attenuation problem

[chikkenguy]

steps 6-11 on a channel of my a12 are not working (no signal). ive been looking at the schematic, but it is more compliacated than i thought it was. are SW2a and SW2b ganged?

the other steps seem to work normally.

how does this attenuation scheme work?

im figuring its proably just a resistor or something, and ill check around with a dmm sometime soon, but if i knew how it worked, it would make it easier for me to trace.

[tpryan]

Check for continuity between pins 1-3 and 8-12 on the front deck of the rotary switch (SW2a on the schematic). The attenuator is a simple resistive divider across the secondary of the input transformer. Each successive resistor decreases the voltage 5dB. But it sounds like your problem isn't the attenuator.

[chikkenguy]

i dont get it. pins 4-8 on sw2a go through the resistor chain in what looks like a logical fashion to me, but pins 1-3 and 8-12 are all common and will get divided all by the same amount.

what purpose does sw2b serve in the circuit? it looks to me like volume control by using neg. feedback.

pins 1-3 and 8-12 all have continuity.

if you dont think the problem is the attenuator, what else could it be that would be dependant upon the position of sw2?

[tele_player]

The overall gain through the preamp is controlled by the combination of attenuation on the input (sw2a) and adjustment of the negative feedback on the opamp (sw2b and the trim control).

I'd check if pin one of the opamp has zero resistance to ground in any position of the rotary switch. That would definitely kill the sound.

[chikkenguy]

ok, well i found what it probably is. during soldering, i shorted pins 8 and 9 of sw2a. without really checking, i thought they were not supposed to be connected, so i cut out the solder bridge (as well as the pcb trace that i now see was there upon closer inspection).

i thought though, that i was missing 5 or 6 steps of signal, and resoldering pins 8 and 9 will only give me 4 steps (pins 9,10,11,12) it seems.

anyways, ill give it a try and see how it works.

i also am wondering what is the reason for using neg. feedback for some of the gain steps and then resistive attenuation for the rest?

[tele_player]

A tip - if you create a solder bridge, remove it with solder wick.

[tpryan]

The transformers give a combined voltage gain of about 24dB. If you want the overall gain to be a minimum of 15dB, you've got to use an attenuator somewhere. Either that or run the op-amp with a gain less than 1. This is possible with many op-amps, but the 2520 (and 2520 work-alikes) aren't guaranteed to be "unity gain stable" and need to be run at a higher gain.

[chikkenguy]

why then doesnt it use only resistive attenuation?

[tpryan]

Adjustable gain is the best way to accommodate the widely varying signal levels a microphone preamp encounters. If you ran the amplifier at a fixed gain, you'd have to fix it at its maximum. With a voltage feedback amplifier, maximum gain also implies minimum bandwidth and maximum distortion. It also means the maximum amount of self-noise would always be present at the output. If you attenuate at the input, you'll have a needlessly noisy, distorted, and bandwidth limited preamp. If you attenuate at the output, you'll reduce noise along with signal level, but your amp will clip very easily and have a needlessly high (and probably variable) output impedance to go along with the extra distortion and limited bandwidth. It's simply much more effective to be able to select the appropriate gain for any particular input signal, and only attenuate (at the input) when absolutely necessary.

[drunton]

How about this for an explanation (sure hope I have it all right)

If you look a little closer on the schematic, you will see that pins 1-3 and 9-12 all direct the input signal straight from the transformer to the opamp. The added connection on pin 8 (which took me a while to find ...) adds the divider network to the circuit.

In positions 1-2 and 9-12, R2-R6 are all in series to ground. They don't load down the transformer since the sum is quite high compared to the transformer secondary impedance.

On step 4, you switch some of the attenuator into the circuit. The signal goes to pin 8, through series resistors R3-R6 but R2 is now connected to ground, resulting in -20dB voltage divider type of attenuation.

On step 5, the signal goes through series resistors R4-R6 with R2 and R3 in series to ground and the attenuation is only -15dB.

The pattern continues until you hit position 8+ where there is no more attenuation.

The interesting thing is that Tim put switch positions 1-3 as the highest gain, so the stop pin is set for the lowest overall gain at switch position 4. I can only figure this made the layout easier (any comment Tim?)

Maybe this helps you understand it some - took me a while to figure it all out.

[chikkenguy]

It's simply much more effective to be able to select the appropriate gain for any particular input signal, and only attenuate (at the input) when absolutely necessary.

thanks for the explanation. just one more question. why chose to use negative feedback after the amp instead of another resistive attenuator?

[tele_player]

Because negative feedback greatly reduces harmonic distortion.

[tpryan]

thanks for the explanation. just one more question. why chose to use negative feedback after the amp instead of another resistive attenuator?

I thought I answered this above. The negative feedback doesn't come after the amp. In a fundamental way the negative feedback is the amp, and a full explanation of how and why this is so is beyond the scope of any answer you're going to get on a message board. One place to start is the IC Op-Amp Cookbook by Walt Jung. Have fun.