Eight LEDs make a 100-division voltmeter
Use a counter IC to drive two sets of four LEDs.
Raju Baddi, Tata Institute of Fundamental Research, Pune University, Maharashtra, India; Edited by Paul Rako and Fran Granville -- EDN, September 22, 2011
The circuit in this Design Idea makes a voltmeter that reads to 0.99V. The idea uses a counter IC to drive two sets of four LEDs (Figure 1). Each of these two sets represents a BCD (binary-coded-decimal) value. With all of the LEDs off, the voltmeter reads 0V. With all of the LEDs on, the reading is 0.99V. Op amp IC1A generates a predictable voltage ramp.
Click to enlarge
You use op amp IC1B as a comparator
to compare the ramp to an input signal.
The higher the input voltage, the longer
the output pulse from IC1B is. You use
this pulse to gate free-running oscillator
IC2B. A potentiometer on this multivibrator
circuit allows you to adjust the
full-range count. The voltmeter has a
maximum input of 1V and uses three
dual-part packages. You make output
counter IC3 work as a two-digit counter
by strapping the enable pin of the IC3B
part to the MSB (most-significant-bit)
output of the IC3A part.
A dual op amp is used to create the
comparator function and the ramp
generator. The design also uses a dual
555-type timer chip. You use IC2A to create the ramp and to reset it and
the output counter, and you use IC2B
as a free-running oscillator that drives
the counter chip. To blank the output
LEDs when the chip is counting,
Q3 disables drive current to the LEDs
when IC3 is incrementing. You use IC4
to derive a reference of 2.5V.Tests of the design use TL084 op amps, but you can also use an LM358. A top view of the 15×15-hole prototype board shows component placement (Figure 2a). Figure 2b shows a bottom view of the board, with the connection and three resistors. You might use flat-green LEDs with the sides painted black or covered with black-plastic sleeves for good visibility.
Click to enlarge
Talkback
-
This idea is a replacement for moving coil meter. But of
course in conjunction with an OPAmp and/or RC filter.
-Raju Baddi
Raju Baddi - 2011-15-10 02:29:34 PDT -
Please Note:
1) This circuit works with a single supply of 5 to 5.5V.
2) Input voltage is with respect to the reference voltage of 2.5V.
3) The -2.5V should be 0V.
Further if there are any queries perhaps refering the
circuit layout may help.
Raju Baddi - 2011-14-10 09:32:08 PDT -
1. The Ramp (output of IC1A) will be in the range from min=2.5V if Q2 is closed to max=~3V (Q2 is open, C3 is charged and IC1A output is saturated). The input voltage coming to IC1B according to the IFD is 0..1V. So the comparator IC1B will never see the inputs cross (one input is within 0...1V, another within 2.5V...3V)
2. If, by some magic, IC1B sees the inputs cross and IC1B changes the output -- most likely, without a hysteresis (positive f/b across a comparator) there will be numerous transition between gnd and Vcc --that will immediately bust the counter.
3. As it was already mentioned in some comments -- the idea is over 30 y.o.
Samuel Kerem - 2011-12-10 14:32:07 PDT -
Forgot to mention why this is an interesting design for Battery Cell monitoring: in a typical Electric Vehicle, the battery pack might consist of 100 cells in series, which means that hundreds of volts exist between the top and bottom cell, so one of the ways to connect them to a central display is using opto-couplers. This requires digital signals as most opto's are not good in linearity or thermal stability, but looking at this design you will recognise that the two signals (Clock and Enable) feeding into the counter are digital, so this is easy to transfer using opto-couplers between the analog cell monitoring front end (the opamps) and the digital display. Just an idea...
BTW, you also need to incorporate the change that Hans J Weedon already pointed out, the counter pin 9 must go to ground as there is not even a -2.5V reference voltage - apparently the designer has flip-flopped the design idea between using symmetric 2.5V and a single 5V, with the resulting errors still remaining and the unspecified input voltage reference that I commented on in my other post.
Cor van de Water - 2011-5-10 14:03:06 PDT -
I think the circuit as presented will not work because the datasheet says that the CD4518B counter will advance when either the Clock is low and Enable gets a falling pulse or when Enable is high and Clock gets a rising pulse.
Since the ramp generated starts at 2.5V and rises, the input opamp will generate a high signal on the Clock input, disabling the counting during the measurement period. You want the + and - inputs of the input opamp (acting as comparator) exchanged, so the output is low during the measurement period and the counter will actually count until the ramp gets to the same level as the input signal and the output of the opamp flips high, halting the counting and showing the BCD output on the LEDs.
NOTE1: this means that the max output of 0.99V is shown when not all LEDs are on, each segment only counts to 1001 (binary).
NOTE2: the input pin is not referenced to ground, in fact there is no reference shown where the input is measured against, but as shown it is against the 2.5V reference voltage is the "negative" input, not ground.
NOTE3: Since you are using LM358 which does not need an offset as the input range includes the negative supply voltage level (ground in this design) it is much more convenient to use no offset for the input and only give the ramp generator a small (half division = 5mV) offset to make a positive-going ramp, of course this means a divider on its positive input and a smaller resistor for the RC circuit.
NOTE4: the circuit as presented is good to measure the useful range of the newer LiFePO4 batteries, typical voltage levels are from 2.5 to 3.5V which satisfies the design as presented. But if you want higher input voltage then it is a matter of increasing the supply voltage and if you want a larger range then it requires either dividing it down or using the ground-offset that I described before, this gives you 3.5V input range on 5V supply.
BTW, if you need LM358 series opamps, I have a reel of SOIC-8 LM2904 somewhere around here. Great if you start a Battery Monitoring project or so.
Cor van de Water - 2011-5-10 13:26:30 PDT
