Led voltage indicator

[Chris_b]

Hi everyone,

I rashly offered in another thread to "put together a design.. and post it", thinking this would take half an hour with a pen and paper. Then I thought I should do a quick layout and build something to photograph, and I should make sure it can be built by anyone who can solder and wire things up. So a little later than I planned, here it is.... I'm going to do this in several posts as I can't use the picture uploader recommended to me because I use linux not windoze on my pcs  .

The spec:

- A reproducible circuit that runs off a PP3 (alkaline or rechargeable), measures voltage and displays it on a row of 10 leds
- Led current controlled so they don't blow up!
- Choice of bar display (more volts = more leds alight along the row), or dot display (more volts = lit led moves across the row)
- Basic input range 0 to 5 volts. Can be increased to anything you like with 2 resistors

Notes:

- You don't have to use all 10 leds
- observe + and - signs. The circuit does NOT measure voltages that are negative with respect to the battery negative terminal
- I've set the led current to about 10mA each (can be varied, see later). If you run from a PP3 and use the bar display option, the battery won't last too long. I suggest a rechargeable which would give you about an hour with all 10 leds on. Or you could use a 9 or 12 volt dc mains power unit eg one of the plug-top power supplies.

The circuit and layout on a piece of stripboard is attached in the picture. Parts list for the basic circuit:

a) Piece of stripboard at least 13 strips wide, 14 holes long
b) LM3914 or LM3914-1 integrated circuit (these are quite commonly available in UK-  Maplin, Rapid, etc)
c) 10 uF 16 volt electrolytic capacitor (or bigger is ok if it fits on the board)
d) 3.9 Kohm resistor **
e) 10 Kohm resistor **
f) a few inches/cm of tinned copper wire for links

optional:
g) terminal pins - useful at the edge of the board for connecting wires. Careful as you push them in or the board will break.
e) 18 pin socket for the LM3914

** these resistors set the voltage range, AND the led current (= brightness). More combinations later...

[Chris_b]

Here's pictures working in bar mode

The input range is 0 to 5 volts. As the voltage rises past each 0.5 volt interval, the next led along will light.

I used a bar led display in a socket as it was quicker to put together for testing; this is a set of 10 individual leds in a handy package. You can use whatever you like, mix colours, sizes etc. Forget about the terminal pin in the top left of the photo, the LED +ve wire just connects to the top of the strip, and the link across the board goes on the hole where the pin is (as in the layout picture earlier)

The ic is in a socket - again so I could change it easily to make sure the circuit is repeatable. Not compulsory.

Mode change requires removing/adding the wire link just below the ic in the picture. I unsoldered/lifted the right hand end for the dot picture. Note it isn't connected to the link next to it, that's just how it looks in the photo.

[Chris_b]

...and now in dot mode

[Chris_b]

Details of how to make a 10 volt version.

The basic circuit has a range of 0 - 5 volts (chosen because the input must always be a bit below the battery voltage).

To increase the range, we fit resistors RX and RY to make a potential divider on the input to the ic, and move the input wire (the mauve one) to a different place on the circuit board (labelled "alternate +ve input > 5 volts" on the layout).

For a 10 volt input, use 10 Kohms for RX and RY. In practice, almost any values will do provided they are the same, and the value isn't small enough so it overloads whatever you're measuring. As its name suggests, this divider takes 10 volts in and divides it down to 5 volts to suit the basic range.

You can set almost any range you want with appropriate values of RX and RY; I'll post equations later.

[Chris_b]

Other possibilities:

You can run the leds from a different voltage source to the rest of the circuit by disconnecting the "led +ve wire" from the circuit board and connecting the other voltage source +ve here (its -ve connects to the battery -ve pin at the other end of the board).

This reduces the battery current to just a few milliamps so battery life will be much longer. Provided the other voltage is above about 3 volts, the leds will light. Note they won't change in brightness as the voltage rises above this, because the LM3914 passes a constant current through them, regardless of the voltage. Beware that the LM3914 will get hot and eventually burn out if you use a voltage that's too high. What's too high you ask? Well, it depends on how many leds are lit and for how long. I wouldn't suggest more than 15 volts in bar mode where you might have most of the leds lit. In dot mode you could probably use up to 20 volts.

Varying the led current

As I said above, R1 and R2 do two things, and they are inter-related. They set the voltage range and they also determine the value of the constant current that is passed through a led when it's lit. I calculated values according to the data sheet and the led current was more than specified. I don't know why this is, and I tried several LM3914 bought from different places (so different manufacturing batches) with the same result. So I calculated sets of resistors R1 and R2 that gave decreasing led currents whilst keeping the 5 volt range.

R1 = 1.2Kohm; R2 = 3.6Kohm  led current about 18mA
R1 = 2.2Kohm; R2 = 5.6Kohm  led current about 14mA
R1 = 3.9Kohm; R2 = 10Kohm   led current about 10mA

If you wanted something seriously bright, then use the smallest set of values (and a big battery or a mains power supply!).

Note that high-intensity leds are very bright at the lower current, though they have a narrower viewing angle than ordinary leds.



If you wish to choose your own input range that is greater than 5 volts, I suggest the following

Use 10Kohm for RY

Set RX (in Kohm) = 2 (Vin - 5)   where Vin is your desired maximum input voltage



Hope this is of interest to some folk. Do post if you make one and it works (or even if it doesn't!)

Chris

[Darren]

Hi Chris,
That looks great and I may just make one at a later date for my valve amplifier when the time comes. Thanks for posting all the info, whilst I play with valves ok with ic's they just do my head in 

I tend to blow them most of the time. I know, I know, it's a lack of education 

One question, can this circuit run off the same power supply it's measuring?

[Chris_b]

Hi Darren

Valves are good :-) you could use it as a level meter if you rectified the audio. But I'm not offering to design one of these with valves 

As far as running off what you're measuring, the answer's not really. You need at least 3 volts on the supply before there's enough to turn the leds on at all, and the maximum input voltage has to be less than the supply by a volt or so. I just tried my (10 volt) circuit on a variable power supply with the input connected to the supply. It lit random blocks of leds as the supply volts rose until it got to about 8 volts, then worked ok after that.

Chris

[Darren]

OK ok then,

I had an idea to monitor the 10V that power the heaters on a valve amp. It will remain pretty constant though.

Still I could add another PS transformer without much problem.

[Bernd]

Hi Chris,

Nice circuit design. I've played with small circuits like that.

I have a light level indicator that I built from a kit. It works pretty good. I'll have to find it and post a pic of it.

Bernd

Instead of cluttering up this thread with another post I'll modify this post and post the pics of the light level indicator in this post.

This is the light level indicator. As you can see it uses an LM339 comparator chip and lights the 5 LED's, four red and one green. The back side holds the battery which is about 3/4" in diameter. The chips are what is known as surface mount chips. I assembled it from a kit. I really like soldering. My dad taught me that at a very young age.

Component side and trace side.


Back side.

[sbwhart]

Nice one Chris

Those cercuite things are beyond me:- much o respect o 

I'm shure the chaps will put this information to good use.

Cheers
 
Stew

[PMK]

Hi, Chris_B.
Apologies for reviving an old'ish thread (new member here), but your circuit caught my eye. I've just breadboard'ed your design - works a treat. I made a similar gizmo a couple years ago. The main difference is that the unit is self-powered by the battery that it's monitoring (no separate supply needed). See http://www.pm.keirle.com/2_test/esv/esv.htm
Also, instead of the 10-digit display, a couple circuits shown below using just four LEDs. The 1st ckt uses 3x fixed-value 1% res's; the 2nd ckt uses 2x trimpots.
Designed primarily for the model plane/boat fraternity, can work on any number of cells up to 25 volts max.

Hope these sketches make sense.......

[Chris_b]

Hi PMK

Nice collection of circuits you've got there, this thread could end up as the definitive reference for these!

The original challenge (which I failed to meet entirely) was to produce a self-powered circuit that would indicate the output from a dc generator driven by one of the amazing engineering creations you see on this forum. Folk had tried some basic led circuits and suffered from vapourised leds or a system where the leds didn't come on at consistent increments of input voltage. With a self-powered LM3914 you'll not be able to measured voltages much below about 4 volts, so I cheated and used a battery.

Designing with the LM3914 is not entirely straightforward as the function that set the led current/brightness is mixed up with the function that sets the voltage corresponding to FSD "full scale deflection" or all leds alight. As you'll see from my notes, my circuit did not produce the current I expected so I added a list of resistor values to produce useful led currents.

My circuit also included a potential divider on the input, so the range could be extended above the battery voltage, and on reflection (hindsight is always 20/20) I should have simplified the circuit by fixing the FSD at the absolute minimum which is 1.25 volts by connecting pins 6/7 to ground with a single resistor (that would set the led current) and grounding pin 8.  Then RX and RY could heve been chosen to set the FSD. That way the way you set FSD is independent of the way you set the led currents. It would also have meant the circuit, if self-powered, would have worked down to the lowest possible voltage; mind you, at 4 volts or so it still might not have been low enough for some folks. With a separate battery, and the 1.25 volt FSD above, the first LED would com on at 0.125 volts.

If anyone is interested in the revised circuit, shout here and I'll update the design above and repost.

Chris

[PMK]

"With a separate battery, and the 1.25 volt FSD above, the first LED would com on at 0.125 volts."

Absolutely. That's the reason why your battery-powered version is ideal for measuring my single cells.
In my case, the unit had to be self-powered because, down at the flying field, they wanted it that way. They don't like parting with their ££ for more batteries and it had to be made reasonable small (read weight). They don't like current-hungry gadgets, neither. It's the same deal with the model boating fraternity. So having a 4-led version pulling just a few measly mA is music to their ears.
The '3914 is a great little chip - ideal for these sort of gizmos. I initially started out with using quad op-amps but soon found that the '3914 takes much less space on any given Vero layout. And pulls less current, to boot. Incidentally, if you want to manually vary the brightness of the LEDs, and if you can put up with a slightly larger board size, you could swap out the fixed res' on pins 6 and 7 with a 5k trimpot (another tweaker to play with).

Nice to meet you, Chris.
I'm sure I'm not the only one interested if you want to post your revised version.

<edit>
Forgot to say...

Your issues regarding R1/R2/current draw: I too noticed the discrepancy on the data sheet. Again, in this case, superbright LEDs are probably the better choice because you can skim as much mA as you dare in return for brightness. In fact, the photo above, it lies. The actual LED intensity is more than the camera implies. You could probably hike R4 up to as much 5.6k, maybe more, all depending on how good your eyesight is/whether the sun shine directly on the LEDs, etc. The higher the res' means the more mA you save.
I see that it's now possible to buy 10-digit displays in the superbright flavour. This is good because your R1 and R2 could be recalculated, and likely end up squeezing a bit more milage from the PP3.

[Bernd]

PMK,

I know were I can use that circuit. Going to have to build one for my two six volt battery powered "Rescue Boat".

Thanks for that design.

Bernd

[PMK]

Hi, BernD.

Is there a chance you might have a schematic for your light level indicator? I'm trying to eyeball the tracks in the above photo but it's a bit of a hit-and-miss affair.

[Bernd]

PMK,

If there was a schematic I don't think there's a chace of finding it since it was a kit. I've had this for more than 10 years so finding a schematic is nil at best. I'll see if I can get a better pic of the tracks. Look for something here Monday. Won't have time tomorrow to do it.


Bernd

[PMK]

BernD,

Understood.
I guess I should have realised sooner - being that it was a kit and all. Sorry about that.
A better picture of the tracks might help, but pse don't bust a gut on my account.

By the by....
While you're there, the guy sitting on the tractor in the photo on your homepages, I presume is you, yes? Your face is familiar. Do you play guitar by any chance? Where have I seen you before?

[Bernd]

BernD,

Understood.
I guess I should have realised sooner - being that it was a kit and all. Sorry about that.
A better picture of the tracks might help, but pse don't bust a gut on my account.

I tried to get closer with the camera and get a better picture, but no way. The digital I'm using just won't let me get close enough for a nice clear picture. I might run across the schematic a some point.

If I recollect this came from a kit sold by the now defunct Heathkit company. I believe Zenith bought them up. I have many of there kits I put together. This was a kit when SMD's first came out. They sold it along with other equipment to people who were interested into getting into soldering SMD circuit boards.

By the by....
While you're there, the guy sitting on the tractor in the photo on your homepages, I presume is you, yes? Your face is familiar. Do you play guitar by any chance? Where have I seen you before?

Yeah, that's me alright about 10 or so years ago. No I don't play guitar, can't even carry a tune. If you think you've seen me before the only way you could have is to have come over to the states. I've had people tell me I look like "Billy Crystal" at one time. 

Bernd

[PMK]

I've been to the States, but I'm sure it wasn't California. So obviously, there's another BernD look-a-like out there somewhere.
Like I said, don't bust a gut over the photo. Appreciate you having a stab at it just the same. Tnx.
I probably would have built the thing with bog-standard parts anyhow (not really a fan of SMDs).
I guess I live in a world where Heathkit is still king.

Thanks once again.

[John Swift]

Hi Bernd ,

                  I've just been looking at the light meter pcb
I expect it to be some thing like the circuit I've drawn from your pictures

                                                John

[John Swift]

now traced circuit

this is more like it


well may be not - the leds are connected to the battery positive !
just incase you didn't spot it
should of redrawn the circuit instead of modifying it on the pc  
pencil and paper 1   pc 0