[dgardner]

A few people have asked me about front panel gain select switches for the dynalo and dynahi amplifiers. It seems desirable to have at least two gain settings, one lower and one higher.

The question is: What is the smartest way to implement the switch in the circuit?

My thinking is that if you put a 2 pole, 3 position rotary type switch into the feedback loop, one needs to be careful. Is it better to control the feedback resistor from the output to the inverting input, or the resistor from the inverting input to ground?

Also, assuming you switch in various feedback resistor values to get different gains, would you expect that the compensation capacitor would need to be unique for each gain select position?

If anyone has had success or failure along these lines, please comment.

 

[bg4533]

On your Dynalo board what resistors determine gain? What is the gain with the stock resistors? It is way too high for me. I normally like a gain around 3.

 

[dgardner]

Gain = 1 + (Ra/Rb)
----
Dynalo Gain = 11.
Ra = R16, R46 = 10K
Rb = R7, R37 = 1K

----
Dynahi Gain = 13.
Ra = 1200 ohms typical
Rb = 100 ohms typical
----

So to get a dynalo gain of 3, set R16 and R46 to around 2K.

 

[phobus]

So would it be preferable to change Ra and not Rb, or does it matter?

 

[SnoopyRocks]

I would do it like this

 

[chillysalsa]

I did a similar thing to Snoopy in a Dynahi, it worked and sounded fine.

I personally like to aviod having switches in the signal path if at all possible. This scheme at least prevents the signal from 'seeing' the switch when it's open in one gain setting anyway...

 

[nikongod]

i personally like the method of using 2 resistors in series rather than a single resistor, and a switch in parallel to one of the resistors.

simple.

 

[dgardner]

So if you use the nikongod method, and you want to have gains of 3, 6, 9, and 12, you could make the feedback resistor (R2 below) out of four resistors in series and short them out sequentially with a rotary switch.

Code:

[left]R1R2GAINR2-AR2-BR2-CR2-DR2-TOTAL 31002003200000200 6100500620030000500 910080092003003000800 121001100122003003003001100[/left]

 

[Tomo]

Hi,

A word. ... The rotary switch should never "break." There should always be some feedback resistance when switching. This is because when there is no feedback you know you get the direct gain of the opamp which is pretty darn high. In essense, you might get a burst of signal which may be damaging. ...

Take care of your sweet cans.

Tomo

 

[nikongod]

if you want 4 gain levels, you can do it with 2 "extra" resistors, and 2 switches (per chanel obviously)

simply select the various resistors so that they are not similar. if you want to get fancy, calculate it out.

example: feedback resistor: 10k.
non shorted resistor: 15k
shorted 1: 7.5K
shorted 2: 15k

with both "shorted" resistors shorted, you ahev gain of 2.5 with only the 15k resistor shorted, you ahev gain of 3.3(mental math) with the 7.5k shorted, and hte 15k in line, you ahve gain of 4, and with all resistors in line, you ahev gain of 4.75 (mental math)

my descriptions are terrible, but draw it. im too tired.

 

[bg4533]

I prefer the method of resistors in parallel. You get 1 untainted setting with this method. I really don't see the need for more than 2 gain settings. The gain settings are too close together. I listened to the Headamp GS-1 on Saturday. It has a gain switch (2.3 and 7.5). Even a gain of 2.3 was enough for my HD650. I think those settings are well chosen and should be good for anything.

 

[phobus]

Great info

 

[skyskraper]

i reckon a simple toggle for 5 and 11 would be perfect for the dynalo

havent built a dynahi yet, but im sure once the basic procedure is worked out it would be easy to change the actual gain figure

 

[dgardner]

One reason to consider more than two gain settings is to have the option on a stereo amp board to support running it with one channel of a balanced input.

In single ended stereo mode, the unit might need only a high and low gain setting. Now assume that you feed in a balanced signal via an XLR connector. Since the output is driven differentially, the output signal will effectively have 2X the single ended voltage gain (I think). So, you might need to lower the gain even more in balanced mode.

I guess I'll just have to play with it for a while to see what settings make sense.

What would be nice would be to have the minimum number of gain settings that allow for a pair of RS1's and a pair of HD-650's to be happy in single ended mode *and* in balanced mode (dual amp units required).

 

[SnoopyRocks]

Here are two more options with 3 gain settings. I thought about it a little and decided that it's better to have the switch connect to the output than the the input. For a balanced configuration, simply substitute a 4pdt on/off/on switch for the dpdt on/off/on switch in the schematics.

Parallel

Series