Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

THP210: Unwanted leakage currents affecting charge amplifier design

Part Number: THP210
Other Parts Discussed in Thread: TMUX1134

Hello, I have taken over an amplifier design, on which questions have already been asked regarding stability (see THP210: Stabilisation of a current-to-voltage differential amplifier (with a 20KHz bandwidth and a pass-band gain of +25dB) - Amplifiers forum - Amplifiers - TI E2E support forums).

On a THP210 development board, the circuit did not initially perform well - we had a large DC offset on the output and both input and outputs were slowly moving around by a volt or so. We stabilised this by adding 1Meg resistors between Vocm and the +ve and -ve inputs to the device.

This now performs with excellent symmetry on the output signals and, when floating, the inputs before our input series resistors, are at Vocm (2.5V) as expected. THD is good - typically better than -105dB


However, in our design, we have 24-channels of this circuit on a small 32mm x 150mm PCB, and for test purposes we have the input signals going through analogue switches. We cannot get a channel on our PCB to have a zero DC offset. The inputs sit at around 4.4V, and there is often a difference of a few 100mV in the mean output signal levels.

If I take the analogue switches out of the circuit and inject the signal directly onto the 500R series resistors, then the circuit behaves as expected.
If I simulate 10nA of leakage (e.g. a possible leakage path from an analogue switch) into the circuit (see attached) then a transient simulation is very close to what I see with a large DC offset at the output

I appreciate that this then isn't really a TI problem. but I wonder if you may be able to offer some advice regarding the use of analogue switches in this kind of circuit / how it might be desensitized to such unwanted currents. Clearly adding a 2nd current (flowing in the opposite direction)  to the other input cancels the effect, but I can't see a practical way to achieve that.

I am in contact with Vishay to get a better understanding of the leakage paths involved (DGQ2788AEN-T1-GE4).

Many thanks in advance,
Nick


ChargeAmp_THP210_with_leakage.TSC

  • Hello Nick,

    As you have explained, when the THP210 FDA circuit is configured into a current-to-voltage configuration with a 500MΩ feedback resistor, the circuit will be quite sensitive to any leakage current of the analog switches.

    We offer low leakage, precision analog switches.  For example, the TMUX1134 offer relatively low leakage current is in the ±0.1nA max (±0.003nA typical, at room temperature and ±0.350nA over the -40C to 85C temperature range. An analog switch with low leakage should reduce the errors in the circuit. Below is a link with a list of low-leakage analog switches offered by TI.

    https://www.ti.com/switches-multiplexers/analog/products.html#3306=1.8%3B2.5%3B3.3&3193max=0.0004%3B0.005&sort=3193max;asc&

    If you have detailed questions about analog switches, you could submit a post to the Switches and multiplexer forum link is below.  (I can also refer this e2e request to the switches and multiplexer team, if you wish as well.) Below is a link to the switches and multiplexer E2E forum, where the team may provide you detailed suggestions:

    https://e2e.ti.com/support/switches-multiplexers-group/switches-multiplexers/f/switches-multiplexers-forum

    It may be worth mentioning, although, I am not certain if these would be acceptable in your design, you could also consider reed relays. Reed relays offer a lower leakage and more ideal switch impedance characteristics when compared to solid state switches, although the switching speed / settling time is much slower, and they have other disadvantages.

    Thank you and Kind Regards,

    Luis

  • Hi Luis, Many thanks for your response. I will follow your advice and try the TI switch you have recommended. Reed switches (even SMT) are physically too large for our application. Thanks also for the link to the other team, I'll look into this.
    regards,
    Nick

  • HI Nick,

    Thank you, 

    Kind Regards,

    Luis