0
\$\begingroup\$

I currently have a 2:1 switch / multiplexer which is being used to switch between two different batteries which are to provide power to a small module on my board. The switch that I am using is the TS5A3154 and I am wanting to switch between the channels using an output pin from my microcontroller (EFM32G222F128).

The issue that I am having is that the battery voltage is going to be approx. 2.7 - 3V which will be at both of the analogue inputs (NC and NO) and the supply voltage for this switch is going to be only 2.5 - 2.8V - due to V+ coming from a dual schottky diode which draws the current equally from each battery and also prevents on discharging through the other in the event of a battery failure so this part is needed.

Because of this small difference between the analogue inputs and V+, the Enable pin does not function properly (ie. when there is a high signal at the pin there is a high output when it should actually be off completely) and because of this I need to replace the switch with either a new component or a new solution which would only have a small voltage drop across it or be very low resistance.

I have already had a look around for alternative switches or multiplexers but it appears to me that the same problem would always be present for any alternatives that I looked at.

Low cost is also fairly important but at the moment I am more interested in any solution as opposed to a low cost one. The solution would also need to have a similar Enable function the current switch has in that the power to the module can be cut off at some point - either just by turning two separate things off or having one switch that would do it.

A schematic of what it currently is shown below might be of help for understanding a little more:
enter image description here

\$\endgroup\$
1
\$\begingroup\$

Why not use two devices then each can be powered by their individual local battery supply.

If not then TI have a range of power/load switches that may be suitable. I'm thinking of the TPS2080 listed on this page amongst other devices that may also be suitable. The TPS2080 is a dual supply with <0.1 ohms on resistance and supports a change over contact. It has a built-in charge pump that should overcome the problem you have.

If all else fails then create a low power "higher" voltage rail and use this for the chip supply.

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

Would it be worth placing a diode in line with the NO and NC pin connections to drop the voltage from the batteries to nearer or slightly below the common V+ value ?

Are you sure the problem is not that the control signals are higher then V+ ? A simple resistor voltage divider on the signal pins could solve that.

\$\endgroup\$
  • \$\begingroup\$ If you have any continuous pulses in the system you might use that to create a small charge-pump input to the V+ pin. A small cap on the V+ pin will help store the charge. \$\endgroup\$ – Nedd Jan 14 '15 at 14:17
  • \$\begingroup\$ It can't really afford to have a diode drop as the module it will be powering needs to be on for as long as possible and a drop in voltage would already reduce that time significantly. And it wouldn't be constantly pulsing - only every now and again \$\endgroup\$ – MrPhooky Jan 14 '15 at 15:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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