Measuring very small voltages and currents

[K Murty]

Hello.
How would this be done? Using a step up transformer?

 

[scottdave]

Can you be a little more specific? How small? A transformer will only work with AC, not DC. What is it that you are measuring? If you have access to an oscilloscope, that might be a way.

 

[K Murty]

Can you be a little more specific? How small? A transformer will only work with AC, not DC. What is it that you are measuring? If you have access to an oscilloscope, that might be a way.

I am not measuring anything, but I am wondering how one would measure AC waveforms that are very small in magnitude, even smaller than the micro prefix. I know transformers are used to step down high voltages so they can be measured by voltmeters and for safety. So I was wondering if the opposite use, to step up small voltages and currents so they can be measured by instruments, is a use for transformers?

 

[berkeman]

https://en.wikipedia.org/wiki/Lock-in_amplifier

picoammeter at https://en.wikipedia.org/wiki/Ammeter

https://en.wikipedia.org/wiki/Radio_receiver

We use a sigma-delta receiver with decimation to achieve about 80dB quiet line receive performance...
https://en.wikipedia.org/wiki/Delta-sigma_modulation

 

[dlgoff]

So I was wondering if the opposite use, ...

For audio signals for example;
from http://www.vias.org/crowhurstba/crowhurst_basic_audio_vol2_033.html

When a transformer is used to step up the audio voltages from a microphone or pickup, it steps up the noise voltage that comes with it.

 

[berkeman]

For audio signals for example;
from http://www.vias.org/crowhurstba/crowhurst_basic_audio_vol2_033.html

When a transformer is used to step up the audio voltages from a microphone or pickup, it steps up the noise voltage that comes with it.

Excellent point.
@K Murty -- can you say a bit about the small signal noise sources that you've been reading about, and how to maximize the signal-to-noise ratio (SNR) when dealing with small signals in the presence of noise sources (natural and human-generated)?

 

[K Murty]

Excellent point.
@K Murty -- can you say a bit about the small signal noise sources that you've been reading about, and how to maximize the signal-to-noise ratio (SNR) when dealing with small signals in the presence of noise sources (natural and human-generated)?

I will reply in some time.

 

[berkeman]

I will reply in some time.

Thanks.

Hint -- the main natural noise source I've had to deal with is thermal electrical noise. That should help your search terms...

 

[K Murty]

Excellent point.
@K Murty -- can you say a bit about the small signal noise sources that you've been reading about, and how to maximize the signal-to-noise ratio (SNR) when dealing with small signals in the presence of noise sources (natural and human-generated)?

I believe you would use some of the devices and methods mentioned in your post. I am having a hard time finding the correct material on how to minimise noise from small signal sources.

I think the integrating A D converted is key here...

Thanks.

Hint -- the main natural noise source I've had to deal with is thermal electrical noise. That should help your search terms...

I would think cooling!

 

[dlgoff]

We use a sigma-delta receiver with decimation to achieve about 80dB quiet line receive performance...
https://en.wikipedia.org/wiki/Delta-sigma_modulation

From the wikipedia link,

Primarily because of its cost efficiency and reduced circuit complexity, this technique has found increasing use in modern electronic components such as DACs, ADCs, frequency synthesizers, switched-mode power supplies and motor controllers.

I'm trying to visualize, circuit wise, how this would be applied to motor controllers. I'll be doing some searching but if you know of any good sources, could you provide them here? However this may be a little off-topic.

edit: Never mind berkeman. I just found this that looks good. http://www.analog.com/en/technical-articles/sigma-delta-conversion-used-for-motor-control.html

 

[jim hardy]

Think in terms of energy present in your signal.
A very feeble AC will get more feeble when it has to energize the core of a transformer.
That's why we usually use an interface circuit to make it energetic enough to drive a conventional instrument.

Tutorial here, don't be scared off by the math on first few pages...
http://www.ti.com/lit/ml/slap103/slap103.pdf

 

[jim hardy]

I think the integrating A D converted is key here...

Integrating ADC does a wonderful job. If you integrate for one line cycle you have really suppressed line frequency interference because integral of a sinewave is zero... but it's slow. old jim

 

[dlgoff]

Tutorial here, don't be scared off by the math on first few pages...
http://www.ti.com/lit/ml/slap103/slap103.pdf

Dang Jim. Where do you find this good stuff? This quote from the tutorial is priceless IMO.

Many engineers, used to getting steady values from their low-resolution converters, are dismayed when they
find the output value of a 24-bit device fluctuating madly in its least significant bits. It’s not hard for an engineer,
long used to working with low-resolution devices, to feel a bit cheated when he finds out that a 24-bit device isn’t
really 24 bits when noise is taken into account!

 

[jim hardy]

Most everything in nature is moving if you look at it close enough. In the plant it was not practical to measure closer than 0.1%, a part per thousand, because every signal has "noise" imposed on it. 0.1% takes only a 10 or 11 bit ADC. We installed a 14 bit (15 including sign) ADC for our computer because its drift and nonlinearity were so small relative to our 10 bit need that it could go for years between calibrations. That saved a LOT of man-hours.

There's a whole niche industry focused on analyzing that 'noise' riding atop process signals.
https://www.kth.se/polopoly_fs/1.469664!/In-Core%20Neutron%20Noise%20Analysis%20for%20Diagnosis%20of%20Fuel%20Assembly%20Vibrations.pdf

typically one uses an analog high pass filter to block the steady state signal and amplify just the noise, then a fast ADC connected to a FFT analyzer.
In my day that took 200 pounds of computer equipment, nowadays they do it with a DSP processor that fits in your shirt pocket.


old jim

 

[dlgoff]

We installed a 14 bit (15 including sign) ADC for our computer because its drift and nonlinearity were so small relative to our 10 bit need that it could go for years between calibrations.

You da man.

 

[Baluncore]

The first thing I would do is amplify the signal using an instrumentation amplifier.
For low voltage precision DC signals, a zero-drift chopper amplifier might be an advantage.

 

[K Murty]

I think I would have to read up significantly more on the posts made here, thank you.