Salas hotrodded blue DCB1 build

[schubert]

Originally Posted by KatieandDad

It's normally consdiered bad practice to put a variable resistor (as opposed to a trim pot) in the feedback loop of an op-amp.

In the feedback position, this can result in quite noisy operation. (I'm talking about a volume type pot. that is constantly being adjusted.)

The normal practice is to attenuate at either the input or the output.

How about a sealed cermet trimpot? After all I'm most likely going to set these pots once and forget them.

[schubert]

Originally Posted by AndrewT

post1847
The left most diagram, inverting, is the one you should use.
R1 & R2 set the gain or attenuation.
With both equal to 10k the amplification factor is 1 or +0dB.

You can alter the gain by simply increasing R2. 30k will give ~+10dB.
or
You can increase attenuation by reducing R2. 3k will give ~-10dB.

That range for a trimmer should be adequate for any sources you have or plan to buy in the future.

You can replace the resistors with a pot. The wiper goes to the -IN. But you must fit some protection in case the wiper goes momentarily open circuit.
You also need to add a fixed resistor before the pot at the input and a second fixed resistor after the pot on the leg going to the output feedback tapping. These resistors limit the range of adjustment.

You can choose any range of adjustment that suits your gear.
+0dB to -20dB or +12dB to -12dB or whatever.

You must use a unity gain stable opamp, or a unity gain stable discrete opamp.
I suggest you use a socket and a cheap & good {NE5534an+compensation}, to get you started. Later you can decide to go swapping or go discrete. Or 5532 for two channels and without added compensation.

OK great - thanks Andrew T - I think I have enough info to start trying things once I get the boards stuffed.

[KatieandDad]

If you are going to set and forget then that is fine.

Just imagine an infinite gain amplifier (an ideal op-amp) with a dodgy connection (a variable resistor) in its feedback loop.

As you adjust the pot. the gain can vary from (say) 1 to (say) 1000000. The slightest crackle will be immediately amplified over 1 million times.

Personally I would use the pot to ascertain the fixed resistor that you need.

[schubert]

Originally Posted by KatieandDad

If you are going to set and forget then that is fine.

Just imagine an infinite gain amplifier (an ideal op-amp) with a dodgy connection (a variable resistor) in its feedback loop.

As you adjust the pot. the gain can vary from (say) 1 to (say) 1000000. The slightest crackle will be immediately amplified over 1 million times.

Personally I would use the pot to ascertain the fixed resistor that you need.

Ideally, yes - but I may have to use it in another room some day, in which case I'd have to reset it.

[syklab]

Just finish building the DCB1

Neg rail measure 10.06VDC
Voltage drop across the 10R is 1.909V

Pos rail measure 10.3 VDV
Voltage drop across the 10R is 2.028V

DC offset
Left 2.0mV
Right 1.6mV

AC input 16.52 and 16.54VAC

The relay kicks in 15 sec after power on. Next thing is to connect a 20k volume and start the listen test.

[KatieandDad]

Schubert,

I would still build fixed gain stages and vary the volume as opposed to the gain with trimpots at the inputs.

[syklab]

Picture of my DCB1 buit, no case yet.

[KatieandDad]

I'm surprised you need a heatsink on the LM7812 ????

[Salas]

Most sink it, there is enough drop over it with 2X15 Tx. Not essential, but helps reliability.

Nice build syklab, congrats.

[AndrewT]

If you are using 5V relays without current saver and a mesmerize then a sink may be necessary.

But a Hypnotize with a 12V relay should not require a sink on the regulator. If you add a current saver then both Hyp & Mes should not need a sink for 12V relays.

I have a Mes running day and night on 15+15Vac, with two 12V relays pulled in, a current saver active and the reg without sink, is running cooler than my finger, i.e. Tc<36°C