Best way to bypass a circuit without pop

[daniel1010]

Some good and enjoyable short reads on the subject:

http://www.muzique.com/lab/pop.htm
http://www.muzique.com/news/pulldown-resistors/
http://www.geofex.com/Article_Folders/box_pop.htm

I was a little surprised that the topic of true bypass has not yet been properly picked up at GroupDiy.

Exactly there I am also already mentally landed and had said yes an understood "silent" bypass to achieve will probably still take some time.

Thanky you.

 

[Tubetec]

With tube stages its really important to always provide a high resistance path to ground after a coupling cap , if you dont the cap can get charged up to plate voltage , that voltage discharging will cause not only a big loud bang but can also damage subsequent equipment .

 

[CJ]

Pultec eq has some popping problems also. 10 meg on the center pin was the answer. But any engineer who messed with the freq switch during a Streisand session would get picked up by the loony wagon and taken to the farm out in rancho Cucamonga,

 

[saint gillis]

Here's a practical example, how would you switch that ? (4 different options)

 

[Gold]

To get truly click free switching down to 20Hz you need either a zero crossing detector or a ramp/ramp down gain cell. Most requirements aren't that stringent but there are IC's made for large audio switching matrixes like used in broadcast.

 

[JohnRoberts]

Zero crossing detectors are generally overkill for basic audio bypass switching (you can get HF clicks from switching in the middle of a sine wave). There are techniques to minimize those clicks but probably more complexity than needed.

you need to manage the from and to audio nodes for 0 DC volt difference steps that can cause thumps and worse.

JR

 

[Gold]

If you are only concerned about switching with no audio present the above approaches are sufficient. It gets much more complicated if switching with audio present is a requirement. The lower the frequency that needs to be switched cleanly the more difficult it is.

 

[JohnRoberts]

It also matters what audio you are switching between... Back last century I designed a midi controlled patch bay with software controlled bypasses. When switching between the input and output of some effect carrying the same audio signal switching related perturbations were minimal. I did use an old click reducing trick****.

JR

**** to reduce HF content from step clicks apply HF pre-emphasis to the signal before switching, and complimentary de-emphasis after switching. The de-emphasis HF roll-off reduces the clicking.

 

[saint gillis]

you need to manage the from and to audio nodes for 0 DC volt difference steps that can cause thumps and worse.

I've added numbers to the circuits in my previous post, Ian mentioned the 4M7 resistor trick, as in circuit 3, but does that really minimizes DC across the switch's contacts?

 

[saint gillis]

The idea is indeed to prevent DC provoking oxidation on the switch's contacts after a certain period of use, because then you hear scratches in audio when you use the switch, right?

 

[JohnRoberts]

The idea is indeed to prevent DC provoking oxidation on the switch's contacts after a certain period of use, because then you hear scratches in audio when you use the switch, right?

No not as far as I know... Avoiding DC is to prevent clicks and thumps from the sudden step function.

JR

 

[saint gillis]

No not as far as I know... Avoiding DC is to prevent clicks and thumps from the sudden step function.

JR

Alright, thanks for the clarification

 

[Matt Syson]

Using a 2M2 (or any decent high resistance when the capacitor is a 47uF may still not prevent a 'click because unless the instantaneous level of the left and right signals are the SAME you will still experience a 'step' voltage change. in this instance a 'fast ramped FET' or opto resistor even would probably be more appropriate. At vast UK taxpayers expense (no doubt) the BBC in England worked out the maximum step level change related to the time to change was formulated as they were using stud contact 'potentiometers' (attenuators) for fading signals in and out. So there are a couple of mechanisms going on when switching. establishing there is NO DC potential difference (for when there is no audio signal) and then the rate at which any audio signal can change from one possible instantaneous voltage to another for which I think a 'fast ramp' of about a couple of milliseconds is deemed to be 'about optimal'. The BBC also defineda 'click' as an artifact that caused a PPM meter to indicate a 4dB 'peak' above the resting noise level with the switch in either position so switching on to off would have the gain boosted (after the switch) so that the measuring PPM would read 0dBu (scale mark 4) so a line amp output being switched may have about 90dB of gain added to 'energise' the meter!. There were 'allowances' for sharp 'tick' sound or low 'thud' disturbances.

The 4M7/2M2 resistors mainly work when you are switching poly (whatever) non polarised caps up to maybe 0.47uF

 

[Nick Nack]

At vast UK taxpayers expense (no doubt) the BBC in England

No UK tax dollars were spent- the BBC is funded by a licence fee, and various services it sells.