Analog Mux/Switch Versus Relay for bipolar signal : Advice needed !

[prateek_k_chd]

Hi
This is what I'm trying to achieve :


I need your advice for the "Selector" in the above circuit.

I need something with
a) low resistance (Pref. less than 50 Ohm)
b) Commonly available (not the here-today-gone-tomorrow parts)

Can anyone suggest something other than a relay(They would take up lot of space)?

I almost decided to go for HC4066,before realizing that it can't do bipolar signal

Thank you so much guys for your very valuable advice !

Cheers

 

[prateek_k_chd]

I am basically trying to measure the (widely) varying feedback resistor. If anyone can suggest a different method (that doesn't employ DC voltage) then that would be great too !

 

[prateek_k_chd]

anyone ????? :sad:

 

[kak111]

Select here..........

**broken link removed**

**broken link removed**

Multiplexers (Muxes) | Switches/Multiplexers | Analog Devices

Analog Switches | Switches/Multiplexers | Analog Devices

 

[dick_freebird]

You are not saying a thing about the signal, or the system,
that would drive selection.

You can get relays in the same form factor as a packaged IC
(at least DIP-scale).

A relay will be more resilient to many faults, will default to
an open condition with power off, and does not require
that your control and signal have any common connection.

An IC mux may, or not, provide specific fault protection.
Some offer power-off high impedance as a special feature.
All constrain the input signal to be within the supply rails
for normal operation (low impedance on condition). An
IC mux will almost certainly have lower supply current
demand than a normal relay, more than a latching relay
if the toggle rate is low.

 

[barry]

I think you need to give us more info. If you are trying to measure a resistor, why are you using a sine wave and all that other stuff? Just put a constant current source into your resistor, and measure the voltage, right? Why not DC?

 

[poorchava]

I'd say you could use a multiplying DAC. It's a which is designed to have reference voltage rapidly changing. Thus when varying control word you actually vary the fraction of reference voltage to output, since DAC's are esentially digital potentiometers.

What might pose a problem is voltage levels. Your signal is +/-5V sine which is unacceptable for most DAC's. I think that you would need to divide the voltage and add some common mode or play with voltage levels in some other way.

There are multiplying DAC's which are capable of handling this situation as-is, for example AD55xx and AD54xx but they are expensive.


On the other hand you could use a variable gain amplifier such as LMH6505 for example (but that's a one with 120 MHz BW, a bit of an overkill)

 

[prateek_k_chd]

Sorry I didn't state the scenario....I thought that detailing the design might make the post too large ...

I'm building a (liquid) conductivity meter - conductivity is essentially the inverse of resistivity. So i'm measuring resistance. But I can't use DC as that will cause electrolysis.
So I'm using the liquid's resistance as the feedback resistor of an op-amp (which is wired as an inverting amplifier & amplifies a bipolar wave of a fixed voltage -> +5v and -5v)
Now since the value of the liquid's resistance can vary wildly (between 1 ohm and 1 Mega ohm) , I need some way to select between one of the other few resistors that go from opamp's -ve input to ground (so as to select between different ranges of resistivity)

The toggle rate is going to be low and i can afford ~25mA of current. So that makes relays a decent choice. But they increase the cost significantly and hence I wanted to know if there is any cheaper solution that uses commonly available parts.
All the ICs I've seen (4066 etc etc) either have very high ON resistance (1K or more - and that too is supposedly varying ; which means I can't compensate for it) or they can't do bipolar waves or they are just very exotic- not commonly available and/or expensive

Anyways,sorry again for the confusion and thanks for the replies !! I guess relays are the best choice in this case.

 

[chris109]

I think you could use TRIAC's, they are cheap AC switch's.
And the gate current would be less than a relay, also the on resistance should be low.

I have little experience with them though, so hopefully someone on here can either confirm if I'm right or not.

Regards
Chris

 

[prateek_k_chd]

Hey I didn't think about that ! Though me too don't have enough experience with them...can anyone here please confirm what chris109 said??
Thank you !!!

 

[Sajjadkhan]

Your problem is not that complex. The op-amp input is really high so Analog multiplexers resistance is negligible. Look at these Multiplexers below

Multiplexers (Muxes) | Switches/Multiplexers | Analog Devices

they have R in ohms.

Even if you use Relay or Analog MUX, just be care fully with the switching delays (if you are feeding the output to some ADC or likewise).