Hi folks,

I'm trying to build a simple LED indicator for a hobby project, but while I know a few basic things about circuits, I'm a little out of my depth and need some help working out what's possible here.

I've found some designs online that are close, but they're not quite what I need and I'm wondering if a simpler circuit is possible. The best examples I've found so far are these two, a voltmeter, and a low voltage indicator:
http://www.electroschematics.com/5816/led-volt-meter/
http://www.rmvhf.org/wordpress/?page_id=889

What I really want is this:
- I'm testing a single cell LiPo battery, so the possible voltage range is 3V to 4.2V
- I'd like a green LED to light up if the voltage is over 4.1V (easy!)
- I'd like a red LED to light up if the voltage is below 4.1V

From my reading online I know I can create a low voltage indicator using zener diodes, but I'm wondering if a simpler design may be possible, and whether I can use the green LED lighting as a means of extinguishing the red LED.

It feels like something that should be possible, I simply don't have the knowledge to know if it is, or how to design it.

Thanks all,

Myx

 

The op amp should be connected to the 9 volt battery and ground (9 volt battery should also be connected to ground with the other ground symbols.

All parts should be at RadioShack (op amp is TL082)

The battery listed below is connected to the triangle (4.17 or 4.05 volts in images below).

Adjust the potentiometer until your voltmeter says 4.1 volts. Use a 10k ohm LINEAR potentiometer.

 

Thanks. Does this need to have a separate 9V supply though? Since we have 3V or more available at all times from the LiPo's I was hoping to power the LEDs directly from there.

 

You are using LEDs that require over 2 volts to turn on. With only a 3 volt supply, maybe somebody could polish the numbers and get it to work, maybe not.

Gopher's circuit will run for a 40 hour week on one battery. How many batteries do you need to test?

 

Here is the simlelist one I can think of.
It uses a voltage reference so it can be made accurate,
If you decide you want to build it let me know and we can pick resistors for 4.1 volts. I just got close for the simulation.

 

Thanks guys, this is a huge help.

To give you some background, this is something I'm looking to build for my hobby of RC Helicopter flying. These models are powered by LiPo batteries with 1-7 cells and up to 5000mAh capacities. It's very easy for mistakes to happen with these batteries which can easily mean either a nasty LiPo fire, or a crash. One example I saw recently was somebody who had a fuse blow on their charge board, and while that protected the battery during the charging, since they didn't realise it had blown, they then tried to fly with a pack that hadn't been charged at all.

While I've already got a battery meter, these tend to work off the total voltage which can miss some cases. They also don't work at all for the single cell LiPos I use for night flying.

So what I'm looking to do is build a simple tester that can live in my flight case. The ideal is for it to not require any additional power source, meaning it can be pocket sized and just sit there until needed. The concept is to just plug a LiPo in before any flight and ensure the voltage is good. Since I've already got a balance board adaptor that I can plug in to give me individual pinouts between every cell in the battery pack, my idea was to build up a set of 7 circuits. If I plug it in and get all green lights I'm good to fly. Any red lights at all indicates a bad cell.

The idea sounded easy, but I reckon I'm going to have 50+ elements to solder together to build this. More complex than I originally thought, but at least I now know it's actually possible

To keep the size down my thinking is to use 0805 SMD resistors and LEDs, while they're horribly fiddly I am used to soldering those for my night flying builds and should allow me to make a pocket sized tester.

Ronv, if you can let me know what resistors would be needed I'll get some parts ordered to build my first test circuit and see how I go

Myx

 

Here you go.

 

Excellent! That's so slick I thought MikeML was on the job. (The 431 seems to be a chip he's really good with.)

 

Yes, I've stolen his idea several times. It really is a simple comparator.

 

Yes, I've stolen his idea several times. It really is a simple comparator.

Can you add a potentiometer to the voltage divider so the OP can dial in the exact voltage.

 

I thought he just dialed it in to 1% in post#7.

 

I thought he just dialed it in to 1% in post#7.

To paraphrase Audioguru, simulations are based on typical chip values and, unfortunately, you cannot go to the store and ask for a typical chip - Ya get what ya get.

A trimmer just seemed easier than 1% resistors

 

Oh. I guess I learned something again today.

Using 1% resistors fooled me into thinking the chip was accurate to 1%.
My bad.

 

Hey guys,

I'm afraid my lack of knowledge means I need to ask some more questions. It seems to me that a potentiometer is a good idea, even a 1-2% variance in the circuit is going to make a big difference to the results.

I'm not sure where the potentiometer would need to go though, or how accurate it would be. On the assumption that it's the ratio between R2 and R3 that's important I would guess that I just need to replace R2. I can pick up a 100kΩ 150mW SMD potentiometer pretty easily, so my first question is would it be sufficient to drop that in place of R2 and use a multimeter to set it to around 50kΩ before fine tuning it with a known source voltage?

But after a bit more thinking, would it be more accurate and safer to use a 47kΩ resistor with a 10kΩ potentiometer in series? That seems like it would make this more of a fine tuning exercise with less risk of frying things?

These projects always sound easy before you learn about all the subtle little details

Thanks again

Myx

 

Here is the simlelist one I can think of.
It uses a voltage reference so it can be made accurate,
If you decide you want to build it let me know and we can pick resistors for 4.1 volts. I just got close for the simulation.

One more point, I am not sure the LEd is completely off. The voltage drop across a red or green LED should be in the range of 2 to 2.2 volts. If you look at the scope chart in the simulation, the drop is much less and the current change is quite low (high efficiency LEDs would be required in any case).

Also, why does it say diode 1N914 in the lower right of the schematic - that is not an LED part number.

 

One more point, I am not sure the LEd is completely off. The voltage drop across a red or green LED should be in the range of 2 to 2.2 volts. If you look at the scope chart in the simulation, the drop is much less and the current change is quite low (high efficiency LEDs would be required in any case).

Also, why does it say diode 1N914 in the lower right of the schematic - that is not an LED part number.

Never mind, I see that the 723 is essentially a bipolar comparitor.

 

Hi Gopher,
I didn't have any red LEDs so I just used 2 diodes in the simulation.
I suppose you could add a pot, but most components are manufactured to 6 sigma now days so the odds of getting 2 at the very top and one at the very bottom in the right place is like winning the lottery.
You could change R2 to a 51.1k and add a 2k pot in series; however it might still be a little touchy to adjust. For another $1 you can buy 0.1% resistors. Then you are just left with the reference. I think you can buy those down to .2% as well. But anyway right now it's set to 4.06 volts so at 1% overall that would put it at 4.1996 on the high side and 4.0194 on the low side.

 

Would I be right in thinking that if I just bought the components I could test the circuit and see what voltage it's switching over at, and swap R2 to fine tune if that proves necessary?

I could use a simple circuit for testing 1 cell batteries anyway, so the mockup has some value to me.

 

Maybe the easiest would be to buy them as is then buy one around 500 ohms and another around 250 ohms. If it turns on to soon you can add the 500 in series with the 50k, to low add the 250 in series with the 20k. You can make a little tee - pee with the resistors.
We can start a pool on the switching voltage.

 

Would I be right in thinking that if I just bought the components I could test the circuit and see what voltage it's switching over at, and swap R2 to fine tune if that proves necessary?

Somebody else's signature: One real test result is worth a hundred theories.
My point of view: Electrons can't lie. Assemble the parts and you will know the truth of it.