Stepped Pan Controls

[abbey road d enfer]

I won't check your calculations, but this looks correct to me.

 

[earthsled]

I understand. You have been immensely helpful!

Just to add some notes:
Here is a formula for Excel that will calculate 1% resistor values: http://electronicdesign.com/content.aspx?topic=excel-formula-calculates-standard-1-resistor-value I went back and recalculated my shunt values so that they were standard 1% resistors.

I'm noticing that the calculation with injection resistors adds a range of 0.424 to 0.007 dB of extra attenuation to the steps (compared to without them). This this is of little concern being that the sin/cos scale is only off by under half by a dB.

Preparing to order parts for a prototype!

 

[earthsled]

A couple questions about components...

I'm planning to use rotary switches that are shorting (make-before-break). Hopefully this will minimize pops and clicks while using the pan controls.

Also, thinking about the bus connections - does it matter to have outputs of the attenuators "daisy-chained" together connecting to the main L/R out? Or, would it be best to have the outputs be connected in a "star" wiring?

Many thanks!

 

[abbey road d enfer]

It doesn't make much differnce in terms of sound quality, except the "star" bussing may introduce excessive parasitic capacitance. Since it is also much simpler in terms of construction, I would definitely choose daisy-chain.

 

[earthsled]

Hi again.

I'm wondering if it makes a difference to split the mono inputs before or after the series resistances of the attenuators.
If I use the "B" setup, should I increase the series values to 2k?

Thanks for your help!

 

[abbey road d enfer]

I wouldn't recommand config B, because one side loads the other, twisting all your calculations!
Almost nothing to be gained (what's the cost and bulk of two resistors?) and OTOH, you may have a lot of difficulties adjusting the value of the shunt resistors if needed.

 

[earthsled]

Okay, thank you. Just wanted to double check.

 

[earthsled]

I'm still having trouble with figuring out the input and output impedance.

The input impedance must be at least equal to the min. recommended load impedance of the preceding equipment. 50 ohms is the actual source impedance, but min. rec. Z is probably 600-2000R. The min impedance presented by the attenuator is when the shunt resistor is zero; then the impedance is 1200R, which should be adequate for most equipment.

When you say 50 ohms is the actual source impedance, is there a calculation for this? How is the minimum recommended impedance derived from this? Then, how do I calculate the output impedance based the 20k mixing resistors and 24 channels?

Thank you!

 

[abbey road d enfer]

I'm still having trouble with figuring out the input and output impedance.

The input impedance must be at least equal to the min. recommended load impedance of the preceding equipment. 50 ohms is the actual source impedance, but min. rec. Z is probably 600-2000R. The min impedance presented by the attenuator is when the shunt resistor is zero; then the impedance is 1200R, which should be adequate for most equipment.

When you say 50 ohms is the actual source impedance, is there a calculation for this?

You mentioned this figure regarding the Digi002.

How is the minimum recommended impedance derived from this?

I don't derive it from the specs. It's based on the fact that Digi use standard audio AOP's. All audio AOP's can drive a 2kohms load. Some can drive less than that. A typical 5534-based floating-balanced stage can drive 600 ohms. In order to drive significantly lower loads, it takes  buffered outputs or DOA's, not very likely to happen on a box that's built for cost.

Then, how do I calculate the output impedance based the 20k mixing resistors and 24 channels?

The output impedance of the attenuators vary little compared to the 20k series resistors. Each branch has an output impedance that's a little over 40k, the combined output impedance is 40k/24.

 

[earthsled]

I'm sorry. As you can see, I was confused.

So, the input impedance of this unit is simply 2k, and the output impedance is 40k/24 (with 24 pans together). 2k seems low for an input but, you are suggesting that most op-amp outputs should be able to drive this impedance with no problem.

Would there be any reason to attempt a higher input impedance? Or, would this cause other issues?

I appreciate your help!

 

[abbey road d enfer]

In order to increase the input impedance, all the components should be increased by the same factor; the injection resistors would be significantly higher and the resultant source impedance presented to the make-up amp would be higher too, which would be detrimental to the noise performance. You may also start having problems with stray capacitance (frequency-response and crosstalk).

 

[earthsled]

Great, I'll stick with the original plan for series resistors and impedance. This seems to be a good solution.

I'm considering adding a switch to the stereo bus outputs that would allow for a summed-mono check. Could this be done with a simple DPST to connect/disconnect both poles of the left and right? Or perhaps resistors would need to be added to reduce the summed level?

Thank you.

 

[earthsled]

Hi again.

I wanted to pose a question about a mono-summing switch again. This seems like a useful feature to have. Is there a simple resistive circuit that can be used to combine the balanced left and right outputs?

Thanks!

 

[abbey road d enfer]

You can use a DPDT for paralleling the L & R busses.
A 3PDT would allow you to add a resistor that would load the resulting bus in order to reduce level. I would suggest a 10kLog potentiometer with a 1k series resistor; this combination would give ca. 6dB attenuation fully CCW, ca. 1dB fully CW and ca. 3dB at mid-rotation.

 

[earthsled]

Interesting solution!

If my primary use of the mono-summing switch is for combining phase-coherent signals, would a fixed value at 6 dB attenuation be appropriate?

I'm assuming most console manufacturers would use a fixed attenuation for a "mono check" switch. Do you happen to know if there is a standard in this regard?

Thank you.

 

[abbey road d enfer]

Depending on manufacturer, mono sums are at either 0dB, -3 dB, -6 dB or adjustable...

 

[earthsled]

Would you mind showing me the equation used to calculate the attenuation based on this load resistance?

Thank you for your help.

 

[deuc224]

no helpful data i can add, but i just want to say thank you to abby for contributing to this thread so much

 

[abbey road d enfer]

Would you mind showing me the equation used to calculate the attenuation based on this load resistance?

Thank you for your help.

Just out of my head...
Bus impedance 40k/24 = 1600R
Two in parallels 1200R. Put 1.6k in parallels, you lose 6dB. Put 1.2k x2 in parallels, i.e. 2.4k, you lose 3dB.

 

[earthsled]

Thank you for the resistance vales. I have to admit, I'm still a bit fuzzy on the math. Is there a formula for finding this?

Also, just to be sure I'm setting this up right, I have attached a schematic....