0
\$\begingroup\$

I'm building a uCurrent Gold clone (https://www.digikey.ch/reference-designs/de/sensor-solutions/current-sensing/2637) - with a digital interface (measuring the voltage directly with a microcontroller). However, I also want to automatically select the Shunt-Resistor by two solid state relay or two Mosfet's.

However, is this feasible? Is there a Mosfet with a very low Rds_on (<10mOhm), very low Gate and Drain leakage current and very low Drain-cut-off current, such that there are nearly no unwanted effects (similar to a regular switch with on resistance of ~10mOhm)?

This is what my intention looks like (out goes to input of OpAmp, pull-down on MOSFET should be 100k):

enter image description here

EDIT: ADDITIONAL PARAMETERS

Rds_on <10mOhm

Gate & Drain leakage current should be 0nA, since they would sophisticate results very heavily in the nA range.

Drain-cut-off current should be around 0.1uA

-> As there isn't any MOSFET with this characteristics, especially for the Gate leakage current, maybe a Photo-Transistor would help. But the Drain leakage current would still remain. I would also be satisfied with an electromechanical solution, like a relay or a reed-relay, as long as the mechanical dimensions aren't to big (Footprint of SOIC-8 would be acceptable).

\$\endgroup\$
4
  • 4
    \$\begingroup\$ Not claiming this is the best solution, but I would use a reed relay here. \$\endgroup\$
    – Hearth
    Commented Mar 21, 2020 at 13:59
  • \$\begingroup\$ If you are set on avoiding mechanical switches, you'll have to be clearer about what you mean by "low" for gate and drain current, and the bandwidth you are interested in (so we can see if device capacitance matters). And in what way the devices you have found in the obvious way fail to meet your needs. Otherwise this is both a lmgtfy AND a shopping question. \$\endgroup\$
    – user16324
    Commented Mar 21, 2020 at 14:25
  • \$\begingroup\$ Your best with a mechanical solution here... At least for the uA range. \$\endgroup\$
    – MadHatter
    Commented Mar 21, 2020 at 15:22
  • \$\begingroup\$ This IS a shopping question.. an electromechanical solution, like relay's would also be fine. However, I have never seen small relay's which would fit on a small PCB.. I will adjust the parameters. \$\endgroup\$
    – cortesis
    Commented Mar 21, 2020 at 17:03

2 Answers 2

1
\$\begingroup\$

As described by the maker of the uCurrent in this video https://www.youtube.com/watch?v=xSEYPP5Xsi0 , this may be a solution (I'll try it out):

Example Circuit

In this setup, the Rds_on doesn't affect the measurement (only the burden voltage)

\$\endgroup\$
0
\$\begingroup\$

This design is not true to the original as is. The original is also changing the measurement point when the shunt is changed. I can't really tell why unless the circuit is extremely sensitive to trace resistance. There are many low Rdson NMOS if you search digikey but you need to specify more parameters.

\$\endgroup\$
2
  • \$\begingroup\$ I assume its because they exchanged the switch from the previoud design (alternatezone.com/electronics/ucurrent/uCurrentArticle.pdf). In my opinion, it doesn't really matters because: 1. In the uA mode, the voltage drop across the 0.01 Ohm resistor is negligible (<1‰)). 2. In the mA mode, the current flowing into the MAX4239 is really small, therefore there won't be a voltage drop across the 10 Ohm resistor. \$\endgroup\$
    – cortesis
    Commented Mar 21, 2020 at 16:53
  • \$\begingroup\$ Agreed. Try using a reed relay. They're not too big. \$\endgroup\$ Commented Mar 22, 2020 at 17:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.