconnects, keeping the battery safe.
 

Attachments

  • TLV431-voltage.png
    TLV431-voltage.png
    8.8 KB · Views: 5

gregfox

Mar 25, 2013
164
Joined
Mar 25, 2013
Messages
164
I must add re the above, I can get the load to come on, by adjusting VR1, but if I set it, say, to go off at 3.5V, it will not go back on at any, up to 4.2V or down to 3.0V. It just stays off, and if I set it to go ON at any voltage(3-4.2v) it will only dim as it get down to 3.0V, and not turn off. I thought the TLV431 would act like a Schmitt trigger, but in this case..no.
 

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
Here are all the readings,
Your REF voltage is way off.
It should be 1.24V at the trip point.
The pot setting should be adjusted until you get that voltage where you want the circuit to change states.
 
Top