Shunt Attenuator Myth

Status
Not open for further replies.
Oh, I meant an attenuator using conventional resistors. I didn't mention S&B TVC which I would probably choose over any resisstive attenuator, but only if properly implemented.

I also have Placette as one of my references.

But I might give a try that Lightspeed in a future.

It's hard to discuss attenuators using different systems, what works in one may not neccessarily work for other. I learned about it already few times.
 
FWIW, the Lightspeed handily beat my Audio Consulting TVC. The latter added, by comparison, an edge to HFs, and a bit of smear, was to that degree less resolved, etc. I speculate that the switches in the AC unit + the lousy dielectric on those yards and yards of (albeit excellent silver) wire are probably to blame.
 
atarashi said:
well, one thing to consider (in favour of shunt att.) is, that the series resistance that the following circuit sees is almoust constant, so the hf rolloff point does not change that much.

a negative is, that the input imp. can change quite a lot...
so the stage in front of the att. has to cope with it.

No, both the input and output impedance can change quite a lot.

The following circuit doesn't see just the series resistor, it sees the parallel combination of the series and shunt resistor. Which can vary from zero ohms to the value of the series resistor.

se
 
Peter Daniel said:
I agree with Steve though, that the shunt resistor is equally important, if not more important than a series resistor. I didn't use Ohm law to figure that out, but my listening tests indicate just that.

Well, my post was purely on a technical point to address purely technical claims. What sounds good, bad, or worse to a particular individual is a whole other matter.

se
 
AndrewT said:
Steve,
you do realise that your formula is incomplete.

Yes. Intentionally so.

Source sees Rx + (Ry//Zin).
If Zin is > 10* Ry then the error is small, but as Ry increases the errors become very significant.

Right. I was just sticking to the basics and looking at the attenuator on its own under ideal conditions (i.e. zero source impedance and infinite load impedance).

se
 
AndrewT said:
it appears that most are agreed that AN has placed emphasis on the wrong (series) resistor.

Well, there seems to be this somewhat common notion that the "signal path" is a straight, series path from input to output.

However, there appears little consensus on whether both resistors are equally important or if the quality should be placed in the shunt position.

I don't see how you can separate the two. The current that flows through the series resistor must also flow through the shunt resistor. While the input doesn't see the voltage dropped across the series resistor, any deviation from the ideal of the series resistor must result in a deviation in the current flowing through it which will manifest itself in the I/R voltage drop across the shunt resistor.

So I guess the best you can say is that an ideal series resistor won't make a less than ideal shunt resistor any worse and vice versa.

se
 
My version of series shunt volume control

I would like to share with you on using a modify series shunt voulme control. I am very please with the result. this is what I do :-

There are two equal value ressistor in series with the signal path.I/P --> RX + switch selector
+ to --------------------> O/P
+ RY ----> Ground

The lower volume level switch position has a RYn connected to ground ( parrallel RY and RYn)

The Hioh volume level switch position has a RXn connected to I/P
( Parallel of RX and RXn)

You can work out the resistor value by using a excel spreadsheet
In this configuration, there are alway 3 resistors in the signal path except the switch is set to the mid point ( a selected postion with no connection). RX and RY use the highest quality resistor which you like. RXn's and RYn's can use next best quality resistor of your choice ( I use helco).


Parry
 
Status
Not open for further replies.