Analog switch for 12V

[I14R10]

Hello

I have a ultrasonic transducer that I want to use as transmitter and as a receiver. I need a switch that will alternate between supplying the transducer with power and after that "listening" to what transducer is receiving.

So far I found out there are analog switches, but unfortunately so far I only found ones that accept signals up to 5V, but I have 12V signal that's going to the transducer so those are not an option.

Second option is to make the switch myself, with MOSFETS, like in this circuit:

The problem is that I only can find MOSFETs with body not connected to source in CD4007 IC. I can use that IC but it seems like a waste of resources to use only 1 out of 6 MOSFETs in IC.

I'm now doing a simulations in LTSpice with regular mosfets but so far they don't work as well as those with separate body (substrate) pin unconnected.

Any alternatives, advices?

 

[danadakk]

What is part number of transducer ?

Regards, Dana.

 

[KlausST]

Hi,

there are many analog MUXes that work above 5V. Up to 40V isn´t uncommon.

Just go to your distributor´s internet site and do a parametric search on analog mux.
..or at any analog mux semiconductor manufacturer.

Klaus

 

[I14R10]

What is part number of transducer ?

Regards, Dana.

I'm not sure which one I will use. TCT40-16 is the one candidate but TX and RX transducers are a bit different. Apparently TX one's impedance is lower than RX's. I have to try how bad TX will act as RX and vice versa.

--- Updated Jul 10, 2022 ---

Hi,

there are many analog MUXes that work above 5V. Up to 40V isn´t uncommon.

Just go to your distributor´s internet site and do a parametric search on analog mux.
..or at any analog mux semiconductor manufacturer.

Klaus

https://lcsc.com/products/Analog-Switches-Multiplexers_618.html there are none above 5V

There are some at Mouser

https://eu.mouser.com/c/semiconductors/switch-ics/multiplexer-switch-ics/?supply%20voltage%20-%20min=10%20V~~20%20V&rp=semiconductors%2Fswitch-ics%2Fmultiplexer-switch-ics%7C~Supply%20Voltage%20-%20Min&sort=pricing

but honestly, I would like to avoid it if I can because of the shipping.


I'll go search the Mouser and try to find that part on aliexpress.

 

[crutschow]

The old CD4066 analog switch will work from 0V to 15V as long as the supply voltage and logic signals are also 15V.

 

[KlausST]

Hi,

you noticed that LCSC simply not filled their table with the datasheet specifications?

I recommend a distributor that fills the table data more carefuly, or as already mentioned: go directly to the manufacturers.

shipping: you don´t say where you live, thus we are unable to find out about useful distributors and part availablility..

Klaus

 

[dick_freebird]

Consider whether you truly need a "switch" with all it's idealities (or close as practical) or can you make a more application specific circuit.

In RF T/R modules there is often a "circulator" where each port throws signal to (say) it's clockwise neighbor and nothing CCW. Tx throws to Ant, Ant to Rx, Rx could throw to dummy / GND or a fourth port, etc.

An amplifier that can simply be told to STFU could be OK for half-duplex operation but few LNAs want to see 12V so some sort of gate protection, switched attenuation might be needed. Depending on BW and crest voltage you might be able to use less fragile Rx front end.

 

[I14R10]

The old CD4066 analog switch will work from 0V to 15V as long as the supply voltage and logic signals are also 15V.

Nice! Thanks.

Hi,

you noticed that LCSC simply not filled their table with the datasheet specifications?

...

shipping: you don´t say where you live, thus we are unable to find out about useful distributors and part availablility..

Not really. Now I feel stupid
I live in Croatia. I know about Mouser and TME, both of them ask about 15-20 euro for shipping.
LCSC has lower shipping cost, that's why I prefer it.

Consider whether you truly need a "switch" with all it's idealities (or close as practical) or can you make a more application specific circuit.

I will consider other options too, it doesn't have to be switch or multiplexer or whatever it's called. I just need to at some point disconnect transducer driver and make it a receiver.

 

[KlausST]

Hi,

Not really. Now I feel stupid
I live in Croatia. I know about Mouser and TME, both of them ask about 15-20 euro for shipping.
LCSC has lower shipping cost, that's why I prefer it.

12V MUXes like the CD4066, MAX4066, DG4xx, DG2xx ...are not rare. You should find some in oyur country. Ask around, a dealer, another electronics developer, a company... even in retired PCBs..

Klaus

 

[I14R10]

I have a question regarding DG417/418/419 https://www.vishay.com/docs/70051/dg417.pdf

I don't quite understand V_L pin. I believe it's logic level voltage, which is in the datasheet always put to be 5V.
If I'm using 3.3V logic, should it be 3.3 or also 5?

 

[danadakk]

The Vl pin seems to be TTL compatible, so if you used 3.3V logic levels
it will work but with reduced noise margin compared to if you were using
CMOS 5V logic levels.

Although part seems most specs characterized at Vl = 5V, but this seems
to justify using 3.3 -

Regards, Dana.

 

[KlausST]

Hi,

VL determinnes the switching threshold of the control signals.
The datasheet really isn´t clear with this.

However in Fig 11, they use 3V.

So if you have 5V available then connect VL to 5V.
If you have 3.3V available then connecti it to 3V3.

As long as your logic drive levels is higher than 2.5V (V_OH) then you are safe in either way.

Klaus

 

[fgavin]

Hello

I have a ultrasonic transducer that I want to use as transmitter and as a receiver. I need a switch that will alternate between supplying the transducer with power and after that "listening" to what transducer is receiving.

So far I found out there are analog switches, but unfortunately so far I only found ones that accept signals up to 5V, but I have 12V signal that's going to the transducer so those are not an option.

Second option is to make the switch myself, with MOSFETS, like in this circuit:
View attachment 177325


The problem is that I only can find MOSFETs with body not connected to source in CD4007 IC. I can use that IC but it seems like a waste of resources to use only 1 out of 6 MOSFETs in IC.

I'm now doing a simulations in LTSpice with regular mosfets but so far they don't work as well as those with separate body (substrate) pin unconnected.

Any alternatives, advices?

If you don't want to start from zero, there are already solutions to experiment with ultrasound sensors in Arduino. Look for HC-SR04 modules from your favorite vendors. You'll be taking measurements in less than five minutes.

 

[I14R10]

If you don't want to start from zero, there are already solutions to experiment with ultrasound sensors in Arduino. Look for HC-SR04 modules from your favorite vendors. You'll be taking measurements in less than five minutes.

Thanks but my point exactly is to start from zero. I also don't use Arduino.

 

[dick_freebird]

The Arduino just represents a "generic digital host"
in this context and there ought to be clues about
architecture and components embedded in the design
docs. That's the probable point.

You may want to be clear about just how "ultra" the
ultrasonic element is, and how much power you
will throw and lose (within the switch / mux) at it.
Those old CD4000 parts are pretty slow, at 12V
you may be looking at the even slower CD4000B
series, and CMOS digital parts are not big dice with
great thermals - you may see unacceptable drive
signal attenuation. You may also see cross-channel
contamination of Rx by "off" Tx if your frequency is
so high that ground and supply inductance, with
internal capacitances, degrades isolation (I'd bet
if it's even specified, it's at some stupid low frequency
like 1kHz) and the series R with your tranducer C
presents a final pole in the Tx lineup.

 

[I14R10]

The Arduino just represents a "generic digital host"
in this context and there ought to be clues about
architecture and components embedded in the design
docs. That's the probable point.

You may want to be clear about just how "ultra" the
ultrasonic element is, and how much power you
will throw and lose (within the switch / mux) at it.
Those old CD4000 parts are pretty slow, at 12V
you may be looking at the even slower CD4000B
series, and CMOS digital parts are not big dice with
great thermals - you may see unacceptable drive
signal attenuation. You may also see cross-channel
contamination of Rx by "off" Tx if your frequency is
so high that ground and supply inductance, with
internal capacitances, degrades isolation (I'd bet
if it's even specified, it's at some stupid low frequency
like 1kHz) and the series R with your tranducer C
presents a final pole in the Tx lineup.

40 kHz is the frequency. https://www.vishay.com/docs/70051/dg417.pdf page 6 says about 100dB crosstalk isolation for 40kHz.

LTSpice simulations predict about 20mA at around 12V into the transducer. If I calculated right, that's around 0.5V lost in the switch. I can live with that.