The four resistor pad is electrically the same as the 3 resistor one for differential signals. If you imagine a single Rs with either end sea-sawing (AKA differentially driven) in the centre of that Rs there is no voltage change, i.e. it is a virtual ground. So for differential inputs there is no electrical difference if the centre of the shunt resistance is grounded or not.
The reason the 3 resistor pad doesn't attenuate common mode (AKA common voltage at either end of Rs) is that Rs has no voltage divider; however if Rs is split into two parts with the centre grounded, the the common voltage on each leg is attenuated by Ri/(Rs÷2), the same attenutation as differential voltage. This is because the node at the junction of the two halves of Rs can't vary with the common voltage on either end because it is tied to ground.
I can not understand why you would want to be able to change the attenuation in front of the amplifiers. Once you have optimised the gain structure, what is the point of sub-optimising it?
It's taken 150 posts or so achieve an understanding/acceptance of recommendations made in the first post in response to your initial post, yet you still keep going off at a tangent. Your system has been >>20dB away from optimal, but not for the reasons you thought. Getting the basics right will give huge gains, before getting concerned about minutiae.
The reason the 3 resistor pad doesn't attenuate common mode (AKA common voltage at either end of Rs) is that Rs has no voltage divider; however if Rs is split into two parts with the centre grounded, the the common voltage on each leg is attenuated by Ri/(Rs÷2), the same attenutation as differential voltage. This is because the node at the junction of the two halves of Rs can't vary with the common voltage on either end because it is tied to ground.
I can not understand why you would want to be able to change the attenuation in front of the amplifiers. Once you have optimised the gain structure, what is the point of sub-optimising it?
It's taken 150 posts or so achieve an understanding/acceptance of recommendations made in the first post in response to your initial post, yet you still keep going off at a tangent. Your system has been >>20dB away from optimal, but not for the reasons you thought. Getting the basics right will give huge gains, before getting concerned about minutiae.
Hey John, sorry if it seems frustrating to you. Feel free to take a break any time. I appreciate your input, but I don't want to impose. The reason why I want to be able to set different attenuation values is because for most of my work I'll be listening at a low volume, at which the consistent noise is counter productive. However, for some things, I need to turn things up louder than a fixed pad would allow.
In relation to the 10B with 20dB pads its noise floor will still be below that of the amplifier, and the amplifier will clip before it does. The signal to noise ration cannot be improved by changing the attenuation here. You need to have variable gain after your greatest noise source - the one with the smallest dynamic range - (the Fiio) to keep the signal path gain structure optimal.
We've already established that the way to do this is a variable gain line amplifier in front of the 10B. If you change the attenuation after the 10B, you will no longer be getting the best THD+N even if the background noise in the speakers is lower.
We've already established that the way to do this is a variable gain line amplifier in front of the 10B. If you change the attenuation after the 10B, you will no longer be getting the best THD+N even if the background noise in the speakers is lower.
cheater said:
Member sgrossklass mentioned AES48 in post 154, end of the 1st line of the last paragraph. Sorry for the confusion -- was trying to be a little less formal. A couple of the links he posted were by his name. I was just so sure that you'd have hungrily gobbled up all that great info.
Cheers
Ah, right, sorry! I didn't know that was Stephan. It's funny - when I couldn't find a reference to AES48 I went to the print view and searched each page of the print view (in 50-post mode) for "AES48" and didn't find it... wonder how that happened, maybe I glanced over his post because I confused it with yours. Thanks for pointing it out!
Yeah, the Pin 1 problem. I fixed it by cutting Pin 1 on the cables near the amps. I already have a plan on asking Bryston if they have suggestions about modifying the amplifiers to fix that. They're very receptive - turns out they have a cozy little forum ran by the VP, John Tanner, and he got me in touch with one of the engineers.
Bryston Limited forum on Audio Circle
Last edited: