Uploaded December 2025 | Updated November 2026, 3 hours ago
Day 6: The main benefit of feedback ammeters is elimination of the "burden voltage" seen with shunt ammeters. Magnetic approaches do not have this problem either, but have limited sensitivity unless they also introduce significant series impedances reflected through the current transformer.
The disadvantages are that the instrument has to source or sink the full measured current. For lower currents this is not challenging and the main non-idealities remaining are closed loop bandwidth related.
Presented here is a straight forward JFET op-amp TIA-based feedback microammeter with 5 transimpedances ranging from 1 mA/V to 100 nA/V. On the most sensitive range it may have a slight zero-offset, this can be corrected with an op-amp that supports an offset null, like the TL071 instead of the TL072 dual amp I used to simplify implementing the rail splitter. A much nicer modern op-amp would enable extending the range to picoamperes, but care would be required in the construction and shielding.
In the practical circuit I flipped the input terminals around as the TIA is fundamentally inverting, as discussed in the video this means the +ve input is tied directly to the -ve output which may present a problem for common grounded instrumentation, but is of little consequence as long as you know about it. The bandwidth of the instrument well exceeds 1 kHz on the most sensitive range, but has the option to low pass the measurement strongly at the control loop which can be useful for average DC current measurements. It can source and sink a bit more than 2.5 mA on the highest range, which is fine for its intended purpose.
The burden voltage of my multimeter on its low current range did not allow it to measure the calculator supply current, especially at slightly lower supply voltages, at around 1.2 volts it would not turn on at all with the meter in series! My old AVO MkIII did better on its 50 uA range, as did my MF47B, but both made the display dim a lot with their burden voltages. This circuit has about 3 mV burden voltage which is mostly offset and could be improved.
I should probably make a PCB for this at some point...
Day 6: The main benefit of feedback ammeters is elimination of the "burden voltage" seen with shunt ammeters. Magnetic approaches do not have this problem either, but have limited sensitivity unless they also introduce significant series impedances reflected through the current transformer.
The disadvantages are that the instrument has to source or sink the full measured current. For lower currents this is not challenging and the main non-idealities remaining are closed loop bandwidth related.
Presented here is a straight forward JFET op-amp TIA-based feedback microammeter with 5 transimpedances ranging from 1 mA/V to 100 nA/V. On the most sensitive range it may have a slight zero-offset, this can be corrected with an op-amp that supports an offset null, like the TL071 instead of the TL072 dual amp I used to simplify implementing the rail splitter. A much nicer modern op-amp would enable extending the range to picoamperes, but care would be required in the construction and shielding.
In the practical circuit I flipped the input terminals around as the TIA is fundamentally inverting, as discussed in the video this means the +ve input is tied directly to the -ve output which may present a problem for common grounded instrumentation, but is of little consequence as long as you know about it. The bandwidth of the instrument well exceeds 1 kHz on the most sensitive range, but has the option to low pass the measurement strongly at the control loop which can be useful for average DC current measurements. It can source and sink a bit more than 2.5 mA on the highest range, which is fine for its intended purpose.
The burden voltage of my multimeter on its low current range did not allow it to measure the calculator supply current, especially at slightly lower supply voltages, at around 1.2 volts it would not turn on at all with the meter in series! My old AVO MkIII did better on its 50 uA range, as did my MF47B, but both made the display dim a lot with their burden voltages. This circuit has about 3 mV burden voltage which is mostly offset and could be improved.
I should probably make a PCB for this at some point...










