I am designing Open source DC load with multi channel /Modular design, It would have a touch screen interface along with encoders knobs for fast selection and toggle for menu.

Suggestions are welcome
HMI interface - STM32H750 With 5" or 7" Touch Panel
DAC - 16 bit DAC
Resolution- 0.1mA
Working Voltage 150v
Working Current 10 AMP

 

Welcome to AAC!
It's asking a lot to get 0.1mA resolution at 10A. Do you really need that?
Will all channels have a 10A-capable output?
How many channels?
Is 150V a fixed DC output voltage?

 

It's asking a lot to get 0.1mA resolution at 10A

That resolution is easy, that accuracy and precision would be more difficult.

 

Welcome to AAC!
It's asking a lot to get 0.1mA resolution at 10A. Do you really need that?
Will all channels have a 10A-capable output?
How many channels?
Is 150V a fixed DC output voltage?

Looking at the complexity I believe 0-3 Amp at 0.1 mA would be better and post 3-10 amp 1 Ma will be sufficient!

 

That resolution is easy, that accuracy and precision would be more difficult.

Share some pointers on achieving such a thing.

 

Share some pointers on achieving such a thing.

Actually, I misread the resolution as .1A. At .1mA that is 1 part in 100,000. That is not doable with a 16bit DAC which is, at best, 1 part in 65536.

As for that accuracy or precision? Way out of my league. Someone who could do that would could name his price at any precession lab equipment manufacturer.

Here is a high end instrument that costs over $4000. Its accuracy for 30A is 13mA.

Keysight Electronic Load

And you think you need accuracy if 1/130 of that? Good luck.

 

Actually, I misread the resolution as .1A. At .1mA that is 1 part in 100,000. That is not doable with a 16bit DAC which is, at best, 1 part in 65536.

As for that accuracy or precision? Way out of my league. Someone who could do that would could name his price at any precession lab equipment manufacturer.

Here is a high end instrument that costs over $4000. Its accuracy for 30A is 13mA.

Keysight Electronic Load

And you think you need accuracy if 1/130 of that? Good luck.

I Guess you got the wrong Idea on the Spec sheet, I am trying to achieve 0.1mA at 3 Amps and 1mA at 0-10 Amp.

Resolution=Number of Levels
___________________
Full Scale Range

3A
______= 0.0001A Implies to 30000
2n

Calculate for n (DAC Resolution) n=log2(30000) =14.9 or 16 bits

 

I Guess you got the wrong Idea on the Spec sheet, I am trying to achieve 0.1mA at 3 Amps and 1mA at 0-10 Amp.

Resolution=Number of Levels
___________________
Full Scale Range

3A
______= 0.0001A Implies to 30000
2n

Calculate for n (DAC Resolution) n=log2(30000) =14.9 or 16 bits

Okay, so after altering the specification, it is within the capability of a 16-bit DAC.

But still, if you do not achieve that accuracy, you will give the user a false idea of how accurate the setting is. So I reinstate my original answer, that resolution is easy, but that accuracy and precision are not.

 

Okay, so after altering the specification, it is within the capability of a 16-bit DAC.

But still, if you do not achieve that accuracy, you will give the user a false idea of how accurate the setting is. So I reinstate my original answer, that resolution is easy, but that accuracy and precision are not.

I agree with you, The Community needs a professional like opensource Dc Load with better accuracy and precision. That is why I need inputs from gurus like you to build it.

 

On the EEVBLOG forum, there is a section of Metrology dedicated to volt-time nuts, as they call themselves. Many of the posters there are specialized in high performance voltage and frequency references and measurements.
Suggest that you also have a read over there.

 

Really there are three parameters and "precision" is just a nice word that gets thrown around. Resolution is how many digits your reading has. Accuracy is how close to actual your reading is, and repeatability is the range of readings that exactly the same value will show in the measurement.
Just trying to help make sense of measurements nomenclature. "Precision" will cost you points in some classrooms.
Precision equipment, though, will usually provide better accuracy at greater resolution.

And such a sophisticated load bank makes me wonder what sort of testing will be going on. That may be interesting to know about.

 

I agree with you, The Community needs a professional like opensource Dc Load with better accuracy and precision. That is why I need inputs from gurus like you to build it.

WHY does it need some opensource DC load with better accuracy and precision?

And, if it does, what constitutes better accuracy and precision? How good is good enough? How did you determine what is good enough?

What specific application needs to be able to control the current so well that 2.9999 A versus 3.0000 A makes a critical difference?

When I was working for NIST, we cobbled (well, a lot more went into it than that) a constant current supply that could deliver 4000 A with mA-scale noise. But while the noise was in the mA range, our knowledge of what the DC current actually was wasn't even to within 1%.

 

When we tested a power supply box that was intended to power film cameras and lighting in a crash car, the load we created used a portion of a bank of 120 volt 1000 watt floodlights. Only our measurements were quite accurate, the current was within 2% of the target value. Not so very great resolution and control.But we were quite able to evaluate the performance very accurately.No need to set the load to draw 100.000 amps exactly. In the real world of industry, close enough is adequate.

 

WHY does it need some opensource DC load with better accuracy and precision?

And, if it does, what constitutes better accuracy and precision? How good is good enough? How did you determine what is good enough?

What specific application needs to be able to control the current so well that 2.9999 A versus 3.0000 A makes a critical difference?

When I was working for NIST, we cobbled (well, a lot more went into it than that) a constant current supply that could deliver 4000 A with mA-scale noise. But while the noise was in the mA range, our knowledge of what the DC current actually was wasn't even to within 1%.

Once things went digital, everyone's perception of accuracy became perverted.
All those perfect digits on displays conceal the murky reality, it's all analog in the real world, and accuracy is not for free.

A wise man knows what he needs and spends only for what he needs.

 

It is certainly easy to confuse accuracy and resolution, not only on instrumentation but even in plain math, where somebody starts out with two-figure data and arrives at a six digits result. Resolution, accuracy, and repeatability can combine to deliver precision measurements only if all three are adequate for the task.