You are here

Balanced Volume Control Using Stereo POT - impedance calcs

Customising, building or repairing your own gear? Need help with acoustic treatment or soundproofing? Ask away…

Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Tue Jan 09, 2018 2:29 pm

Hello folks,

I have had some success building a couple of passive attenuators before for use between my interface and power amp and had pretty much got to grips with the impedance considerations and calculations.

I have since stumbled on this thread on DIYAudio forum.
http://www.diyaudio.com/forums/analog-line-level/161593-wiring-balanced-volume-pots-2.html

After reading it I realised that I can build one that runs on balanced connections without unbalancing them using a standard stereo pot or better a stepped attenuator.

I would like to have a go at this. The pot is wired as a single variable resistor, which, from what I can tell allows some bleed of the + and - lines which will allow a varying degree of cancellation and hence attenuation. A pair of matched resistors one on each + and - line are in series prior to the variable resistor that joins both lines.

The calculations for input impedance as quoted from the thread example are :

"The resistors set the input resistance/impedance seen by the Source.
The Source sees the Zin of the amplifier + the vol pot in parallel plus the resistors.
Caclulate the effective parallel resistance. Then add on the two resistors in series.
eg.
amp Zin = 100k
vol pot = 50k
effective parallel resistance = 33k3
Source see 5k+5k+33k3 = 43k3, when set to maximum volume.
gain = 33k3/43k3= 0.77 = -2.3dB

When you set the vol pot to minimum, it places a zero ohm across the amp input and now the Source sees 5k+5k+0k = 10k
gain = 0.0 = -infinitydB"



I have keyed these in to a spreadsheet which is now calculating the attenuation at various different POT positions with different series resistors and variable resistor values and also telling me the input impedance seen by the source looking into the passive Pot and the Amp.

My main question is, what is the output impedance from this circuit and how can this be calculated? Surely there is a limit to the size of variable resistor (POT) that can be used without causing an issue upstream looking into the power amp?

I have searched through several threads on similar subjects for balanced passive Pots wired in this manner and have seen a suggestion that the output impedance is simply the value of the variable resistor. If this is the case, is this in parallel with the output impedance of the source?

In my case I have the following :

RME Fireface 400 = 75ohms output impedance
Passive POT ?ohms
Hypex Ncore 47Kohm from + to ground and 47Kohm from - to ground


The spreadsheet is showing me that the POT behaviour is likely to be a little strange,
I'm guessing that is why this method is not implemented in a lot of passive monitor controllers which appear to either unbalance the signal internally or use a 4 gang POT. From what I understand, the inaccuracies on the + and - lines cause a reduction in CMR offered by balanced connections. Of course in real world applications I can quite imagine this makes no appreciable difference but I'm certainly interested in attempting to build something that is more accurate in this respect if I have the option. 4 gang POTs also seem a little harder to come by.


Sorry massive ramble but I'm just posting hoping to get a better understanding of the figures!
Thanks.
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Tue Jan 09, 2018 11:14 pm

I am not sure what you are trying to achieve Woodnut?

As I understand things you want to drive a power amplifier which has balanced inputs from an RME 400 which has balanced outputs and interpose a balanced volume control. Why?

If the operating level of the RME is too high for the amp, why not just use a simple fixed resistive attenuator and then use the control on the AI?

You should be aware that using 'consumer' grade stereo pots is fraught? They typically do not have very good tracking tolerances and this causes image shifts as the control is moved. Making a spreadsheet of the values is IMHO an exercise in futility since the track values have a 20% tolerance and as I say, the two sections can be dBs adrift. Linear dual pots are better with law bending resistors but not a lot. See Duggy Self's book on Small Signal Amplifiers.

If you are set on a passive controller then the Nano Patch products seem quite good but since you are at RME 'quality' consider the Drawmer CMC2 to which Hugh gave a glowing report in the November 2017 issue.

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Tue Jan 09, 2018 11:41 pm

Hi and thanks for the reply, I will check out the book thanks for the suggestion.

Yes I want to go balanced from the RME FF400 to my power amp, I had considered fixed attenuation and to continue using the digital volume control at a higher level, but I'm missing a physical volume control, i just find it a lot easier and more reassuring in use.

I would use a stepped switch with 21 or more steps so that the resistors could be matched, this would give much better accuracy versus a standard pot which as you say won't be accurate channel to channel.

I have so many bits and pieces left over from other projects such as wire, connectors, resistors and a few small project boxes, I think I would only need the stepped two way switch plus a few of the less common resistor values. So, for a very small outlay I could build something much better than I can get commercially for anything like the same price.

Certainly its a bit of work planning the resistor values but it seems worth it.
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Wed Jan 10, 2018 7:39 am

Self give the resistor values for 3 different dB steps and some invaluable design information.

I don't know what your time is worth to you but this is a LOT of work! When done do you have the equipment to check the steps?. (we all mess up!)

As to the source and sink impedances, 75R seems adequately low although to get the best noise performance the resistor ladder needs to be low in value*, <1k and the RME MIGHT be slightly compromised driving that resistance and an additional op amp stage might be required. What is the input Z of the amplifier? If around 100k it will not cause any significant error but then I presume you just want a high quality attenuator but not 'AP' dB accuracy?

*Else why go for stepped attenuator? Presumably you want state of art performance. Personally I would just save up for the Drawmer! I have considered building a Monitor Controller based around the Self/Baxandall active VC and using high quality op amps (LM4562) to unbalance the feed then balance the output. But who has the time?!

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Hugh Robjohns » Wed Jan 10, 2018 9:58 am

Just spotted this thread... but I'm about to head out for the day and this involves some lengthy replies... I'll pop back later tonight or tomorrow...

H
User avatar
Hugh Robjohns
Moderator
Posts: 20359
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Wed Jan 10, 2018 10:02 am

Hugh Robjohns wrote:Just spotted this thread... but I'm about to head out for the day and this involves some lengthy replies... I'll pop back later tonight or tomorrow...

H

Oh! Got me worried now. QUITE spoiled my day Hugh!

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Wed Jan 10, 2018 1:59 pm

Hugh Robjohns wrote:Just spotted this thread... but I'm about to head out for the day and this involves some lengthy replies... I'll pop back later tonight or tomorrow...

H

That sounds ominous! I hope I haven't opened a can of worms!
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Wed Jan 10, 2018 2:05 pm

woodnut wrote:
Hugh Robjohns wrote:Just spotted this thread... but I'm about to head out for the day and this involves some lengthy replies... I'll pop back later tonight or tomorrow...

H

That sounds ominous! I hope I haven't opened a can of worms!

Yeah! AND! Eee's taking at about a day to think about it!

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Wed Jan 10, 2018 2:23 pm

ef37a wrote:Self give the resistor values for 3 different dB steps and some invaluable design information.

I don't know what your time is worth to you but this is a LOT of work! When done do you have the equipment to check the steps?. (we all mess up!)

As to the source and sink impedances, 75R seems adequately low although to get the best noise performance the resistor ladder needs to be low in value*, <1k and the RME MIGHT be slightly compromised driving that resistance and an additional op amp stage might be required. What is the input Z of the amplifier? If around 100k it will not cause any significant error but then I presume you just want a high quality attenuator but not 'AP' dB accuracy?

*Else why go for stepped attenuator? Presumably you want state of art performance. Personally I would just save up for the Drawmer! I have considered building a Monitor Controller based around the Self/Baxandall active VC and using high quality op amps (LM4562) to unbalance the feed then balance the output. But who has the time?!

Dave.


The spreadsheet hasn't taken much time at all really, as for checking the resistors, yes certainly I would check as I go. I have made a standard stepped "shunt" POT before for a HiFi valve amp that had a failed pot on it before. It didn't take me too long but lots of checking with the multimeter as I went.

This would involve two fixed, closely matched resistors on the + and - balanced lines and just a stepped two gang switch which joins the + and - balanced lines, again with closely matched resistors. It doesn't feel that difficult at the moment aside from the numbers, but perhaps I will change my mind on that if i actually get onto building it!

The power amp is quoted as 47K from each pole to earth. I have had a bit of a play with the numbers and was aiming for >1K input impedance as seen by the RME.

Perhaps I'm missing something, but surely if I use closely matched resistors I will get good accuracy channel to channel?
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Wed Jan 10, 2018 9:03 pm

If the method for calculating input impedance is correct as per my first post taken from another forum, and I have implemented these correctly, it seems that I could usea pair of 3k series resistors and a maximum value bleed resistor of 10K. With the 10K switched in the attenuation is at its smallest and the control is at maximum volume. In my example maximum volume is actually just over -4db if I have calculated correctly.

The source sees about 15K at max volume (-4db) looking through the passive Pot into the amp
If the bleed resistor is switched down to approx 300 ohms for example for approx -26db
the impedance seen drops to 6.3K

I've had a go at calculating the output impedance from the pot, although this is the calculation I'm really unsure about. At its worst point which is maximum volume (as i understand it anyway) with the 10K bleed resistor switched in this looks like it would be 3.7K.

If I have actually got the above correct, then I assume it would just be a case of selecting 20 resistor values smaller than 10K to be switched in that would correspond to the relevant steps of attenuation moving down from max volume (-4db).
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Hugh Robjohns » Thu Jan 11, 2018 12:20 pm

Sorry to be late in on this. I was out and about all day yesterday and couldn't respond in the way I wanted to.

However, all I was really going to say was that you need to be very careful using passive balanced attenuators, especially if using long connecting cables. There is a greatly increased risk of capacitive losses leading to a variable HF response at different level settings, as well as significantly compromised CMRR, potentially resulting in problems with interference and hum pickup. The gain control law is also difficult to get right using standard pots and, for stereo systems you need at least a quad-ganged pot if not an octal-ganged pot (depending on the design of variable attenuator).

A rotary switch with carefully calculated resistor values is a much better idea, such as those from Goldpoint or DACT (others are available).

http://www.goldpt.com/index.html
http://www.goldpt.com/sa1x.html

http://www.dact.com/html/attenuators.html

As for the total resistance value, I'd stick with 10K Ohm switched attenuators in your situation, and definitely nothing higher than 25k.

H
User avatar
Hugh Robjohns
Moderator
Posts: 20359
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Thu Jan 11, 2018 1:19 pm

I had forgotten about CMRR Hugh!

Yes, the " three legged" attenuator (which I think is the OP's idea?) worsens it considerably.

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Thu Jan 11, 2018 1:57 pm

Hugh Robjohns wrote:Sorry to be late in on this. I was out and about all day yesterday and couldn't respond in the way I wanted to.

However, all I was really going to say was that you need to be very careful using passive balanced attenuators, especially if using long connecting cables. There is a greatly increased risk of capacitive losses leading to a variable HF response at different level settings, as well as significantly compromised CMRR, potentially resulting in problems with interference and hum pickup. The gain control law is also difficult to get right using standard pots and, for stereo systems you need at least a quad-ganged pot if not an octal-ganged pot (depending on the design of variable attenuator).

A rotary switch with carefully calculated resistor values is a much better idea, such as those from Goldpoint or DACT (others are available).

http://www.goldpt.com/index.html
http://www.goldpt.com/sa1x.html

http://www.dact.com/html/attenuators.html

As for the total resistance value, I'd stick with 10K Ohm switched attenuators in your situation, and definitely nothing higher than 25k.

H


Thanks hugh for the reply. Yes i was aware of needing to keep cable length to a minimum.

Do you think a more standard configuration is better than what I was proposing?

Its the link below "configuration B" minus the valves on the right and the transformer on the left of course.
http://www.raleighaudio.com/figure_17.htm

I can see from the spreadsheet that the pot behaviour might be a little strange if a switch wasn't used, but as I would be selecting the resistors for each step it seems that this would be a more accurate (channel to channel) balance than any pot could offer, apart from a stepped standard pot of course but this would then need to be 4 ganged greatly increasing parts count, cost and build time. In that example, only a 2 gang is required.
The bleed resistor which is across the two balanced poles would be switched between 21 values each at a predefined attenuation step with the maximum being 10K. The series resistors remain fixed.

I guess in short my question is, are there any disadvantages of that method versus a four gang solution?
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Folderol » Thu Jan 11, 2018 2:01 pm

+1 for switched resistor attenuators.

Apart from the greatly improved precision, and exact attenuator steps, it is also possible to make such a network represent a constant impedance to both source and load (quite hairy maths though). This means that any cable capacitance effects will be the same at all settings so can be more easily compensated.
User avatar
Folderol
Jedi Poster
Posts: 6261
Joined: Sat Nov 15, 2008 12:00 am
Location: Rochester, UK
Just because you know what I'm talking about, it doesn't mean I do!

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Hugh Robjohns » Thu Jan 11, 2018 2:25 pm

woodnut wrote:Do you think a more standard configuration is better than what I was proposing?

Most commercial passive balanced attenuators seem to use configuration shown as Type A in your diagram, probably because that presents a constant load impedance to the source, and a standard audio-taper switched attenuator design can be used in many different applications (balanced, unbalanced, active, passive, etc).

I can see from the spreadsheet that the pot behaviour might be a little strange if a switch wasn't used, but as I would be selecting the resistors for each step it seems that this would be a more accurate (channel to channel) balance than any pot could offer...

Yes, a rotary switch with calculated and matched resistors to tight tolerances offers the best results, with the least CMRR and channel-matching problems.

H
User avatar
Hugh Robjohns
Moderator
Posts: 20359
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Hugh Robjohns » Thu Jan 11, 2018 2:29 pm

However... while I can see the DIY constructional attraction of a passive volume control, I'd opt for an active one every time because the technical performance and usability will always be superior.

DACT offer very elegant solutions with compact buffer amps that can be wired straight onto the stepped attenuators in either balanced or unbalanced arrangements that offer superb performance, for example.

H
User avatar
Hugh Robjohns
Moderator
Posts: 20359
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Fri Jan 12, 2018 2:16 pm

Folderol wrote:+1 for switched resistor attenuators.

Apart from the greatly improved precision, and exact attenuator steps, it is also possible to make such a network represent a constant impedance to both source and load (quite hairy maths though). This means that any cable capacitance effects will be the same at all settings so can be more easily compensated.


Yes I was reading about "H pads" which I think might be the same thing.

Perhaps a little too complex for where I'm at though!
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby woodnut » Fri Jan 12, 2018 2:21 pm

Hugh Robjohns wrote:
woodnut wrote:Do you think a more standard configuration is better than what I was proposing?

Most commercial passive balanced attenuators seem to use configuration shown as Type A in your diagram, probably because that presents a constant load impedance to the source, and a standard audio-taper switched attenuator design can be used in many different applications (balanced, unbalanced, active, passive, etc).

I can see from the spreadsheet that the pot behaviour might be a little strange if a switch wasn't used, but as I would be selecting the resistors for each step it seems that this would be a more accurate (channel to channel) balance than any pot could offer...

Yes, a rotary switch with calculated and matched resistors to tight tolerances offers the best results, with the least CMRR and channel-matching problems.

H


Thanks for your advice, I shall think on this all.

Finally, any recommendations for low capacitance balanced cables?

I guess the usual suspects, Canare, Van Damme, Mogami, Sommer, Belden are fairly similar?

Most seem to quote 50pF per metre from what I have seen. It seems that the star quad arrangement increases the capacitance a little.

Some of the digital coaxial cables have slightly lower figures and look like they could be wired for this use since they have three poles.

I guess its splitting hairs really but I'm going to be making up some balanced runs of cabling so I have the choice.
woodnut
Regular
Posts: 184
Joined: Sat Mar 24, 2007 12:00 am

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby Hugh Robjohns » Fri Jan 12, 2018 2:42 pm

woodnut wrote:Finally, any recommendations for low capacitance balanced cables?

I generally use AES3 cable these days for static installations, and usually the van Damme Purple series -- but there are many suitable brands around, as you say. They work perfectly well for digital as well as analogue mic and line connections and tend to have lower capacitance than traditional analogue-only cables.

It seems that the star quad arrangement increases the capacitance a little.

Usually considerably more, in fact, but it would also be a wasted expense in our case as the only benefit of star-quad is improved rejection of very close interference sources and (a) you're unlikely to have any of those in your application and (b) it only works if you have extremely good CMRR in the receiver which is difficult to achieve with a passive balanced volume control -- even a switched resistor type.

H
User avatar
Hugh Robjohns
Moderator
Posts: 20359
Joined: Thu Jul 24, 2003 11:00 pm
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Balanced Volume Control Using Stereo POT - impedance calcs

Postby ef37a » Fri Jan 12, 2018 2:54 pm

Personally I would buy the best 2 gang log pot I could find and then use high grade op amps, LM4562 say around it or some of the audiophile amps Hugh mentioned.

Dave.
ef37a
Jedi Poster
Posts: 8531
Joined: Sun May 28, 2006 11:00 pm
Location: northampton uk

Next

Who is online

Users browsing this forum: No registered users