Pro Audio Design Forum

mediatechnology wrote: I haven't thought about it but how does a bridged T increase the accuracy?

Resistor sets working in parallel, so you are trimming a preset standard ratio. No serial errors, or at least far less. I submit the existence of 1 and 0.1 dB step controls on the standard decade attenuator as evidence. That's my laymans assessment.

OK, got it. Absence of serial errors is the key IMHO.

I wished I had some of the old Kay Elemetrics attenuators...

I don't know if you can make this topology work in your application but you might consider some form of gain trim that is not a potentiometric configuration. The resistor errors add along the string. IIRC In the mic preamp gain switches I built, I was able to keep the errors <<0.1 dB with 1% resistors by using a single resistor for each step. With 0.1% parts you should be able to do much better.
I haven't thought about it but how does a bridged T increase the accuracy?

The problem as I see it with just switching a single shunt resistor is that I want the attenutor to behave as a smooth volume control. When the adjacent shunt resistors are in parallel during the switch the attenuation will increase significantly. The other thing I don't like about that style is the wildly varying impedance.

With a bridged T the impedance is constant and is dominated by the series and fixed shunt resistors. You switch two resistors per position so accuracy is very good.

True about the drop-out beween positions. In the mic preamp the gain drops to unity for a few ms between steps but it wasn't too audible. Your mastering situation however would require seemless switching. I think that was the motivation for bruno2000 (live) to use pots in his preamp build.

mediatechnology wrote: I think that was the motivation for bruno2000 (live) to use pots in his preamp build.

Oh yeah, no jumps allowed if they ever go in the remote truck.

If you want accuracy in the neighborhood of .1dB then I think a pot is only good for about +/- 3dB. I wanted +/- 12db but the largest switch Shallco makes is 46 position so +/- 11dB it is. It goes to 11!

I finished the prototype input and output amplifiers for the console with the Shallcos.





That's the M/S board in the monitor controller

The 3 position small rotary switch switches between passive stereo through and the M/S board for mono and difference. The toggle switches are mute L and mute R. So I can have mono sum, L mono and R mono. The push button source selectors control a pair of Broadcast Tools 16x2 relay switchers. I have access to 32 stereo sources. The bottom toggle switch switches between the two 16x2 switchers. It's a Preview/Program switch since everything is four channels.

Paul - Your build is awesome. You may hold the record for a rotary switch with the most poles.

Thanks Wayne. It's a full H pad with a 10k input impedance and a 5k output impedance. I thought it was a good compromise to throw away 6dB and get a lowish source impedance but have a bridging load.

BTW I sent you an email yesterday for some more m/s boards. Did you get it?

Excellent work Paul. I hope it lives up to your standards.