0
\$\begingroup\$

We have designed a circuit for a bidirectional current limiter using an N-channel JFET as below:

enter image description here

This limiter is connected to the end of electrode used for EEG measurement.

The requirement for the product is to measure the skin to electrode impedance also.

Idea behind using the circuit was, the current from the current source would pass through resistor R13 which would generate a negative potential at V(G-S).

This in turn will control the amount of current flowing through the circuit.

The current input is in the range of uA. Typically 6uA.

For testing purpose we connect a resistance ranging from 1k to 330k so as to simulate the skin to electrode impedance.

During testing I found that the resistance of the JFET is non-linear for different temperature ranges.

Due to this, the error in calculating the actual impedance is more than 25%.

Can anyone please help with ways to reduce the temperature based non-linearity of the JFET, or else, a circuit which would limit current bidirectionally to about 80uA?

\$\endgroup\$

1 Answer 1

Reset to default
1
\$\begingroup\$

If all you need is about 80 μA, I wouldn't bother doing it manually and just use the ever-handy REF200 from TI (formerly Burr-Brown). The IC contains two 100 μA current sources and a current mirror, but all you need is one of the current sources in conjunction with four diodes (or both of the current sources along with two diodes) to create a bidirectional 100 μA current limiter. (see figure 17 in the datasheet.)

The current sources inside the chip are internally compensated for low temperature coefficient (datasheet claims 25 ppm/°C typical), so you shouldn't have to worry much about that. The variation of current with temperature is nonlinear, however (figure 1), so be aware of that if you do need to account for the few hundred nanoamps of temperature drift.

The price may be a downside. At about $10 each in single quantity, it's far from the cheapest method of doing what you need--but it may well be among the easiest.

Also be aware that this method is limited to 40 volts of compliance voltage; above that point, you're exceeding the part's limits and it likely won't survive.

\$\endgroup\$
8
  • \$\begingroup\$ Thanks for the prompt reply @Hearth. I guess there is a confusion or some data is missing in the question. Let me reiterate. The 24 bit sigma delta ADC which we are using for EEG measurement has an internal current source which provides a current for different amplitude (6nA, 6uA, 24nA, 24uA). For our application, we found that 6uA of source current works well. When this current passes through R13 as mentioned in the circuit diagram, there is a negative voltage on VGS which limits the current if it is more than 80uA because of any fault condition. Contd.. \$\endgroup\$
    – Sushant
    Commented Jun 30, 2021 at 2:53
  • \$\begingroup\$ I am facing problem due to the non linearity of the JFET. Is there any method to reduce the non linearity over temperature? Or else can you please suggest me a circuit for bidirectional current limiter which limits current more than 80uA? The circuit need not have a current source as we are already using the IC's internal current source. \$\endgroup\$
    – Sushant
    Commented Jun 30, 2021 at 2:56
  • \$\begingroup\$ @Sushant A current source and a current limiter are exactly the same thing--what you are describing is exactly what the REF200 does. \$\endgroup\$
    – Hearth
    Commented Jun 30, 2021 at 2:57
  • \$\begingroup\$ I am not sure how a current source can be used in place of limiter. \$\endgroup\$
    – Sushant
    Commented Jun 30, 2021 at 5:15
  • \$\begingroup\$ @Sushant Just think of this as a current limiter then. That's what it is. \$\endgroup\$
    – Hearth
    Commented Jun 30, 2021 at 13:52

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.