After 5 years of researching, developing and testing, I'm ready with the first PCB design:
The main PCB fits in 10x10cm layout and contains 4 modules : Analog front-end with OVP and two voltage ranges, ADC based on LTC2440, Voltage Reference based on LTC6655 and power supply using classic LM317/LM337 LDO.
The PCB does not includes the controller for reading out and displaying the measurement results from the ADC.
I have to double check availability of the components, schematics and layout for possible errors before ordering all PCBs next week.
There are 2 more PCB :
- one for AC input circuit with over-voltage protection using GDT and varistors (based on IEC 60950-1 schematics) ̶a̶n̶d̶ ̶p̶r̶o̶t̶e̶c̶t̶i̶o̶n̶ ̶a̶g̶a̶i̶n̶s̶t̶ ̶i̶n̶t̶e̶r̶f̶e̶r̶e̶n̶c̶e̶ ̶v̶o̶l̶t̶a̶g̶e̶ ̶f̶r̶o̶m̶ ̶t̶h̶e̶ ̶m̶a̶i̶n̶s̶. A friend of mine gave me advice, that main filter is not needed when transformer is used, because it acts like low pass filter for external interference. Only if switching power supplies are used, it is worth to place such of filter. A reader of the blog notice that the transformer is not ideal and always a capacitance exists between the primary and secondary side. Which means that any AC signal (noise) can pass through the transformer. The frequency band is limited to the capacity value and the transformer's frequency loss. So far I was not able to find a tests results about this impact. I found an application note from Schaffner where noise filter is placed between a main line and linear power supply.
- one for AC-DC conversion which includes again over-voltage protection using TVS diodes and CLC ripple filter.
I paid special attention for input over-voltage protection after 2 times literally burning the op-amp in my LTZ1000 based 10V voltage reference due mains AC over-voltages.
Here is the AC input PCB:
The AC-DC PCB supports 2 sizes of transformers.
Based on LTSpice simulation and LM317/337 ripple rejection vs. output load current chart, I expect to have around 15uV noise in the 0.1-0.3A current load range after LDO outputs.
And one more news : finally I bought an Agilent 3458A DMM :-)
The DMM is 10 years old and I will make later a post about the modifications and testing some DMM parameters, I was able to made. This will include also a INL against Programmable Josephson voltage standard.
I think to make a voltage calibrator based on AD5791 DAC and with the help of the 3458a linearity I can measure the real INL of the AD5791. Once this is done, I will be able to measure the INL of the open DCM voltmeter using the AD5791 based calibrator.
