Gain switching, avoiding audible transients

I need to switch the gain of an audio preamp using a reed relay, Kv =
3 (low), and 10 (high). It's important to avoid audible pops during
the switching. The device involved is a lo-noise audio opamp,
NE5534AP, using the relay to adjust one of the resistance values by
switching a second R in parallel. I'm wondering which configuration
would be most likely to be pop-free - inverting/noninverting, which R
to change. Or another scheme? Any other hints?

 

[Phil Allison]

nor...@googlegroups.com wrote:

I need to switch the gain of an audio preamp using a reed relay, Kv > 3 (low), and 10 (high). It's important to avoid audible pops during
the switching. The device involved is a lo-noise audio opamp,
NE5534AP, using the relay to adjust one of the resistance values by
switching a second R in parallel. I'm wondering which configuration
would be most likely to be pop-free - inverting/noninverting, which R
to change. Or another scheme? Any other hints?

** The NE55534 is a low noise and very low THD op-amp, ideal for low level and line level audio, but it has one problem - high input bias current.

The bases of two input transistors require around 1uA of current each and this has to be supplied via feedback and ground reference resistors. With typical values of say 100kohms, a 1uA current produces a voltage drop of around 100mV. Suddenly reducing the value of the feedback resistor by 3 will produce a very audible pop of about 60mV at the output pin.

The simplest solution is to put a film capacitor in series with the switched resistor to isolate this offset so the gain changes for audio signals but not for DC voltages. This cap must be kept charged to the offset voltage by a high value resistor of say 1Mohm.

Another issue is with the coil of that reed relay, applying or removing the drive voltage suddenly will induce transient voltages in nearby wiring or PCB traces. Better have that voltage rise and fall over some time period - 50mS at least.

Thirdly, the mere act of switching the level of an audio signal creates a fast transient which will be audible under certain conditions as a distinct click.

The solution that solves *all* the above simultaneously is to use a suitable opto-coupler instead of a relay.

A LED driving a small photo resistor ( aka CdS cell)connected in series with your switched resistor will do the trick. Typical CdS cells have on resistances under 1000 ohms and off resistances in the megohms. Switching times are in the order of 100mS. No clicks or pops are possible as it is just like turning a potentiometer up and down by hand quickly.

Jaycar and others have suitable CdS cells - also know as LDRs.

You can easily make a opto-coupler by facing the LED against the cell and covering the pair in black shrink tubing to keep light out.


.... Phil

 

[Phil Allison]

Chris wrote:

If you're game and can deal with the various crackpots there,
sci.electronics.design actually has some knowledgeable regulars that
could give some good advice...

** The entire world of audio seems to be a complete mystery to regular posters on SED - myself being a singular exception.

John Larkin for one is openly hostile to its very existence.

Like "Sheldon Cooper" with his distain for Geologists.

" The only people who are glad when others take their work for granite ... "



..... Phil

 

[Adrian Jansen]

On 27/11/2014 12:00 AM, Bruce Varley wrote:

I need to switch the gain of an audio preamp using a reed relay, Kv =
3 (low), and 10 (high). It's important to avoid audible pops during
the switching. The device involved is a lo-noise audio opamp,
NE5534AP, using the relay to adjust one of the resistance values by
switching a second R in parallel. I'm wondering which configuration
would be most likely to be pop-free - inverting/noninverting, which R
to change. Or another scheme? Any other hints?

You are always going to risk pops in this configuration. Anytime the
instantaenous voltage is different from the zero level, and you switch,
there will be a sudden level shift, giving a 'pop'. Best you can do is
to switch "slowly", which means not a relay, but something which
smoothly changes the resistor value over a few msec. A FET with the
gate suitably controlled would be better. Maybe look at some of the
eepot type devices. Some are designed to minimise audio zippering
effects, which amounts to the same thing. Alternately there are audio
opto-couplers using a LED and photo-resistor, which at least gives you
electrical isolation from the circuitry.

--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Note reply address is invalid, convert address above to machine form.

 

[Chris]

On 27/11/2014 1:00 AM, "Bruce Varley" wrote:

I need to switch the gain of an audio preamp using a reed relay, Kv =
3 (low), and 10 (high). It's important to avoid audible pops during
the switching. The device involved is a lo-noise audio opamp,
NE5534AP, using the relay to adjust one of the resistance values by
switching a second R in parallel. I'm wondering which configuration
would be most likely to be pop-free - inverting/noninverting, which R
to change. Or another scheme? Any other hints?

If you're game and can deal with the various crackpots there,
sci.electronics.design actually has some knowledgeable regulars that
could give some good advice...

--
Cheers,
chris.