How to determine the optimal potentiometer / stepped attenuator impedance?
Jun 1, 2015 at 5:39 PM Thread Starter Post #1 of 12

eduardokbb

New Head-Fier
Joined
Jun 1, 2015
Posts
2
Likes
0
Hello! This is my first post, and I'm really sorry if I'm posting that in the wrong place.
 
Well, my doubt is quite simple. Giving I have a DAC/preamp and a power amplifier (all of them solid state, no tubes) with the following specs:

DAC/Pre output impedance: 0,5Ohms
Power Amp input impedance: 40kOhms

How can I determine, exclusively for that application, the optimal impedance for a stepped attenuator?

Thanks for your help and sorry for bad english!

Peace,
Eduardo Barth.
 
Jun 4, 2015 at 1:02 AM Post #2 of 12

OJNeg

500+ Head-Fier
Joined
Aug 7, 2011
Posts
709
Likes
59
A lower value pot will tend to perform better, assuming the DAC/preamp is capable of driving it. Given that your DAC has a very low Zout and is solid-state, you can be confident there. If the DAC's output is AC coupled (capacitor in series) then you need to make sure the pot value isn't too low or you'll lose bass response.
 
I suspect you should be able to put a 10k pot in with no issues.
 
Jun 12, 2015 at 4:24 PM Post #3 of 12

eduardokbb

New Head-Fier
Joined
Jun 1, 2015
Posts
2
Likes
0
Another question: Since I'm using a bit of DIY with my project, what would be better:

a) Using a stepped attenuator/pot to attenuate the line level signal.
b) Using a pot to adjust my amplifier power output according to my needs.

I'm asking that because the second option would make no changes to the original signal and is something that I don't really see very often as a way to adjust an equipment volume level. Does this makes sense that this way would avoid harmonic distortion, giving I can keep both channels with the same volume or perhaps it makes sense but is way more difficult to keep both channels with the same output?

Thanks!
 
Jun 12, 2015 at 8:53 PM Post #4 of 12

OJNeg

500+ Head-Fier
Joined
Aug 7, 2011
Posts
709
Likes
59
Pot on the output is a bad idea and you'll almost never see it used. You'll not only waste power but also screw up output loading. Keep the attenuator at the line stage
 
Jun 19, 2015 at 12:43 PM Post #5 of 12

tomb

Member of the Trade: Beezar.com
Joined
Mar 1, 2006
Posts
10,513
Likes
743
  Another question: Since I'm using a bit of DIY with my project, what would be better:

a) Using a stepped attenuator/pot to attenuate the line level signal.
b) Using a pot to adjust my amplifier power output according to my needs.

I'm asking that because the second option would make no changes to the original signal and is something that I don't really see very often as a way to adjust an equipment volume level. Does this makes sense that this way would avoid harmonic distortion, giving I can keep both channels with the same volume or perhaps it makes sense but is way more difficult to keep both channels with the same output?

Thanks!

What OJNeg said. If the pot is on the output of an amp, it makes no sense because you would have no control over the clipping point of the amplifier.  It could be into full clipping mode and all you'd be doing is attenuating the clipped and distorted output.
 
Plus, if you are talking about the pot ahead of the amplifer in (b), but after the signal stage, that has risks as well.  The O2 did it, but it's the only headphone amplifier I'm aware of that did.  Just about everyone in the DIY community recommended against it.
 
Typically, the power output amplifier section of an amplifier string is distinguished from the signal stage in that it only produces current to supply the load.  The real amplification takes place in the signal stage.  Placing the pot between the two is only possible in a more sophisticated amplifier that has two stages.  In the case of a receiver or separate speaker components, this would be the pre-amp and the amplifier.  Note that volume controls are always on the pre-amp unless the component is a combined, integrated amp or receiver.  In either case, the separate components are inside the same chassis in the receiver/integrated amp, but they're still there.
 
In the case of a sophisticated headphone amp (not a CMoy), there is typically a signal stage and an output buffer.  It is possible to place the pot between the two as the O2 did, but you run into the same issue as placing it on the output of the entire amp - your source could have an amplitude high enough to cause the signal stage to clip, but the pot will have no means of controlling it.  It will simply pass on an attenuated, clipped/distorted signal to the output buffer.
 
I believe NWAVGUY made extensive notes about source selection and determining whether outputs were too strong for the O2.  In any other headphone amp, that's controlled automatically by placing the pot at the input. 
 
Jun 20, 2015 at 3:16 PM Post #6 of 12

OJNeg

500+ Head-Fier
Joined
Aug 7, 2011
Posts
709
Likes
59
The whole pot after the input thing was Nwavguy just being a cheap wanker and not wanting to use a good pot. He wanted to use a cheap one so that any channel imbalance wouldn't be amplified by that first stage. You can read about a lot of people running into issues getting the input stage to clip because of it. But of course the O2 is the pinnacle of headphone amplifier engineering, right guys
rolleyes.gif

 
Jun 20, 2015 at 10:32 PM Post #7 of 12

tomb

Member of the Trade: Beezar.com
Joined
Mar 1, 2006
Posts
10,513
Likes
743
  The whole pot after the input thing was Nwavguy just being a cheap wanker and not wanting to use a good pot. He wanted to use a cheap one so that any channel imbalance wouldn't be amplified by that first stage. You can read about a lot of people running into issues getting the input stage to clip because of it. But of course the O2 is the pinnacle of headphone amplifier engineering, right guys
rolleyes.gif

 
This ^ 
biggrin.gif

 
Jan 13, 2016 at 11:51 PM Post #8 of 12

Soundsgoodtome

Headphoneus Supremus
Joined
Apr 19, 2013
Posts
6,883
Likes
1,273
Location
SEATTLE
Hello DIYers, bringing this thread out of the depths: I'm looking to put a preamp before my active speakers with line-in input. The impedance on the website states "Impedance: 10K ohm or higher" and the output impedance of my source is Output impedance:  32 ohms Output level:  3 Vrms however it's a tube buffered DAC (click here for DAC details).

Is the DAC being tube buffered matter here or will being all SS be the same? Choices I have are
10K, 20K, 50k, 100K, 170K, 250K, 500K, and 750K. How do I choose the right stepped attenuator value and what different will there be between say a 50K and 100K? The 100K will have more attenuation?

 
Jan 14, 2016 at 11:07 AM Post #9 of 12

tomb

Member of the Trade: Beezar.com
Joined
Mar 1, 2006
Posts
10,513
Likes
743
The impedance of WHAT says 10k or higher? Your post is confusing as to what's going on.

If you're talking about input impedance, your discussion about output impedance has no bearing.

Input impedance will combine with the output capacitors of a source to create an RC circuit. This results in a high-bandpass filter that will cut out the bass frequencies if the impedance of the volume pot is not high enough. It sounds like someone is quoting you a minimum input impedance to ensure that doesn't happen.
 
Jan 14, 2016 at 11:54 AM Post #11 of 12

tomb

Member of the Trade: Beezar.com
Joined
Mar 1, 2006
Posts
10,513
Likes
743
And I'm more confused. These stepped attenuator can be used anywhere a normal passive preamp can be used? If so, which rating (100k)?


Lower is better unless your source is using output coupling capacitors. Then the pot rating must rated sufficiently high to keep from cutting out bass frequencies. You also have to know the value of the output coupling capacitors of the source.

The same issue will present itself if your powered speakers have input coupling capacitors.

If you have to guess, 50K is often a happy medium. 100K is almost certain, but will have more noise. 10K is often not enough.
 

Users who are viewing this thread

Top