Digital multimeter

[mafrasm]

dear friends,
i have been asked to build a digital multimeter as a project,so as a start i thought to collect some information on current commercially available digital multimeters,

but for that i need the most important resource(as i think),,the DMM users!!!!
so i appriciate any comments on digital multimeters that you came across (if possible mention the product also)

and also you can give me your valuable ideas on this topic,,,,,(any thing related is welcome)............
thank you....

 

[MrAl]

Hi,

There are a number of ways to start something like this. You can start from regular logic and analog parts, from a dedicated meter chip, or from a microcontroller chip. You'd have to decide which way you want to start first or at least investigate those three approaches. You also might want to think about how complex you want to get with it, and most important what ranges you want to be able to measure and what you want to measure like voltage, current, resistance, capacitance, etc., and what resolution you want for each type of measurement. How many digits in the readout.

 

[RCinFLA]

Most DVM's use an integration based counter. They reset the integrator at beginning of cycle then the counter counts the number of clock cycles until a comparitor stops the count based on the input voltage to the comparitor.

Look up the Intersil chips, ICL7xxx series. There are a number of application notes on their web site.

 

[mafrasm]

thanks for both replies (Mrai & rcinfla),i thought to go with pic microcontroller 16f877a for this project,and also i think a four digit display(in a 16*2 lcd display) of all the measurements is enough for my DMM (but the range should be changed manually or automatically and should be displayed as adequate),further as a first step i try to include simple voltmeter,ammeter,and resistance measuring meter in my DMM.since i dont have much experience with practical electronics(though im somewhat ok with paper works,) i require all of your supports if possible please help me out with some circuit diagrams (and ideas behind that).and i will update my project status often and you guys plzzzzzzzz help me .....

 

[mafrasm]

plzzzzzz any body help me...........

 

[MrAl]

Hi,

Well, you are talking about a fairly big project here if you dont have any idea how to start.

You should actually start with a voltmeter and get that up and running, then go from there. Reading voltage with a PIC chip isnt that hard but you're going to have to study the chip you want to use in detail, as to the register and what they do.
The main parts would be:
1. Read the AD port, and maybe some user switches later
2. Convert the binary AD reading into BCD and display
3. Repeat above indefinitely.

Your display could be made from display chips and seven segment LEDs so you dont have to get into multiplexiing with the PIC itself.
Even simpler would be a serial type LCD display, that you would write to with the PIC.

Once your circuit can measure and display voltages from 0 to 5v you can go from there.

If you want really good resolution however you'll have to add another chip that can read AD inputs to 12 bits or better. It's a lot nicer doing it that way than using oversampling. I still suggest that you start with a simpler circuit with the PIC where you read the input AD port and display the result. That will get you going and it will be easier to proceed from there.

What made you choose the 16F877 chip over the others?
Also, is this your first PIC project or have you built other things before?
Also, what level of mathematics did you complete in high school or college?

 

[mafrasm]

thanks for the reply MrAi,
for the moment i have a circuit diagram of voltmeter and an ammeter,which is working fairly well with the proteous simulation(although still i couldn't extend it to several ranges).i have some experience with pic 16f84A chip but since it has only one ADC pin i thought to go for 16f877A chip.and also regarding mathematics,i think it wont be a problem ,coz im in 3rd year of my engineering course and i almost completed all the mathematics courses prefered.
i dont know how to attach my circuit diagrams here,if there is that facility im willing to attache my circuit diagrams here.........
and also if anybody has any circuit diagrams related to this plz post it.......

 

[KeepItSimpleStupid]

Look up "dual slope integration". That's the basics. You might want to leave AC out of it.
Basic functions are then DC volts, DC Amps, Resistance and for extra credit the diode test.
Auto-Ranging gets messy.
The DC Ammeter can be a Shunt Ammeter rather than a feedback Ammeter.
The next question is number of digits or Counts. You have meters that will goto say 1999 and ones that go to 40000, for instance.

 

[MrAl]

thanks for the reply MrAi,
for the moment i have a circuit diagram of voltmeter and an ammeter,which is working fairly well with the proteous simulation(although still i couldn't extend it to several ranges).i have some experience with pic 16f84A chip but since it has only one ADC pin i thought to go for 16f877A chip.and also regarding mathematics,i think it wont be a problem ,coz im in 3rd year of my engineering course and i almost completed all the mathematics courses prefered.
i dont know how to attach my circuit diagrams here,if there is that facility im willing to attache my circuit diagrams here.........
and also if anybody has any circuit diagrams related to this plz post it.......

Hi,

To attach a file, click on the button "Go Advanced", then scroll to the bottom and click on the button "Manage Attachments". That will allow you to upload a file.

 

[mafrasm]

can any body suggest me a chip to do auto ranging?or any techniques related to auto ranging?

 

[MrAl]

Hello again,


I dont know of any offhand but maybe someone else here does, but one thought that comes up is to use a digital pot to adjust the gain of the front end and that would allow various ranges. I dont know how good this would work as i've never actually tried this. You also have to incorporate some sort of voltage limiting so a higher voltage doesnt blow out the front end circuit. An alternate idea is to not use autoranging but instead use a multi position selector switch so you can switch ranges that way.

As you might be finding out by now, it is not that simple to build a full DMM using small parts, and most people would probably opt to use a dedicated chip that does everything in one chip just about. You can still add stuff to boost the functionality but it will be much less work.

 

[mafrasm]

i have being trying to implement a resistance and a capacitance measuring meter through monostable 555 timer circuit(**broken link removed** a pulse is produced with a width of 1.1RC seconds,so i measure the pulse time/width and i calculate the resistance value for a known capacitor and vice versa.

but i get problem when i go for small value resistance,coz the time or the pulse width is not enough to measure it accurately,what can i do for this?
(if the question is not clear plz ask)

 

[MrAl]

Hello again,

Have you read the Microchip app notes on measuring these kind of things? There are some notes on that site that should help you with these circuits.
Measuring small value resistors is a problem in itself. Sometimes you need a decent gain amplifier to amplify the voltage across the resistor so you can get a voltage that is high enough to measure with 'normal' equipment or the usual AD input. The gain amplifier amplifies the small voltage up by at least 10 times which then makes it much easier to measure, then you divide by 10 in the algorithm.
What size caps do you intend to measure?
What range of resistances do you intend to measure?

 

[KeepItSimpleStupid]

I designed an I-V converter and the LabVIEW application implemented auto-range. Nothing really magic, but I had a -10 to 10 V linear output V. The actual range was probably higher and relays to select the range. So, you just keep bumping the range until you get something between 1 and 10 volts or -1 and -10 volts or your at the lowest range/highest gain possible.

Protection mechanism were built into the circuit design which is the same you would do for a voltmeter.

 

[mafrasm]

I designed an I-V converter and the LabVIEW application implemented auto-range. Nothing really magic, but I had a -10 to 10 V linear output V. The actual range was probably higher and relays to select the range. So, you just keep bumping the range until you get something between 1 and 10 volts or -1 and -10 volts or your at the lowest range/highest gain possible.

Protection mechanism were built into the circuit design which is the same you would do for a voltmeter.

im bit confused....
did u build something which converts i(current) into v(voltage)?if so can you plz give me the circuit diagram and the codes or else can you plz guid me on how to achieve this.....anyway thankyou for the reply......

 

[KeepItSimpleStupid]

Total cost was about $5000. It used 2 system DMM's, one 4 channel system DMM (3 used) and had suppression and biasing capability. It was also 2 or 4 terminal. 4 ranges: 100 mA, 10 mA, 1mA and 0.1 mA FS. 3 dual tracking power supples and one logic supply.

Designed primarily as an analog Front End. It was stable into capacitive loads.

It wasn't easy to do and it could have been better. AC measurements (front end) was excellent. DC was good. The 25 uV of offset at the input wasn't compensated for, so I has an error on the order of 25 pA.

1% resistors were used in the FB loop, and calibration constants could be adjusted in software. The AC calibration could too after it was processed by a $5000 lock-in amp.

The inputs were gaurded too.

It also incorporated an AC overload indicator.

 

[mafrasm]

Total cost was about $5000. It used 2 system DMM's, one 4 channel system DMM (3 used) and had suppression and biasing capability. It was also 2 or 4 terminal. 4 ranges: 100 mA, 10 mA, 1mA and 0.1 mA FS. 3 dual tracking power supples and one logic supply.

Designed primarily as an analog Front End. It was stable into capacitive loads.

It wasn't easy to do and it could have been better. AC measurements (front end) was excellent. DC was good. The 25 uV of offset at the input wasn't compensated for, so I has an error on the order of 25 pA.

1% resistors were used in the FB loop, and calibration constants could be adjusted in software. The AC calibration could too after it was processed by a $5000 lock-in amp.

The inputs were gaurded too.

It also incorporated an AC overload indicator.

thank you for the reply............but seems things are tooo technical ..... idont understand alot...........

 

[mafrasm]

Hello again,

Have you read the Microchip app notes on measuring these kind of things? There are some notes on that site that should help you with these circuits.

i went through there site but icouldn find it can u plz help me to find it???

 

[KeepItSimpleStupid]

Most multimeters measure the voltage across a fixed resistor keeping the voltage below a diode drop at full scale. A feedback ammeter generally is used for currents below 10 mA. It gets messy when you want currents of 10 mA and voltages below 1 V.

 

[MrAl]

Hello again,


You only quoted one part of my last post and didnt answer any of the questions. We have to know what you want to do to help and that means knowing exactly what you want to measure...

Hello again,

Have you read the Microchip app notes on measuring these kind of things? There are some notes on that site that should help you with these circuits.
Measuring small value resistors is a problem in itself. Sometimes you need a decent gain amplifier to amplify the voltage across the resistor so you can get a voltage that is high enough to measure with 'normal' equipment or the usual AD input. The gain amplifier amplifies the small voltage up by at least 10 times which then makes it much easier to measure, then you divide by 10 in the algorithm.
What size caps do you intend to measure?
What range of resistances do you intend to measure?

You'll have to answer those questions at least.