[volt-nuts] How to measure micro-amp currents and have low impedance?
bruce.griffiths at xtra.co.nz
Sat Sep 26 23:56:12 UTC 2009
Bruce Griffiths wrote:
> Bruce Griffiths wrote:
>> Use the Patterson transdiode connection with npn and pnp transistors as
>> feedback elements.
>> It works much better/more predictably at low currents than when signal
>> diodes are used.
> To handle 1A either a power opamp or an opamp with a high current output
> buffer will be required.
> Transistors with a collector current rating of 1A or more will also be
The best solution is to use a log amp or equivalent for the 1nA to 10mA
range and a conventional current shunt with a DVM or equivalent to
measure the 10mA to 1A range where log amps dont work to well unless one
takes heroic and expensive measures. The log amp and the conventional
shunt are connected in series.
The real problem is to ensure that the voltage drop across the log amp
input is low even when the current exceeds 10mA.
One way of doing this is to use a high current shunt diode across the
log amp inputs. When the log amp isn't saturated the voltage across this
diode is very low as is the associated diode leakage.
When the log amp opamp output current or voltage limits the voltage
across the diode increases turning it on.
However this introduces a diode drop in series with the supply at high
currents. In principle one could sense the voltage across the current
shunt and use this to turn on either a high MOSFET shunt switch
connected across the diode shunting the log amp inputs. This can in
principle reduce the voltage drop to 10mV or less with the right choice
of switch. Depending on the desired response time one could even use a
relay for this function.
More information about the volt-nuts