Milli-Volt Meter with 24bit ADC (LTC2400)

[Dbldutch]

Hi,

I would like to make you aware of a project that has been going on in other parts of the Web for a little while that details the building of a very nice, stable and accurate (milli) voltmeter. The project is based around the rather popular LTC2400 ADC that will get you a nice addition to your tool box, without breaking the bank. I am very impressed and happy about my two meters, and if anything, I want you to be aware of this development as well.

It is not my project, although I wrote some more code for it, and designed a power section for a re-chargeable 9V NiMH/NiCAD cell for it.
The original idea came from Louis Scully and he produced a set of video's on Youtube. One of his followers, Greg (Barbouri), produced a very nice PCB that is available through OSH Park. His own website details his built, and I based mine on his groundwork. Since then, also Louis made updates and enhancements.

Because I cannot post pictures here, I have to ask you to follow this link to my own blog, which gives you access to the above mentioned websites, other available information, in addition to my own version of the (Arduino) source code.

http://www.paulvdiyblogs.net/2016/09/building-6-digit-digital-milli-voltmeter.html

The two voltmeters I built have specifications that I find impressive for the amount of money I spent. (about 60-70 Euro's each)
With the current hardware design, the voltmeter has a maximum input of 50V, is DC only, but goes down to the micro-Volt level with 6 (usable) decimal digits.

If you have a need to go to this sub-volt level or with this accuracy, I highly recommend you have a look at this tool.

The software is truly open, so if you have suggestions, improvements or extensions that can be made, feel free to chime in. This tool has a lot of potential and can benefit from the collective wisdom, wizardry and experience that is available on this Forum.

Enjoy!

[jeremy]

Very interesting, thanks for sharing. Do you have problems with the seebeck effect at the low end? Are you using low tempco leads? I only ask because I have seen some commercial devices which work in the microvolt range but the bits in those things do add up. Would be neat to see how far you can get accuracy wise with a low cost device such as this.

[Dbldutch]

Hi jeremy,

I do not have the knowledge or the equipment to really investigate the accuracy at the micro-volt level.
As I just finished the built, I'm still "playing" & tweaking to get myself familiar with the meters.
My hope is that others can chime in so we can learn more, or see what the limits really are.

Cheers!

[Dwaine]

Nice project.  Maybe when I get some time, I'll build one.   With a nice OLED display....

[femtohertz15]

It might be simple enough hack to graft the 6 digit readout onto a manual ranging DMM to get the full complement of ranges, current, and ohms, and just soft-cal out the divider errors and tempco's.   Well the latter might be quite an adventure.

[JohanH]

Is there anyone else that built this meter?

I've built it recently based on Barbouri's 2.11 hardware design https://www.barbouri.com/2019/08/07/millivolt-meter-version-2/ I used another voltage regulator, though (imported it to KiCAD).

Barbouri's latest code (v3.34) didn't really work for me, so I changed the adjustment/calibration to include a 0V calibration point and changed it to use three different curves/ranges to calculate voltage. I've uploaded the modified code here: https://github.com/hevanaa/6-digit-millivoltmeter

With the modifications, it's now possible to adjust it quite accurately at the calibration points that Barbouri used (0.4096V, 3.6864V, 4.096V, 36.684V) and with the additional 0V point.

[BD1QMP]

cannot see the sCH on Barbouri's website。

[Kleinstein]

A problem with the circuits shown here seems to be that the amplifier can not cope with negative voltages, even just a few µV from thermal EMF or noise.  A mV / µV meter inherently needs to be able to measure on both sides of zero, at least a little on the negative side. One kind of need at least a little (e.g. 100 mV) negative supply for the amplifier. There is a good reason the LTC2400 has a little negative range.

[newbrain]

The schematics, and especially the calibration procedure, as presented in the original video are total .

I don't remember if he improved the design of either one later, but here's some discussion about it.
https://www.eevblog.com/forum/projects/which-ad-converter-for-measuring-10v-with-five-significant-digits/msg4854611/#msg4854611