audio electronic switch from Maxim

[Dimitree]

hi everyone
I noticed this device:
https://www.maximintegrated.com/en/products/analog/analog-switches-multiplexers/MAX4762.html

I'm already aware about:
1) you need 5V to power it, 9V would kill it
2) the IC package is enemy of DIY
3) you may need additional circuitry to interface it with a switch (as usual with electronic switches)
4) other possible downsides: price, availability, etc

apart from that, what about the specs?
are cross-talk and off-isolation good enough? I don't have any examples to compare to.
I'm curious if it would be suitable for guitar signals..
Would it work at its best without the need to use additional conditioning parts on the "signal" pins, in order to use it just like a mechanical switch?

thank you

[armdnrdy]

Look at SSM2402 & SSM2404 audio switch ICs from Analog Devices.
They both run at a much higher voltage.
Unfortunately...they are both out of production...but...can be found as NOS.
See if you can research a replacement.

[Dimitree]

Thanks, yes I know them. But I' trying to understand if those Maxim ones are good enough too. As I said, package/price/supply V are not a problem.
What I don't understand is if those can be used without signal conditioning, even with passive guitar signal

[anotherjim]

The specs are fine for this side of the mixing desk. Better than 60dB xtalk etc is plenty.

What have they got that can't be done with CD4053?

What I see for the Maxim part is very,very low on resistance. Very suitable for switching low voltage speaker/headphone feeds where you don't want to lose any signal volts in the switch.

For signal (not speaker) switching, I'd stick with a 4053.

[Dimitree]

The pros, in my application, comparing to 4053 or SSM2404 or similar, is that those have a much smaller footprint. Also, if I can avoid in/out capacitors + bias resistors, that would be perfect.

that's what I am trying to understand.
And also what the max input AC source should be. I guess the signal can go below 0V, but how much?

[Rixen]

Although not in a small footprint, the TLP175 (and ilk) is an opto-relay (MOSFET output) from toshiba which has input completely isolated from output, no power supply and hence no bias or buffering requirements. Closest thing to mechanical contacts you can probably get in silicon.

[Dimitree]

This one looks cool. I just need to understand if I can adapt my application for them. I'd need 4 of them and complementary control signals, to build a DPDT. That means higher costs and higher footprint.

Any reason why you guys don't like that Maxim device and suggest me different parts? :-D

[PRR]

Interesting. Yes, it would take +/-4V audio with only a 5V supply.

Note that the "off" position throws a ~~100r resistor to ground. That would discharge leakage voltage on an "off" output, good. Also clamp strong signals against sneaking into other paths. However if you wanted to USE that off signal for some other purpose (in complex switching) it would be frustrating.

I'm not worried abut crosstalk etc.

[Dimitree]

sorry for the ignorance, but how should I understand from the datasheet that it would take +/-4V?
I read that "Analog signal range" is min VCC -5.5 and max VCC, doesn't that mean -0.5V to 5V, if VCC is 5V?

about the "off" resistance to ground, how that translates in common DPDT applications that we usually use in a single stompbox bypass?

[anotherjim]

That's how I read it. If Vcc were 3v, it can pass signal that dips negative by -2.5v.

[PRR]

I mis-read.

At 3V supply you could swing to -2.5V, which does pass most guitar-cord signals.

[Dimitree]

ok, so I'd say it's better run it at 3V instead of 5V for guitar application, where signal is bipolar

I think I will study the TLP175A opto relay way..

[Rob Strand]

What I see for the Maxim part is very,very low on resistance.

+1 on what Jim said.

4053's a are pretty good but do have their limits.

Analog devices make a lot of switches.  Much better performance and wider applications than the older switches.

http://www.analog.com/en/products/switches-multiplexers/analog-switches-multiplexers.html

[anotherjim]

Where you don't want the trouble of changing the DC bias and let the signal keep it's 0v bias, it does look as though around 3v might be optimal. If you want maximum head room, then you could bias in and out with resistors to 1/2 supply and use 5.5v for 2.75v (Error!  I mean 5.5v) peak to peak headroom, just as you would have to with a CMOS switch.

[Dimitree]

my main goal was to remove the decoupling capacitors on the switches.
I guess a better way then is to use a charge pump and power the switch with bipolar supply

or choose the photorelay

[anotherjim]

Ok, how much headroom do you need? Most guitar pedals might put out a few volts peak to peak, although some minimum parts "play with logic/timer chip"  or 386 overdrives might give 8v p-p (because they are very naughty pedals and need a good padding down).

[Dimitree]

not that much for guitar, but I usually use the same pedalboard for my synths too, and so I'm a bit worried about the clipping on bypass (the effect would clip anyway..)