Microcontrollers based multimeter

[elecLear78]

I have got an idea of designing a micro controller based multimeter with all the features like continuity and some additional features etc but at some reasonable price. Is it practical?

 

[Nigel Goodwin]

I have got an idea of designing a micro controller based multimeter with all the features like continuity and some additional features etc but at some reasonable price. Is it practical?

Of course, it's probably how many modern instruments actually work.

 

[elecLear78]

I am sorry I am mistaken all multi meters have controllers in them I think.

 

[Nigel Goodwin]

I am sorry I am mistaken all multi meters have controllers in them I think.

Not all, cheap ones usually have a specific analogue to digital converter chip in them, which directly drives the display.

 

[DirtyLude]

Not an EE or very good at this stuff, but...

My understanding from the teardowns on the eevblog channel that most modern multimeters have a either a one chip does all design with a chip specifically made for multimeter use, or has a two chip design with a purpose made multimeter measurement chip plus a microcontroller that handles the display and higher functions.

I'm not certain what you mean by 'practical'. You wouldn't be able to make them as cheap as they are currently made. The Chinese ones are built down to a price. I get a bunch of $10 ones with continuity that I use for non-critical low voltage measurements and testing.

 

[ronsimpson]

I have got an idea of designing a micro controller based multimeter

Great hobby.
You can not compete against the BIG companies.
**broken link removed**

 

[MikeMl]

The AD converter in most micro-controllers is 10bits or less. That is a resolution of 1/1024, or only 20mV on a +-10V scale. Having owned various Fluke digital meters, I wouldn't be happy with such a crude measurement.

You would need a 14 to 16 bit A/D to compete...

 

[Nigel Goodwin]

The AD converter in most micro-controllers is 10bits or less. That is a resolution of 1/1024, or only 20mV on a +-10V scale. Having owned various Fluke digital meters, I wouldn't be happy with such a crude measurement.

10 bit is 0.1% resolution, which considerably exceeds the accuracy of normal multimeters, including most Fluke ones (and pretty well equals the rest)

So I don't see as 10bit is any particular limitation?.

 

[MikeMl]

My go-to meter...

 

[Mikebits]

My understanding from the teardowns on the eevblog channel

I often enjoy Dave and his video blogs, very informative but he tends to take a bit long to get his point across, and my ADHD usually kicks in before the video is done. A DVM sounds like a good project, just for learning sake.

 

[BobW]

I often enjoy Dave and his video blogs, very informative but he tends to take a bit long to get his point across, and my ADHD usually kicks in before the video is done. A DVM sounds like a good project, just for learning sake.

I think that comment applies to 99% of all techie type videos. Most of them could be edited down severely to less than a quarter of their original length. The useful info in most of them could be summed up in about two sentences, and I could read that in less time than it takes the video to start loading.

 

[NorthGuy]

I think that comment applies to 99% of all techie type videos. Most of them could be edited down severely to less than a quarter of their original length. The useful info in most of them could be summed up in about two sentences, and I could read that in less time than it takes the video to start loading.

Remember how is it at school - you write a composition, and they say you need to write at least this number of pages. This teaches you to put out lots of words which are necessary to fill the required pages - this translates into repetitiveness, using lots of extra words. Then, you grow up and you make blog videos ...

It is opposite in the professional word - you have 30 seconds and you must fit the whole story in.

 

[radioratiometrics]

Hi eleclear I landed on your post by coincidence I am currently thinking of building the same thing. Please see the attached file. What you see in there is a prototype but in the end the oversampling microcontroller will get replaced with a 16bit SPI ADC. I have started testing my voltage attenuator, and ohmmeter part. I need to carry on with the other blocks. The final microcontroller will change as well and the PWM will be replaced with MCPXXX dacs. I am not trying to build an industrial multimeter but rather have one I can use and fix myself if something goes wrong. Its also more a learning project and I try to keep it fun.

 

[radioratiometrics]

A more simplified block diagram of what I am currently working with

 

[elecLear78]

Thank you very much for sharing the circuit. Really I am not sure if I will start this project.

 

[DirtyLude]

A more simplified block diagram of what I am currently working with

Are you just using the first PIC as an ADC? Is there an advantage to that rather than just using a dedicated ADC?

 

[radioratiometrics]

Hi DirtyLude I wanted to find a way to build an oversampling ADC using a pic. The reason I did that in first place was because I had defined an array filled it with all samples and them average all the values. Since the 16F877A has only 368 bytes of RAM and getting 255 samples in was challenging so I initially went for 2pics. I then found a work around where I used a different cumulative technique instead of an array and I was able to get my 255 samples in. The second reason why I used 2 pics was for response time. Once microcontroller would be busy sampling while the second one would take care of handling LCD display and control. If my second scheme ends up being slow I will have to go back to 2 pics