Watch out for those two clamp diodes connected to the (+) input of U1 and your range switch; unless they have EXTREMELY low reverse leakage current, they will cause a substantial offset error due to the resistance of the voltage divider. Pick them carefully.

 

Watch out for those two clamp diodes connected to the (+) input of U1 and your range switch; unless they have EXTREMELY low reverse leakage current, they will cause a substantial offset error due to the resistance of the voltage divider. Pick them carefully.

Yes! I looked into these barrier diodes and from what I see, you have to choose between 3 basic specs: Forward Voltage, Forward Current, and Reverse leakage. You have to give up some to get some of another it seems.

For what I’m doing here I am thinking the forward current would be my least important of the three. A low Vf is good so as the input rises to the OV state the current is directed away from the opamp sooner. Also Reverse leakage should be as minimal as possible as to have as little impact on measurements as possible.

I’ll find those old Part Numbers I had but I also welcome suggestions in terms what I should use if anyone knows off the top of their head. I know the one I was going to use was 250mV Vf and the leakage was low but i don’t remember how low it was.

 

I believe this was one of the ones i found. I have to do more math with it when i get home.

BAT20J
http://www.st.com/web/en/resource/technical/document/datasheet/CD00012201.pdf

 

...................
The only calibration point on the diagram I drew here is the trimmer pot for the Reference voltage. We discussed adding possibly trim resistors in the divider as well or compensating for error in code. (I would think this could be done by measuring the ACTUAL resistor values with considerable accuracy and entering them into the code, have the code perform the calculations and then compare it to the voltage measured). Which method do you think would be more stable?

Measuring the resistor values will only correct the error from that source.
It won't include any op amp or A/D gain/offset errors, or changes in resistor values over time.

 

LTC2440 24 bit ADC

If measuring low voltage levels is your thing, you may want to consider the differential input version of the 2440 (2442 I think).

Also, 24-bit adcs are not that easy to handle from analog perspective. Starting with a 16-bit adc would be a good idea.

 

If measuring low voltage levels is your thing, you may want to consider the differential input version of the 2440 (2442 I think).

The LTC2440 has differential inputs, both for signal inputs and for the reference. The LTC2442 has an input multiplexer that allows it to function with 4 single-ended inputs or two sets of differential inputs.

 

I believe this was one of the ones i found. I have to do more math with it when i get home.

BAT20J
http://www.st.com/web/en/resource/technical/document/datasheet/CD00012201.pdf

That would not be a good choice for your protection diodes, as it's got EXTREMELY high reverse leakage current (as do most Schottky rectifiers). Look at the data sheet. For this kind of application, you need leakage currents on the order of picoamps, not microamps. I'd recommend using diode-connected JFETs (i.e., use the gate as the anode, and short the source and drain together to serve as the cathode) rather than regular silicon diodes. An MMBF4117, for instance, has a maximum gate leakage of 10 pA at 25°C, which will probably be OK.

Don't worry about forward voltage; just insert a small resistor (a few hundred ohms or so) between the junction of the protection diodes and the opamp (+) input to guard against the possibility that the opamp's internal protection diodes might conduct before your external diodes.

 

You will need a low pass filter on the input, with cutoff below the half of sampling frequency.

 

Since you are using a mcu, you may consider C8051 parts that have a built-in 24-bit adc. I have used them in the past and they are really convenient. Determining the sampling frequency + oscillator frequency requires some work.

Another not-so-good alternative is HX711 - I have a few examples of using it on my blog. Sub-par performance but not bad for a <$5 part.

The key to the success of your project will be AFE.

 

That would not be a good choice for your protection diodes, as it's got EXTREMELY high reverse leakage current (as do most Schottky rectifiers). Look at the data sheet. For this kind of application, you need leakage currents on the order of picoamps, not microamps. I'd recommend using diode-connected JFETs (i.e., use the gate as the anode, and short the source and drain together to serve as the cathode) rather than regular silicon diodes. An MMBF4117, for instance, has a maximum gate leakage of 10 pA at 25°C, which will probably be OK.

Don't worry about forward voltage; just insert a small resistor (a few hundred ohms or so) between the junction of the protection diodes and the opamp (+) input to guard against the possibility that the opamp's internal protection diodes might conduct before your external diodes.

I...didnt even realize that you could use those like that. That's a great idea! With most of the shottkeys i've looked through that was one of the best reverse leakage ones I have found. I will add some of those JFETs to the list and try those instead.

You will need a low pass filter on the input, with cutoff below the half of sampling frequency.

Good call on that I didnt think of that either. Its been a while since i looked at filters ill come up with one and draw it into the next drawing. thanks!

Since you are using a mcu, you may consider C8051 parts that have a built-in 24-bit adc. I have used them in the past and they are really convenient. Determining the sampling frequency + oscillator frequency requires some work.

Another not-so-good alternative is HX711 - I have a few examples of using it on my blog. Sub-par performance but not bad for a <$5 part.

The key to the success of your project will be AFE.

What do you mean by "AFE"?

 

If you looking for an analog part of the digital voltmeter, you can check this blog : opendcm.blogspot.com.
The blog contains useful information for voltmeter accuracy and noise. You can find also schematics.

 

I...didnt even realize that you could use those like that. That's a great idea! With most of the shottkeys i've looked through that was one of the best reverse leakage ones I have found. I will add some of those JFETs to the list and try those instead.



Good call on that I didnt think of that either. Its been a while since i looked at filters ill come up with one and draw it into the next drawing. thanks!




What do you mean by "AFE"?

Analog front end.

 

If you're going the autoranging route you could use reed relays to do the switching then you don't need to worry about '4066 resistance and leakage.

 

If you're going the autoranging route you could use reed relays to do the switching then you don't need to worry about '4066 resistance and leakage.

That actually is a far more common technique than most people realize. Instrumentation has been pretty much "all solid state" for 40 years, but if you tear apart today's high-end stuff, there often are small relays doing internal things.

ak