Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

"Clickless" Analogue Switch

Status
Not open for further replies.

vielle568

Member
Anyone out there know of an analogue switch that's "clickless/popless" and in a DIP package? I am looking for a switch to handle audio signals in a prototype circuit and the current analogue switch (MAX4619CPE) makes far too much noise. Ideally an IC running off a single 5V supply would be prefect. I'm not bothered if it's SPST or SPDT or how many channels it can handle because the wiring can be changed to suit the component.

All I can find are switches in other packages and I can't figure out how to mount them into the prototype. Most of them are far too small as they are intended for use in mobile phones.

Any suggestions welcome. Thanks!
 
There is nothing inherently noisy about that IC vs any others. Perhaps a more likely problem is that the signal is coupled to it through capacitors for DC blocking, and the charging and discharging of these capacitors is causing a popping sound. This is fairly common. The way to fix it is to DC bias the input and output of the switch so that there is no charging or discharging when the switch changes state.
 
Hi RadioRon,
Thanks for your rapid reply. Yes, you're probably right. The input signals to the analog switch do have a capacitor in line and could be causing the noise as you suggest. Sorry to admit it but I'm an ME and not an EE, and so how do I add the DC bias to correct the problem? :)
 
The easiest solution depends on the circuit that you have so far. It might be easier if you can somehow sketch the circuit that you have around the IC and post that.

The key to fixing this is to ensure that the DC voltage on either side of the switch is the same whether the switch is open or closed. In this way you avoid any charging or discharging of capacitors.


For now, I'm going to guess that they have one side of the switch biased to around 2.5V DC and the other side simply has a coupling capacitor going to the IC pin. You could ensure the same DC on both sides by connecting a 100K resistor to that one side of the switch, followed by a 1uF capacitor to ground and then another 100K ohm resistor to the other side of the switch.

This suggestion is a crude fix. The problem with this is that it compromises the isolation of the switch at frequencies below the cutoff frequency of the RC filter created by the 100K resistor and the 1uF cap. However, this cutoff is only about 2Hz so it might be just fine. At frequencies above 50 Hz the isolation would be up to about 45 dB (rough guess). You can always increase the values to get better isolation.

A more ideal fix is to bias both sides of the switch through 100K ohm resistors from a low impedance source of 2.5V like a regulator or opamp voltage buffer.
 
Last edited:
You need to operate the switch slowly through a linear region like they do in pro audio gear. Try adding a RC filter to the gate pin of the switch, so it ramps on/off over a few mS, you need more than 5mS to avoid a click, 20-50mS even better still.

I haven't used that MAX4619 switch but many of the older style analog switches were like a FET you can just drive the gate linearly to provide a slow turn on ramp.
 
Thanks for your suggestions.

I've attached a couple of diagrams of the relevant parts of the circuit; one showing the analog switch control system and the other listing the three different analog signals passing through the switch.

What am I building? The circuit is being used in the yellow instrument shown in the photo; it is used as an electronic rhythm generator replacing the string on the acoustic version. A series of small holes are drilled around the perimeter of the wheel and these act as an on/off for an IR LED and a sensor. The wheel speed (frequency) is monitored by a LM2917N speed switch that triggers when the crank is accelerated. This acceleration is a brief pulse and is controlled by the musician; it connects pin 5 of the speed switch to ground and is used to activate the analog switch. The analog switch in turn has a selection of inputs that will pulse when activated and link to an audio mixer (along with the melody and drone signals) and out to the amplifier.

Switching frequency is low, >20Hz maximum and audio signals will be under 20KHz.

The instrument is built and the control part of the circuit seems to work fine. I mentioned that there was a capacitor in the input line to the analog switch. Sorry, I made an error. The cap is at the input to the speed switch; there are no caps on the input lines to the analog switch.

I realise that there are probably several errors in the circuit design. If someone would kindly correct my faults I'd be most grateful. Many thanks!
 

Attachments

  • numérisation0001.jpg
    numérisation0001.jpg
    1.2 MB · Views: 1,613
  • numérisation0002.jpg
    numérisation0002.jpg
    1.7 MB · Views: 733
Last edited:
I don't know if this will help. I made a 4-channel aircraft intercom several years ago and during the initial design had the "click" problem each time the VOX switched in a channel. By biasing the capacitive-coupled input amps at Vcc/2 and also the capacitive-coupled output of the analog switch at Vcc/2, the click problem disappeared.

https://www.electro-tech-online.com/attachments/acintercom-pdf.25999/

Sorry about the PDF image. If you zoom-in in Acrobat or Foxit it's easy to read.

Ken
 
From the data sheet, it looks as if the MAX 4619 is of similar construction to 4000 series CMOS analogue transmission gates like the 4066 or 4051/52. If that is so it is very difficult to eliminate all the switch clicks. Slowing down the 'gate' turn on time is not an option with this type of analogue switch. The attached jpeg, from a very old National audio data handbook, shows some of the detail of the FET alternatives - and describes the transmission gate limitation.

Having said that, it is possible to get pretty close to silent switching from these ordinary analogue switches, if you follow KMoffett's suggestion of maintaining a fixed DC high resistance 'half' rail' bias, on both sides of the switch - and capacitively couple the signal in and out

The other problem you have, from the diagrams you have posted, is that it appears that you are switching the output from the pickup directly into the analogue switch. I'm guessing there's quite a lot of gain after that, and even with the DC bias problem dealt with, the remaning switch 'clicks' are going to be magnified --probably to an unacceptable level.
Ideally, you need to do the switching after the pickup gain stage -it's going to be very difficult (if not impossible) to get rid of the switch clicks otherwise.
 

Attachments

  • Noiseless switching.jpg
    Noiseless switching.jpg
    504 KB · Views: 844
Last edited:
Thanks for your helpful suggestions. I can follow what you've done biasing the intercom circuit and also the data attached related to the noisless switching designs will be useful too. I shall add the pre-amp for the pick-up prior to the analog switch because you're right, there's no need to amplify any extra noise that's generated.

It seems that my last MAX4619CPE stopped switching this evening. I don't know why but the IC no longer functions. I have several others that have all had a similar short lifespan. I don't know if I am overloading the chip or it is ESD or what (the signal levels are all within the datasheet limits). Anyway, perhaps I should just replace the switch with another. I can't find a "popless" analog switch in a DIP package but the MAX4514CPA
https://www.electro-tech-online.com/custompdfs/2009/10/MAX4514-MAX4515.pdf
can be used for analog signals and it will run off a 5V supply. It is only a single SPST switch in a 8-DIP but one switch for each audio signal would be fine.
 
Some digital potentiometer circuits wait for a "zero crossing" in the audio then switch when there is no voltage. The result is pure silence during the switching.

Some circuits ramp down the volume, switch, then ramp up the volume.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top