hi there,

I have been testing the TL431 in my circuit which latches a relay once the voltage is lower than the set point. I am using a feedback resistor for a hysteresis loop so it gives me a high and low cut off point.

The circuit itself is pretty damn accurate, however when the voltage trips the low set point, the internal resistance seems to bring the battery voltage back up to 12V+ and over the high set point of the hysteresis loop.

This isn't very good as the circuit latches and switches itself back on again. Is there any way to totally cut off the circuit to stop it latching back on again once it has tripped at the low limit?

Another problem with this is that i can't switch on the circuit unless the battery is fully charged/voltage is pulled right up passed the high trip point.

I could do with the circuit being able to manually latch/turn on the circuit at any voltage point and cut off at the low limit and not switch back on until the user switches it off etc.

Any one enlighten me with any more ideas?

 

hi kh,
If you post your complete LTSpice 'asc' file for this circuit, which shows the component values, I will try it in simulation.
E

 

And while at it, consider the purpose of the hysteresis and whether you have enough.

 

sorry for the late reply. Had a lot on recently.

I thought it might be good to have hysteresis since there was a low and high limit to cut off but i dont think its entirely what i want. Even without the feedback resistor, the relay doesnt trip off and stay off. it just constantly turns on and off at a ridiculous speed.

I need something precise for cut off so when the battery is under a certain level it will cut off and stay off until you switch it back on but i am all out of ideas :/

I have attached my circuit anyway (although i have changed a few values and things to what i am using). I dont have a great deal of info on how to design other types of cut off circuits either. Ideally i dont want to use a relay because they wear out quite easily compared to mosfets with no mechanical parts.

 

So does power to your circuit get cutoff when the relay opens? That would cause oscillation however much hysteresis you add. Assuming you don't then in the "cutoff" state the battery still has to drive the relay coil. As a minimum then after increasing the hysteresis I would put some resistance in series with your coil shunted by a large electrolytic. That way the relay sees the full 12V to pull in but holds in the open state with less current.

Really this is best done with a latching relay so all load can be removed. You can buy such relays fairly easily. In the past I have modified magnetic circuit breakers into latching relays but it's very fiddley to do.

 

Soon as you push and latch an on switch to supply 12v the relay engages to power the rest of the system. once the voltage dies off to the set point of the resistor divide for the tl431, the relay trips out and turns shuts off (normally closed?).

But i need it to stay off until the user can turn it off properly using the latch push switch to go recharge the batteries up.

 

Here is one I posted on ETO years ago: This one doesn't need hysteresis because I use a manual push-button to start it.

The disconnect trip point is shown at different pot settings.

 

It sounds like your battery is "bouncing" back. That might be because you have a large load on a small battery so the voltage sags quite a bit. You could move the power for the circuit to the load side of the relay contacts. This would make it stay off. Then you would need a switch across the relay contacts to turn it on again. Or you could make the low trip point lower. Where do you have everything set now? What is the battery and the load?

 

Here is one I posted on ETO years ago: This one doesn't need hysteresis because I use a manual push-button to start it.

The disconnect trip point is shown at different pot settings.

I take it the push button starts the system but it won't stop until the battery voltage is lower than the set point?

I need to be able to turn it on and off as well as trip off while the system is on. i guess all these ideas will soon add up

 

It sounds like your battery is "bouncing" back. That might be because you have a large load on a small battery so the voltage sags quite a bit. You could move the power for the circuit to the load side of the relay contacts. This would make it stay off. Then you would need a switch across the relay contacts to turn it on again. Or you could make the low trip point lower. Where do you have everything set now? What is the battery and the load?

Yeah it's definitely the voltage that bounces back over a period and latches the circuit back on. I use a 10 cell 12v 4.5Ah and also a 10cell 12v 9aH battery. TL431 was set to trip and cut off at 9V on the battery.

I was thinking earlier, is there something i could use such as a 4013 d-type flip flop where the on switch will latch the circuit and then when the tl431 senses the 9v cut off the d type will latch the relay or mosfet low and hold it low until the system is turned off resetting the whole system, to then recharge and turn on and off at any point?

Its just the main problem i have is not holding down the switching component to the off state so the load doesn't kick back in and start running again.

my layout is battery > cut off circuit > buck-boost converter (for stable output) > Load circuit

yet the buck-boost and load circuit combined is really the load circuit for the cut off circuit.

Any ideas?

 

Here is one I posted on ETO years ago: This one doesn't need hysteresis because I use a manual push-button to start it.

The disconnect trip point is shown at different pot settings.

Hi Mike,

You posted the circuit on the tl431 battery cut off.

My guess is that the momentary switch can switch the mosfet on to allow the voltage from the battery to be sensed by the tl431 by the divider causing it to trip out.

Is there any way of having the circuit a manual on/off power switch to start the tl431 battery cut off circuit and still automatically trip out and turn the mosfet off even though there is still the battery voltage at drain of the mosfet? Until the user manually switches it off of course.

I am racking my brains as to have a manual on/off with auto cut off which will hold the mosfet or relay low until it is manually switched off completely. I have looked at many different tl431 circuits and nothing stands out to the way i want the circuit to operate.

Any suggestions?

Regards,

Martin

 

Using your circuit. Move your circuit to the load side of the relay contacts. Then place a momentary switch across the relay contacts. This becomes your start switch. Then break R2 into 2 resistors say 22k on top and 10 k on the bottom for the same total of 32k. Then put another momentary switch from the junction of the 2 to ground. This will become the off switch.

 

use another relay. a small reed relay that is in the self latch of the relay you are using.

When the self latch circuit is broken by the reed relay, the original relay will lose its latch and nothing the reed relay does will matter until the circuit is reset.


I will draw a diagram and post it in a few minutes.

 

 

The circuit I posted back in post #4 of this thread is already a "latch"; meaning that a momentary push of the "start" switch will connect power to the load provided that the battery voltage is above the trip point. The load will be on until the battery voltage drops.

If you want the ability to "unlatch" it, then add a second momentary normally-closed push-button in-series with the battery lead. Pushing that button will turn the load off and reset the latch.

This will give you the ability to start it, stop it manually or it will stop automatically when the voltage drops...

 

The circuit I posted back in post #4 of this thread is already a "latch"; meaning that a momentary push of the "start" switch will connect power to the load provided that the battery voltage is above the trip point. The load will be on until the battery voltage drops.

If you want the ability to "unlatch" it, then add a second momentary normally-closed push-button in-series with the battery lead. Pushing that button will turn the load off and reset the latch.

This will give you the ability to start it, stop it manually or it will stop automatically when the voltage drops...

Hi MikeML, Thanks for the reply. Is there away to do this electronically without the use of 2 hardwired switches using only a single switch?

 

This would need a bit of work to sort out but i get the general idea. I wouldn't really have so much of a problem if i could just get the tl431 circuit to hold itself low once it trips off the battery voltage.

That circuit would still be draining off the battery of the low voltage cut out relay as the cut off circuit would still switch on but wouldn't fully be on until someone pushed the mom switch would it not?