Hi out of curiosity I decided to try a passive volume control.
The reason was to see if the cd via a dac could drive my amps.
I was hoping for a greater sound stage.
I read as many threads etc i could find on the subject and understand there is a relationship between output and input impedances
Due to that fact and not fully understanding the concept I ordered the cheapest passive I could find thinking if it worked I could change the pot, connectors and rewire for greater clarity overall.
The cd player Rotel RCD02 is listed as 10Ohms output
The dac CA Dacmagic outputs <50ohms 2.1v
The Power amps 2x bridged mono Rotel RB03 input sensitivity is 33kohms 1.0v
The RCAs are Cable Talk Monitor 3 all quite short
The pot used (allegedly) is alps rk16 10k
Plugging in last night I definitely fall into the HF roll off camp so not one to give up just yet I am curious.
The set was able to drive volume with no problems, however there is definitely a drop in HF I understand this should be due to impedance matching.
Is there a way to modify the source, cd and/or dac output or the power amps input to improve the matching which would not detrimentally affect the system if it still is less than favourable?
I'm quite comfortable swapping out caps/resistors.
Could some diy low capacitance interconnects help?
Does the pot have any effect? e.g swapping to a genuine alps rk27 with point to point wiring and better RCA's
Or is this simply a its not going to work scenario?
Thanks in advance Ian
The reason was to see if the cd via a dac could drive my amps.
I was hoping for a greater sound stage.
I read as many threads etc i could find on the subject and understand there is a relationship between output and input impedances
Due to that fact and not fully understanding the concept I ordered the cheapest passive I could find thinking if it worked I could change the pot, connectors and rewire for greater clarity overall.
The cd player Rotel RCD02 is listed as 10Ohms output
The dac CA Dacmagic outputs <50ohms 2.1v
The Power amps 2x bridged mono Rotel RB03 input sensitivity is 33kohms 1.0v
The RCAs are Cable Talk Monitor 3 all quite short
The pot used (allegedly) is alps rk16 10k
Plugging in last night I definitely fall into the HF roll off camp so not one to give up just yet I am curious.
The set was able to drive volume with no problems, however there is definitely a drop in HF I understand this should be due to impedance matching.
Is there a way to modify the source, cd and/or dac output or the power amps input to improve the matching which would not detrimentally affect the system if it still is less than favourable?
I'm quite comfortable swapping out caps/resistors.
Could some diy low capacitance interconnects help?
Does the pot have any effect? e.g swapping to a genuine alps rk27 with point to point wiring and better RCA's
Or is this simply a its not going to work scenario?
Thanks in advance Ian
The question isn't impedance matching exactly. Its that you have introduced a low pass filter into the system. A passive volume control has an output resistance/impedance that typically varies with the volume setting. The RCA cables and power amp that follow it have capacitance. Between the resistance and capacitance you have created an RC low pass filter. It may also be that the HF loss varies with the attenuator setting.
You could try various things to lower capacitance after the attenuator, or maybe install a high input impedance unity gain buffer immediately after the attenuator. Simple FET buffers can work quite well for that and sound excellent.
You could try various things to lower capacitance after the attenuator, or maybe install a high input impedance unity gain buffer immediately after the attenuator. Simple FET buffers can work quite well for that and sound excellent.
Thanks Mark
I did notice at high volume, way beyond casual listening the roll off seemed to be less prominent.
I'll have a search for FET buffer
I did notice at high volume, way beyond casual listening the roll off seemed to be less prominent.
I'll have a search for FET buffer
The highest output impedance of a 10K ohm passive will occur when the pot is set to mid resistance point. It becomes the 5K ohm portion to ground in parallel with the 5K to source plus the source's output impedance. Since in your case the latter is tiny, 5K in parallel with 5K is close enough or 2.5K ohm maximum 'preamp' output impedance. Over 3,000 pF of interconnect + amplifier input input capacitance is required to cause roll off starting at 20 kHz, which seems very unlikely.
Mid-point is 6 dB below max volume, typically very loud. At normal listening levels the volume control's output impedance will be substantially lower. Though it's been ages since measuring it I recall a 10K pot's roll off became measurable in the 50 kHz plus range.
Mid-point is 6 dB below max volume, typically very loud. At normal listening levels the volume control's output impedance will be substantially lower. Though it's been ages since measuring it I recall a 10K pot's roll off became measurable in the 50 kHz plus range.
Depending on cables and their lengths sometimes funny things can happen that affect SQ. Could try putting a little zobel at the output of the dac, say, maybe 220R in series with .01uf, to ground for each channel. See if any change. Otherwise I would try a buffer or maybe different cables. May help to put the attenuator right at the power amp input rather than closer to the dac output. Ultimately, if the sound changes with attenuator setting in a way not accounted for by Fletcher-Munsen then we may not be modeling the system with sufficient completeness to fully understand the problem. In that case a little experimentation can sometimes help tease out a solution.
Thanks for the replies
So it appears there is no obvious reason for the roll off?
The interconnects are short approx 500-600mm therefore 1-1.2m each. The amp ones need to be sufficient to reach the amps on shelves below but I could buy cables at ~200mm for dac to pot.
I haven't opened the pot box as if its simply not going to work I may return it.
Measuring at RCA terminals in+ out- I get 46kohm min vol 92 max.
across +/- on input 92k regardless of position
across +/- on output 9.5ohm at min 92 max
Not sure if anything can be read into those measurements?
For the record I don't trust the claim of an alps rk16 pot as its a very cheap unit.
So it appears there is no obvious reason for the roll off?
The interconnects are short approx 500-600mm therefore 1-1.2m each. The amp ones need to be sufficient to reach the amps on shelves below but I could buy cables at ~200mm for dac to pot.
I haven't opened the pot box as if its simply not going to work I may return it.
Measuring at RCA terminals in+ out- I get 46kohm min vol 92 max.
across +/- on input 92k regardless of position
across +/- on output 9.5ohm at min 92 max
Not sure if anything can be read into those measurements?
For the record I don't trust the claim of an alps rk16 pot as its a very cheap unit.
Okay, its really more or less around a 100k pot inside, not 10k. A FET buffer should work okay with it, or you could try returning it and getting a 10k attenuator instead.
Thanks Mark
Would that be enough for an audible roll off?
Its potentially good news as I originally intended to swap the pot for an alps blue 10k.
At least Its worth a shot?
Would that be enough for an audible roll off?
Its potentially good news as I originally intended to swap the pot for an alps blue 10k.
At least Its worth a shot?
Well, in a case where a 10k pot would give a 50k low pass corner frequency, then a 100k pot, which is 10 times higher value, should give a 10 times lower corner frequency. So, yes, it's within reason that some audible HF loss could occur with the original pot.
Also, while perhaps not a big concern in this case, pots are known to have distortion mechanisms. The resistive ink is a bit nonlinear with voltage, and the wiper contact can also produce some distortion. Therefore good quality audio pots may sometimes be worth considering.
Hope that helps 🙂
Also, while perhaps not a big concern in this case, pots are known to have distortion mechanisms. The resistive ink is a bit nonlinear with voltage, and the wiper contact can also produce some distortion. Therefore good quality audio pots may sometimes be worth considering.
Hope that helps 🙂
Last edited:
I've used passive 10k pots for years and it works extremely well. Please try it.
The 100k pot is way too large in value. But use an excellent cable to the amplifiers.
Line stages are very poor for the most part.
After trying the 10k pot for a while, consider a metal film switched resistor volume control.
When recording direct to disc LPs, Sheffield Labs set the level pots in a rehearsal, and then
replaced each pot with a pair of metal film resistors. It made a big improvement in the sound.
A friend worked there.
The 100k pot is way too large in value. But use an excellent cable to the amplifiers.
Line stages are very poor for the most part.
After trying the 10k pot for a while, consider a metal film switched resistor volume control.
When recording direct to disc LPs, Sheffield Labs set the level pots in a rehearsal, and then
replaced each pot with a pair of metal film resistors. It made a big improvement in the sound.
A friend worked there.
Last edited:
Would agree that many line stages could be considered poor. Need not be the case though if we don't need gain.
The reason I asked you to check the impedance of the pot was exactly this: based on what you described and human ears you had clearly a steep low pass filter that couldn't be explained with impedance and cables... hence the first culprit being the pot.
Get a 10k pot and be done with it. Buy a decent one, at least RK27. Or a cheap one to try first if you like it. Note, as you seem to like soldering etc., you could also make a try building yourself an voltage divider of a fixed value... the ratio being the one you are currently using your cheap pot when doing most of your listening. Dead cheap, admittely only one position, but then you get a favour of what you can get. Plenty explanation on ladder attenuators etc. on the net. Seek for 10k output impedance
I hope this helps
Claude
Get a 10k pot and be done with it. Buy a decent one, at least RK27. Or a cheap one to try first if you like it. Note, as you seem to like soldering etc., you could also make a try building yourself an voltage divider of a fixed value... the ratio being the one you are currently using your cheap pot when doing most of your listening. Dead cheap, admittely only one position, but then you get a favour of what you can get. Plenty explanation on ladder attenuators etc. on the net. Seek for 10k output impedance
I hope this helps
Claude
Thanks all
I'll pop the case later and order a 27 over the weekend.
Not sure at present if its worth binning the pcb for point to point silver wire? I'm not fully convinced 60mm silver wire wrapped in teflon or different rca's are actually audible. However kinda like when I modded the crossovers, whilst your in there....
I'm not sure there is space in this little box for much more than an RK 27
Those attenuators with resistors everywhere looking like a 1930s aircraft engine do look very cool though
I'll pop the case later and order a 27 over the weekend.
Not sure at present if its worth binning the pcb for point to point silver wire? I'm not fully convinced 60mm silver wire wrapped in teflon or different rca's are actually audible. However kinda like when I modded the crossovers, whilst your in there....
I'm not sure there is space in this little box for much more than an RK 27
Those attenuators with resistors everywhere looking like a 1930s aircraft engine do look very cool though
After a while, for fun you can bypass that pot (leave it in the box), put in a couple of resistors
per channel, and see what happens. Just measure where the pot is usually set to pick the parts.
Keep their sum to be 10k per channel. That's a no cost way to see what a switched resistor control
would sound like.
per channel, and see what happens. Just measure where the pot is usually set to pick the parts.
Keep their sum to be 10k per channel. That's a no cost way to see what a switched resistor control
would sound like.
Rayman what would you consider "excellent" cable?
By metal film i'm guessing its something like this?
They do look cool, a bit eye-wateringly priced
Attachments
Look for expanded polyethylene dielectric cable, a very good type.
For example, Belden 1800F cable is excellent.
https://www.markertek.com/Attachments/Specifications/Belden/BL-1800F-500-Specifications.pdf
Those red resistors are PRP metal film (Precision Resistive Products), and are very good.
https://www.prpinc.com/pr9372-series
Dale CMF/RN series metal film are very good, but discontinued. Some may still have stock.
The company does have other series that are similar but more expensive.
The Takman REY series metal film resistors are also very good.
https://www.partsconnexion.com/takman-metal-film-rey-series-resistors.html
All these 1% metal film resistor types are less than $1 each singly. There is no reason to have a switched
volume control with a large number of steps. Only 12 steps of 2dB each are plenty for me, and I normally
use fewer than that. You do want a rotary switch of the highest quality, such as Elma or Grayhill.
Shunt 10k types are to be preferred, both fewer parts and switch contacts are needed.
For example, Belden 1800F cable is excellent.
https://www.markertek.com/Attachments/Specifications/Belden/BL-1800F-500-Specifications.pdf
Those red resistors are PRP metal film (Precision Resistive Products), and are very good.
https://www.prpinc.com/pr9372-series
Dale CMF/RN series metal film are very good, but discontinued. Some may still have stock.
The company does have other series that are similar but more expensive.
The Takman REY series metal film resistors are also very good.
https://www.partsconnexion.com/takman-metal-film-rey-series-resistors.html
All these 1% metal film resistor types are less than $1 each singly. There is no reason to have a switched
volume control with a large number of steps. Only 12 steps of 2dB each are plenty for me, and I normally
use fewer than that. You do want a rotary switch of the highest quality, such as Elma or Grayhill.
Shunt 10k types are to be preferred, both fewer parts and switch contacts are needed.
Last edited:
