Help on a LM3914 Voltage indicator 14V to 24V (1 LED per Volt)

Thread Starter

mohala

Joined Nov 30, 2017
14
Hello,

I was wondering if someone could give me a circuit design which would work from a voltage line from 14V to 24V.

I tried the a couple of the ideas online but all of them seem to fall short at around 19V. One of them was youtube video I have linked here:
I noticed another thing with the indicators if I wanted it in Dot mode some voltages I would get two dots and not a single dot.

Ideally I was hoping to achieve a Battery/DC Voltage Rail Level Indicator from 24V to 14V with each 1-volt range is a LED

e.g:
14-15 - LED 1
15.1-16 - LED 2
16.1-17 - LED 3
17.1-18 - LED 4
18.1-19 - LED 5
19.1-20 - LED 6
20.1-21 - LED 7
21.1-22 - LED 8
22.1-23 - LED 9
23.1-24 - LED 10

(Sorry for the long-winded example it felt cathartic to write that out)

Hope to hear from you soon!

Mo
 

bertus

Joined Apr 5, 2008
20,051
Hello Mo,

Do you have a schematic of the circuit?
You will have to scale down the signal to get it within range of the comparators.
The LM3914 has a low and high input for the comparators.
You could adapt the circuit to cover the wanted range.

Bertus
 

ericgibbs

Joined Jan 29, 2010
8,733
I wanted it in Dot mode some voltages I would get two dots and not a single dot.
hi mo,
So you want the 1st LED to come on at 14Vin and the 10th LED to come on at 24Vin.?
As requested please post the diagram of the problem circuit.
E
 

Thread Starter

mohala

Joined Nov 30, 2017
14
Sorry here is the circuit:



I forgot to attach with with the youtube link.

Also, that what I'am confused about this High and low Input for the comparitors, would there be a equation to work out what the resistor would need to be for the High and low?

I tried using it with the 10K variable resistor I found in an arduino kit but its just too fiddily and was hoping to get set resistors to work.

Mo
 

Attachments

eetech00

Joined Jun 8, 2013
1,660
Hi

Including Bertus Suggestion..

Battery_Monitor_14-24V.png

Edit: The LED range is 14-24v.
The first LED illuminates at about 14.1 volts.
 
Last edited:

Thread Starter

mohala

Joined Nov 30, 2017
14
Oh wow thank you to you both I will get back to you how it goes.

For 14K Ohms Resistor I had to put 12K and 2K in series as that's what I had in my cupboard, unfortunately maplin is shut for good, so no more popping down there any more.

I just put a quick pick of what resistors I have at the moment.

Mo
 

Attachments

Thread Starter

mohala

Joined Nov 30, 2017
14
Hello,

I'm back I put it on bread board to test it out. I put my power supply to automatically start from 12V and go up in increments of 1V it seems that the first LED only activates at 17V and when it reaches 24V all LED are active.

When putting the LM3914 to dot mode (to ground) similar issues arises but it doesn't reach the last LED (LED 10). Also, for some reason LED 1 is dimly lit.

I have attach image I'm not sure as to why.

P.S After futher checking on my PSU, its about 16.5V where the first LED activates.

Also, seeing at it only activated at 17V should I reduce the R3 Resistor to 10K?

Mo
 

Attachments

eetech00

Joined Jun 8, 2013
1,660
Hello,

I'm back I put it on bread board to test it out. I put my power supply to automatically start from 12V and go up in increments of 1V it seems that the first LED only activates at 17V and when it reaches 24V all LED are active.

When putting the LM3914 to dot mode (to ground) similar issues arises but it doesn't reach the last LED (LED 10). Also, for some reason LED 1 is dimly lit.

I have attach image I'm not sure as to why.

P.S After futher checking on my PSU, its about 16.5V where the first LED activates.

Also, seeing at it only activated at 17V should I reduce the R3 Resistor to 10K?

Mo
Hi

Make sure all voltages are nominally correct. Refer to schematic in post #6. Make adjustment in the order shown below.

1.Set the power supply voltage to 24v

2. Adjust pot P3 until signal at pin 5 is 6.0 volts.

3. Adjust pot P1 until all leds just turn on. Voltage at pin 7 should be about 6 volts.

4. Test by slowly reducing power supply voltage in 1 volt step. Each led should turn off at each 1v step (each led won’t turn off exactly at 1v...but close).

eT
 

-live wire-

Joined Dec 22, 2017
887
Also, batteries do not discharge in a linear way.
Lithium ion:

So even if there are a lot of cells, it's almost impossible to tell between 80 and 20% with 1V resolution. You'd need to be accurate down to 10s of mV, which is certainly doable, but you need better resolution. To get a difference of a few volts, you'd need 100 cells in series (370V)!
If it's SLA:

It's more linear, but still best to not use something that's linear.

NiMH:

Pretty terrible, like lithium. And, like the other batteries, it also needs load based adjustment.

So if you need to know the actual percent of a battery, it gets a little complicated. Maybe consider a microcontroller with a few buttons to select the battery type and discharge rate. You can get a cheap buck boost converter, arduino nano, and pushbuttons. Or at least choose the references based on the battery you are going to be using.
 

ericgibbs

Joined Jan 29, 2010
8,733
hi Mo,
You say you want the 1st LED to come on at 14Vin and the 10th LED to come on at 24Vin, in 1V steps.?
You may have 10 LED's but there are only 9 voltage intervals.!
Thats one reason you have the double LED as the voltage is increased.
E
 

DNA Robotics

Joined Jun 13, 2014
500
I had that problem with the first LED staying on when it shouldn't in dot mode. I think the solution was to put a resistor on that LED that would make it not noticeable unless it is intentionally on.

If someone has a better solution, I would like to hear it too.
 

Thread Starter

mohala

Joined Nov 30, 2017
14
Hello,

Thank you for the advice and guidance eetech00 and DNA Robotics. It's now working as expected. And yes you are right about that issue with double LED ericgibbs.

Doing a bit more looking into it live wire you are correct, and I'm a bit annoyed to myself why I didn't think that the battery was non-linear when it discharges, but the circuit is still good for another part of the project which again I will need to test that out when the time comes (also that part is linear as an output)

As for battery indicator the minimum requirement for this project is a low battery indicator, thought of going one step further and do a range. Would you say something like a single comparator would work or just remove all but the first LED on the LM3914 and put it on dot mode?

I think my answer is that LM3914 is overkill for single low battery indicator.

Mo
 

ericgibbs

Joined Jan 29, 2010
8,733
hi,
If it is permanently connected 'low' battery indicator, it will have a current drain on the battery.
So the lower the current drawn by the indicator the better.
E
 

Thread Starter

mohala

Joined Nov 30, 2017
14
I was googling something about voltage monitors, im gonna have a further check on this but they seem to be very low current draw circuits.

Mo
 

Thread Starter

mohala

Joined Nov 30, 2017
14
Ok a bit more searching I have found that TPS3711 which has a low quiescent Current 7 μA, might do the trick!
It uses a voltage divider circuit on the sense pin and runs a open collector at the output.
So I guess if my battery is a 18V Li-Ion and max discharge is 15V, I could set the TPS3711 to sense about 10-30% of the battery remaining (which I will have to check on the graph from the manufacturer) what that voltage is and set the LED with its current limiting resistor to that voltage.
2018-09-08_19h44_54.png 2018-09-08_19h50_42.png
 

ericgibbs

Joined Jan 29, 2010
8,733
hi mo,
That looks OK for the project, note the limit of 10mA on the Out pin, for driving the resistor/LED 'low battery' indicator.
E
 
Top