Gating a signal on pulse

[Mo3hre]

Hi,
I was hoping you could help me with a non-guitar pedal project. I am trying to gate a signal (white noise at around +/- 10V) with a short (500 ms) pulse that I trigger from an Arduino. How could I implement this gate? I'm thinking of a device that is closed unless it receives a pulse, then stays open and passes the signal as long as the pulse is high, then shuts if the pulse is low again. The time it takes to switch is important and should be on the order of microseconds (or faster). I have the strong feeling that this is the most basic TTL operation, but I'm a complete newb here! I assume a transistor is what I want? If so, what transistor would be suitable and to which pins do I deliver the signal/pulse/ground?

I greatly appreciate your help.

[GibsonGM]

Short on time (have to go to work), but maybe you would look down the avenue of holding an arduino output high, and when you want the pulse, bring it low - triggering a 555 timer that is set up to give your 500ms pulse as its output. This pulse could be used to open/close a fet used as a gate similar to the Boss pedals' bypass switching (without the flip flop stuff).   The arduino can be set up to oscillate H/L, or do it on a timer, or by mechanical switch. 

"Switch A"...  https://www.electrosmash.com/tube-screamer-analysis#jfet

I'm not sure if this would meet ALL of the criteria, but it may offer one possible approach, HTH....if this is something you think would work, you will certainly get more info on the specifics here!

[Mo3hre]

Thanks for the suggestion and the link! I realise I wasn't clear in my first post: The Arduino is already set up to deliver a 500 ms TTL pulse, I just need a gate that gates a white noise signal as long as the pulse is high. Could I just deliver the pulse to the base of a NPN transistor? With emitter and collector in the signal path?

[amptramp]

It sounds like a diode bridge gate would be best for the signal levels you are looking at.  This diagram shows a pair of diode bridges being used to select between two inputs.  If you look at the bridge on the left-hand side, you should imagine the capacitor in the bridge is not there and the diodes in the bridge being the only ones there.  If the upper EFF/BYP line is high and the lower line is low and both amplitudes are larger than the signal amplitude, all diodes are reverse-biased and the signal does not go from BYPASS to the op amp input.  If the upper EFF/BYP line is low and the lower EFF/BYP line is high, all diodes conduct but the signal is loaded with the parallel 100 K impedance (making it 50 K) so the source must be able to drive this load in addition to whatever is connected to it (in this case, the op amp input resistor of 100 K), making the total impedance of 33K for the source.



This was originally part of a thread on silent switching for effect/bypass selection here:

https://www.diystompboxes.com/smfforum/index.php?topic=120006.msg1122294#msg1122294

but the principle is the same.  You can use a flip-flop and a level shifter to get the signals up to the amplitude you need.

[Vivek]

Audio Mux CD4000 ?

[anotherjim]

Swing VCA?

Depends on if you care about distortion or clicks. If you just snap it open or shut with a logic control you will probably get a click. A VCA means you can slew the rate of change of the on/off control with a simple RC filter so there's a more gradual open/close. This can be very fast to do the job, around 20ms.

[ElectricDruid]

Jim makes a good point, Mo3hre. If you open and close the gate as fast as you suggest (usecs) you *will* hear clicks. In order to avoid this, you either have to only switch at a zero-crossing in the audio, or you have to do a very slight fade-in and fade-out over a period of milliseconds.

What's the need for usec switching?

[Mo3hre]

The white noise pulse is part of an experimental paradigm where I record brain activity at a sampling rate of 30 kHz. Since I'm synchronising the pulse with that data, it has to be as fast as possible. A click wouldn't be too bad, I assume. Thank you so much for the suggestions, it'll take me some time to read through them and understand them. I was kinda hoping that a single transistor would do the trick...

[ElectricDruid]

Ok, if it's a simple on/off for a noise signal, it's true that a bit of click isn't going to be the biggest problem.

The Roland swing-type VCA Jim posted above would probably do the job, and that's a single transistor. But it doesn't deal with bipolar signals, except by treating them as unipolar. For audio, sometimes you don't care - the signal will be connected in and out via a capacitor, so the offset doesn't matter. It's cheap'n'cheerful, but if it does what you need, it's a good solution. Roland certainly thought so, so who are we to argue?!?

Otherwise, you could use a single SPST CMOS switch (there are loads: https://pt.mouser.com/c/semiconductors/switch-ics/analog-switch-ics/?configuration=1%20x%20SPST). The DG series work with +/-15V supplies and use 5V for the switching logic, so that'd be a good choice for a +/-10V signal and would interface to the Arduino directly.

How about?:

https://pt.mouser.com/ProductDetail/Vishay-Siliconix/DG418DJ-E3?qs=6sq0UYnmXrtm00TkQT9cZg%3D%3D

or

https://pt.mouser.com/ProductDetail/Vishay-Siliconix/DG419DJ-E3?qs=3bFTlaDNFuV0juGT9hJcaQ%3D%3D

First is SPST, second is SPDT.

Switching times for both of these are sub-200 *nanoseconds*, so no problem there.

HTH

[Mo3hre]

Awesome, thanks! I'll try both the VCA and the SPST switch and see what works best.