Perhaps an Explanation

[sr1200]

I seriously dont know what i could possibly put as a title for this post, so apologies for that.

Im digging into a piece of gear that I am trying to mod for a friend of mine and i ran into something kind of odd (well, odd to me since ive never run into it before). There is a switch on the unit that takes the hot and cold input signal and runs the hot through resistor A and cold through resistor B.  The switch has 4 settings, but only one of these settings does this where hot goes through say a 100K resistor and cold goes through a 110K resistor.  From there it seems the signal goes into a transformer and then to the rest of the circuit.  Im guessing that the other functions (where the resistors are matched) is just helping to load the transformer a little differently, but the one where they are mis-matched... whats the science behind that?

 

[Bo Deadly]

Sounds like it's just a pad that uses the load on the secondary as the shunting element.

Assuming that the resistors are really just in series with the signal lines, it doesn't matter that they're not matched. It's a two terminal device and therefore any current into one terminal must be balanced by an equal amount of current out of the other. There's simply no where else for the current to go. So you could literally have one 210K resistor and it would perform exactly the same. Although maybe not quite exactly the same because 100K/110K is enough that it might interact differently because of the capacitance of the transformer windings.

I have actually used a pad like this before but I have to wonder why you never see this configuration. It is MUCH more common to see a conventional U pad in front of the transformer. I'm not particularly fluent in magnetics but there are folks here who are. Maybe they know of a difference between using an explicit shunt resistor vs using the secondary load as the shunt resistance. I would be curious to know myself actually.

 

[abbey road d enfer]

Just to make sure I understand it all, a schematic would help.
If squarewave's guess is correct, I would say it's a solution that needs to be analyzed.
The main issue I see is that transformers prefer low source impedance.
In that case, the source Z is 210k. Not knowing the rest of the circuit, it's hard to speculate.
[SPECULATE] Let's say it creates 20dB attenuation; that would mean the reflected impedance at the primary is 23.3k.
Another option is to use a U-pad. Let's assume a 20k input resistance is acceptable. The pad woul be constituted of two 9.1k in series and a 2.4k in shunt. Then the source Z for the xfmr would be a little over 2k. That means a possible VLF distortion reduction of 40dB!
Indeed the practical benefit would be dependant on many other factors. [/SPECULATE]

Now the impedance reflected at the primary is not perfectly linear with frequency; how much depends on the construction and how the secondary is loaded. So teh actual frequency response may be somewhat compromised.

Another point to be taken is that some xfmrs are designed to be operated in current mode. Their secondary is usually loaded with a VE stage, so the voltage is very small and the flux nearly constant, which minimizes distortion.

 

[sr1200]

Interesting.  The funny part, is that this part (selection) of the circuit is supposed to be a "lofi" effect and if anything ADD distortion.  The unit is completely passive so padding down 40db is kinda crazy considering that you only lose about 6db at back side of the unit.

 

[JohnRoberts]

An unwritten truth about merchandising floobydust is that to claim your SKU is better, it needs to sound different from a flat, linear path.

Any audible difference can be sold as better, to a receptive audience.

JR

 

[abbey road d enfer]

Interesting.  The funny part, is that this part (selection) of the circuit is supposed to be a "lofi" effect and if anything ADD distortion.  The unit is completely passive so padding down 40db is kinda crazy considering that you only lose about 6db at back side of the unit.

Where does the 40dB attenuation come from?

 

[sr1200]

Where does the 40dB attenuation come from?

sorry 20db.  I mis-read the 40db as attenuation, was VLF dist. reduction.

@johnRoberts, yep... its a "FEATURE", up charge it!! lol

 

[Whoops]

Interesting.  The funny part, is that this part (selection) of the circuit is supposed to be a "lofi" effect and if anything ADD distortion.  The unit is completely passive so padding down 40db is kinda crazy considering that you only lose about 6db at back side of the unit.

And whats the deal with the secrecy about what unit that is?