1
\$\begingroup\$

I am using a comparator (TS3021) as a zero crossing detector (ZCD). The ZCD toggles its output based on the zero crossings of a sine wave generated across a transformer coil. One end of the coil is connected to the +ve input of the comparator via a voltage divider and the other end of the coil is connected to the -ve input of the comparator. The comparator is powered from a 2.5V LDO output supplied by a 3V battery source. The voltage swing of the sine wave is less < (VDD-VSS) of the comparator and doesn't violate the common mode voltage requirements (rail-to-rail) of the comparator. I tested this configuration and I was able to see the comparator output toggle at the zero crossings.

enter image description here

The comparator doesn't have a shut-down/enable pin and consumes around 73 uA of quiescent current. I do not need the ZCD to operate continuously, hence I tried to disable the comparator by inserting an NMOS transistor between the negative supply and the VSS pin of the comparator to cut the tail current. I connected the gate of the MOSFET to VSS to test if it disables the tail current. I noticed that the output of the comparator still toggles, but doesn't go all the way to the negative rail. Also, the VDS voltage is around 0.2V. When I connect the gate of the MOSFET to VDD, I am able to resume normal operation with full rail-to-rail output swing. My question is regarding the case when gate of the MOSFET is tied to VSS. Why is the output still toggling when I have cut (I assume I have) the tail current? Is there a better way of implementing shut-down functionality?

\$\endgroup\$
12
  • 2
    \$\begingroup\$ "Is there a better way of implementing shut-down functionality?" Yes. Use a comparator or opamp with shutdown. \$\endgroup\$
    – Olin Lathrop
    Dec 30, 2015 at 17:17
  • 2
    \$\begingroup\$ Probably current is flowing to GND through the inverting terminal (which is grounded). Technically, by grounding the inverting terminal, and cutting off the flow of current from the VSS pin to GND, you are exposing the comparator to a common mode voltage far below VSS, which is probably not allowed (unless this is a very special comparator). If you want to disconnect the comparator from power, you will need to simultaneously disconnect all the inputs and outputs from any source of power, including bias networks or pullups or pulldowns. \$\endgroup\$
    – user57037
    Dec 30, 2015 at 17:33
  • 1
    \$\begingroup\$ I don't think a CD4066 would work down to 3.0v reliably, but something like this could disconnect the other inputs quite easily. Easier yet is as Olin suggests, just get a comparator with a shutdown pin. \$\endgroup\$
    – rdtsc
    Dec 30, 2015 at 20:37
  • 1
    \$\begingroup\$ Not a good circuit because the top of the transformer coil will go negative with respect to ground so you are mistaken in believing it is OK. \$\endgroup\$
    – Andy aka
    Dec 31, 2015 at 1:01
  • \$\begingroup\$ @OlinLathrop I have looked into comparators with shut-down feature (linkLMV761). Please note that the circuit I am working on is a low power one (<10uA @ 2.5V). I am not getting the right combination of propagation delay and the quiescent current needed for my application. Also, the LMV761 has a turn-on time (from shut-down) of 6us, which is not desirable in my case. Then we have the ADCMP608 which has a turn-on time of 150 ns, however the leakage current during shut-down is around 250 uA, which again is not desirable in my case. \$\endgroup\$
    – ashare
    Dec 31, 2015 at 1:20

1 Answer 1

Reset to default
1
\$\begingroup\$

Here is an idea, based on your existing circuit. Basically, you need to modify the circuit so that all pathways to GND are cut off by the NMOS. See below. I think this may work, but I really can't say I recommend it. It would be so much easier to find a component with an acceptable quiescent current, or a shutdown pin. Also, if you do this, you need to make sure that the output of the comparator has no path to GND. It is OK if the output is pulled up to VDD, or is floating (connected to CMOS input of a powered up device), but there cannot be any type of pulldown or anything that will sink current to GND.

alternate-switch-idea

\$\endgroup\$

Your Answer

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

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