12v Battery Low Voltage Monitor

[JimmyJazz]

Hello

I have an old car with an alternator with no wiring provision for a charge warning light. I thought that maybe the circuit below would simulate this as I don't want to fit a meter.



1. Would the 100K variable resistor be OK to set 13.6v.

2. I think I would also need to isolate the circuit from the alternator output using a filter, otherwise the electrical noise from the alternator would kill it. What fillter would I need and where would I place it in the circuit.

Thanks for any help

 

[Hero999]

1) Yes, you could do that, or you could use a couple of fixed resistors so you don't have to tweak it.

2) Yes, put a 10µF capacitor and 100nF capacitor across the power leads.

13.6V sounds too high, I'd suggest a lower voltage, maybe 10.5V?

There should also be a feedback resistor from the output to the non-inverting input.

I haven't checked whether you've got the pins connected correctly. The schematic should have the inverting and non-onverting inputs labelled rather than just providing pin numbers which aren't very informative.

I'd also recommend replacing the 741 with a real comparator IC, such at the LM311 or LM393. The 741 isn't really designed to be used as a comparator, although it can be used if it's the only IC available to you.

 

[crutschow]

1. 100k is OK but 10k is preferable. Also add 10k in series with the top lead of the pot.

2. Add a small resistor (≈100Ω) in series with Vin to the op amp and 100µF from the junction of the resistor and op amp to ground.

A battery is pretty well dead at 10.5V. I used about 13V to turn off a car seat heater when the engine stopped and that seemed to work well. Just adjust the pot to barely turn off the LED when the engine is stopped.

 

[MikeMl]

The 741 is not suitable. It makes a poor comparator. It is not tolerant of transients in an automotive environment. Here is better circuit:

I would set the trip point to about 12.2V. If a fully charged battery is disconnected from the alternator, or if the alternator quits charging, the battery voltage will almost immediately drop to 12.6V. You only want the LED to light if the battery voltage drops further, probably less than 12.2V, which would be an indicator that the alternator has been off for a while. You don't want the alarm to light at idle....

 

[Hero999]

Where's the hysteresis?

That circuit will oscillate when the voltage is near the threshold.

You're right that it needs to be set higher than 10.5V because you want it to indicate that the alternator isn't working. I was thinking about the kind of low voltage cut-off for an inverter.

 

[MikeMl]

Where's the hysteresis?

That circuit will oscillate when the voltage is near the threshold.

You're right that it needs to be set higher than 10.5V because you want it to indicate that the alternator isn't working. I was thinking about the kind of low voltage cut-off for an inverter.

If you are talking about the TL431 circuit, it will not oscillate. It draws much more current when the LED is on, which is also when the input voltage is low, meaning that if there is even the slightest resistance in the wiring, the hysteresis is provided by the wiring resistance. I have built several of these, and there has never been any oscillation.

 

[JimmyJazz]

Thank you for all the replies. Being a novice to electronics I am afraid I don't fully understand your circuit diagram. Would it be possible to list all the components need so as to give me a clearer Idea where they all go or maybe a more "idiot proof" diagram. Sorry for the dumb questions and lack of knowledge, but I hope you'll forgive me while I learn.

Alan

 

[Hero999]

If you are talking about the TL431 circuit, it will not oscillate. It draws much more current when the LED is on, which is also when the input voltage is low, meaning that if there is even the slightest resistance in the wiring, the hysteresis is provided by the wiring resistance. I have built several of these, and there has never been any oscillation.

Yes, you're right, that's absolutely true, the increase in current reduces the voltage and provides positive feedback.

 

[JimmyJazz]

Are you saying then that this circuit is OK for my application or indeed better than the one I suggested. Will it need the noise filters adding. I have looked again and I am sorted with all the stuff in Mikes circuit accept the reference to O1, not sure what that means and finally is a 42.2K resistor and unusual size, it seems a bit difficult to source.

 

[Hero999]

Yes, it is better better than the one use designed.

42.2k is a standard E96 precision resistor value but I don't see why it can't be 43k as a pot for adjustment is being used anyway.

 

[JimmyJazz]

Thank You, and finally do I need to input the extra resistors/capcitors as suggested in the previous post to filter the noise

 

[MikeMl]

Thank You, and finally do I need to input the extra resistors/capcitors as suggested in the previous post to filter the noise

No. The circuit will work as designed. If you use other than the 42K resistor, you may have to use a 2K trim pot instead of the 1K depicted to be able to hit your selected trip point.

The "REF" is just a node label which is used to probe that node during simulation. If you look at the datasheet for a LM431 or TL431, you will see that one of the pins is called REF, so that happens to be that input...

 

[JimmyJazz]

Sorry to be so stupid but electronics isn't my thing and I'm a real learner

Can someone point out the error of my ways or help me with the circuit I have tried to build as per mikes diagram



Apologies for the rudimentor pic, but I cant think of another way to show it

Alan

 

[MikeMl]

I cant quite tell from your picture, but I see a couple of things:

It looks like the 4700Ω(4.7KΩ) resistor might be 4.7Ω?
It looks like the bottom black jumper misses connecting to the anode of the LM431?
What is the value of the resistor in series with the LED?
Is the cathode of the LED grounded?

 

[JimmyJazz]

I cant quite tell from your picture, but I see a couple of things:

It looks like the 4700Ω(4.7KΩ) resistor might be 4.7Ω? Yes wrong resistor silly mistake
It looks like the bottom black jumper misses connecting to the anode of the LM431?This is in the right place I have just shown it wrong on picture
What is the value of the resistor in series with the LED? 470Ω
Is the cathode of the LED grounded? LED Cathode goes to -tve supply is this correct?

Thanks for the guidance I will get the correct resistor and report back

Alan

 

[JimmyJazz]

So I have got he correct resistors now and I still can't get it to work. Is there anybody out there that could show a simple picture or build an example of this post it up so someone as STUPID as me can see what to do as I am obviously missing something.

 

[Diver300]

Remove the TL431 for a moment to test the circuit. The LED should light.

Current should flow though R1 and turn on Q1, so that more current is available to flow though R5 to light the LED.

If the LED does light, go to (X) below.

If the LED does not light with TL431 removed, bypass Q1 (collector to emitter) with a wire. If the light now lights, you've got R1 or Q1 wrong. If the led does not light, it's faulty, the wrong way round or R5 isn't connected, the wrong value or faulty.

(X)
Connect the bottom of R1 to ground with a wire. The LED should go out. If not, you have Q1 fitted wrong or it's faulty.

Assuming now that the LED lights, and goes out when you short R1 to ground, the problem lies with the TL431.

Replace the TL431, and remove the potentiometer. Connect the reference input of the TL431 to ground, and the light should be on. Connect the reference input of the TL431 to the cathode and the light should be off. If that's not working, you've got a faulty TL431 or it is wired wrong.

Once that works, connect the potentiometer. If the LED is on all the time, R8 is too big or not connected, or R6 is too small or shorted.

I hope this gives you some clues.

 

[mneary]

Your trim pot leads are not in the correct rows. Two of them are in the same row and thus shorted together. The resistors that connect to the ends of the trim pot must also be in the correct row to connect with the end points of the trim pot.

 

[MikeMl]

Here are two pictures, showing LED off and LED on. For simplicity, I left out the fixed resistors above and below the pot, and used a 50K trim pot instead. I arbitrarily set the trip point to 12.00V. The gain of the TL/LM431 causes the LED to go from fully off to fully on with about a 10mV change.

If those resistors are not the values shown in the schematic, you may not be able to hit your desired trip point. Temporarily remove the LM431, and set the Ref input to 2.500V when the input voltage is 12.00V. Then hook up the LM431, and sweep the input voltage from 11.5V to 12.5V; that should switch the LED on/off.

 

[MikeMl]

Here are two more sims. The pot is centered. First, 12.70V = LED ON node voltages, followed by 12.75V = LED OFF node voltages. Should help you debug yours. Use a DVM to compare voltages in your circuit to the ones coming from the sims.

Oh, and by the way. If you powered your LED from 12V while you had only a 4.7Ω resistor in series with it, you might have vaporized the LED, and possibly the NPN transistor.