Well i have some very newbie questions. Their is no indication on the pcb for the polarity of 10 uf caps ? I assume the + are on the side of the attenuator ?
Also thx to the bom to indicate 100 uf caps instead of 10 (and far bigger caps also), i will have to order again
Anoter question. I use the hans bord, i need to connect the grd 1 -2 of the volcb to the "net tie 10 and 4". To be sure, these "net tie" are next the R5 and C4 inscription on the BPBP pcb ?
Also thx to the bom to indicate 100 uf caps instead of 10 (and far bigger caps also), i will have to order again
Anoter question. I use the hans bord, i need to connect the grd 1 -2 of the volcb to the "net tie 10 and 4". To be sure, these "net tie" are next the R5 and C4 inscription on the BPBP pcb ?
expand your PCB layout on screen until you can easily read the area you are interested in. Then screen print and copy that into a print programme.
Then you have aprint next to you as you start soldering.
The two Net tie points are VERY close to other components and have very small pad areas (lands). Do these solder connections early, even if only wires for later use.
Then you have aprint next to you as you start soldering.
The two Net tie points are VERY close to other components and have very small pad areas (lands). Do these solder connections early, even if only wires for later use.
I'm probably showing my lack of knowledge here, but I understood the choice for a linear pot over a logarithmic pot was only made for the track linearity between the tracks of the type of pot. This led me to believe that a switched attenuator can be used and that it being logarithmic or linear doesn't matter, as long as 10K.
Could anyone please educate me on why a linear switch is preferred over a logarithmic one? Thanks!
Rene
Could anyone please educate me on why a linear switch is preferred over a logarithmic one? Thanks!
Rene
The "linear" part of B.Putzeys' explanation went right over my head.
It was eventually explained to me that the current density through the two resistances on either side of the opamp's -IN node MUST be IDENTICAL, before I got it.
it's the way an inverting opamp works that results in a cancellation of the resistance distortion if that condition is met. Even for very bad vol pot types (carbon included).
But for that condition the vol pot MUST be a linear type.
OK, that covers the vol pot type for distortion cancellation.
If instead one adopts ultra low distortion resistors (meaning selecting those that have ultra low Tempco and low voltage coefficient) then one can substitute fixed value resistors where the current density through each does not need to be equal.
Now that the current density requirement has been dropped, it gives the Builder the option to use any fixed values of low tempco low Vcoef resistors to give the attenuations and gains that the Builder requires.
Most of us like an equal dB change in SPL for an equal vol pot rotation. One can calculate the required resistor values to get that dB vs rotation law. or any other law that the Builder wants/needs.
It was eventually explained to me that the current density through the two resistances on either side of the opamp's -IN node MUST be IDENTICAL, before I got it.
it's the way an inverting opamp works that results in a cancellation of the resistance distortion if that condition is met. Even for very bad vol pot types (carbon included).
But for that condition the vol pot MUST be a linear type.
OK, that covers the vol pot type for distortion cancellation.
If instead one adopts ultra low distortion resistors (meaning selecting those that have ultra low Tempco and low voltage coefficient) then one can substitute fixed value resistors where the current density through each does not need to be equal.
Now that the current density requirement has been dropped, it gives the Builder the option to use any fixed values of low tempco low Vcoef resistors to give the attenuations and gains that the Builder requires.
Most of us like an equal dB change in SPL for an equal vol pot rotation. One can calculate the required resistor values to get that dB vs rotation law. or any other law that the Builder wants/needs.
I took the plunge.
I ordered the 10kOhm input resistance version of that ebay L168-45 (128 steps ladder type) volume controller kit, which the seller offered to me as being available instead of the 50kOhm version.
Will see if that does what it is supposed to do... but it will take several weeks to get here.
I ordered the 10kOhm input resistance version of that ebay L168-45 (128 steps ladder type) volume controller kit, which the seller offered to me as being available instead of the 50kOhm version.
Will see if that does what it is supposed to do... but it will take several weeks to get here.
Last edited:
hfxrzw
As I understand it...
The resistance materials used to make a pot are non-linear - ie they distort. This is nothing to do with whether it's a "linear-law" or "logarithmic-law" pot, it's the resistive material itself.
The distortion non-linearity is dependent on the current density in the resistive track.
Because of the way the pot is arranged around the opamp, for any particular value of signal the current in the "input resistor"is the same as the current in the "feedback resistor". If a linear-law pot is used then fit any particular current, the current density is constant throughout the whole of the pot. If a log-law pot were used then imagine a printed track that is narrow (more resistive) at one end and wider (less resistive) - here, for any particular value of current there will be varying current density along the track
With a linear-law pot the distortion in the input resistance part of the pot is cancelled by the distortion in the feedback resistance part of the pot
So the linearity being referred to had nothing to do with comparative linearity between the two tracks
It took me quite some time to be clear on this too. That's why I keep referring people back to the original thread - all discussed in great detail and explained over there.
As I understand it...
The resistance materials used to make a pot are non-linear - ie they distort. This is nothing to do with whether it's a "linear-law" or "logarithmic-law" pot, it's the resistive material itself.
The distortion non-linearity is dependent on the current density in the resistive track.
Because of the way the pot is arranged around the opamp, for any particular value of signal the current in the "input resistor"is the same as the current in the "feedback resistor". If a linear-law pot is used then fit any particular current, the current density is constant throughout the whole of the pot. If a log-law pot were used then imagine a printed track that is narrow (more resistive) at one end and wider (less resistive) - here, for any particular value of current there will be varying current density along the track
With a linear-law pot the distortion in the input resistance part of the pot is cancelled by the distortion in the feedback resistance part of the pot
So the linearity being referred to had nothing to do with comparative linearity between the two tracks
It took me quite some time to be clear on this too. That's why I keep referring people back to the original thread - all discussed in great detail and explained over there.
And what about having a 5K linear pot, in serie with a couple of good 5K dale resistors ?
That would give less gain, and more flexibility for attenuation, right ?
JMK
You really should read the article, why this would not be a good idea. The whole concept of the way the volume is controlled is about this.
Jan
No, it breaks the distortion correction mechanism for the pot which is provided by the circuit. As you move the pot wiper, whilst the current density in each part of the pot is still the same, the distortion component produced by the input part of the pot + resistor is now out of proportion to the distortion component produced by the feedback part of the pot + resistor, so you no longer get the distortion cancellation that you do when the pot is on its own
(Correct me if I've misunderstood, Jan)
Following on from Jan's comment - I didn't fully understand this after reading the article as it wasn't immediately clear to me which aspect of "linearity" Bruno was talking about, but I did after fully reading the original thread where you've posted the same question. Series resistors etc have already been discussed there
I doubt that, unless you are using a rheostat
Read the original article. Read the main thread; possibly continue the conversation there
Edit:
To quote from the article
Last edited:
It's not the distortion that plays a role here, but how to use a pot that is not super expensive for tracking, and still have a good tracking so that the balance does not shift with level.
Using additional resistors left or right destroys the brilliance of it.
Bruno went to great lengths to describe it. I arranged for it to be published, and even got a company to sponsor free PCBs. Why don't you (general, not you ChrisPa) just read it.
You know, I really get tired of this lazyness. I think I'll call it a day with Linear Audio.
Jan
Last edited:
That's a picture of the kit version.
There is a seller, who delivers even from Germany for a somewhat elevated price compared to the other one. But no additional VAT and duty.
http://www.ebay.de/itm/HIFI-Remote-preamplifier-kit-remote-Audio-volume-control-4-way-inputs-dale-GE-/192129156303?hash=item2cbbca44cf
Same seller, another version / model, even lower price
http://www.ebay.de/itm/Hi-End-2-Channel-Relay-Volume-Control-Board-Stereo-Relais-Lautstarkeregler-GE/142311878512?_trksid=p5411.c100167.m2940&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D20140131123730%26meid%3D39ca59d65b004964b37aefa1f5214410%26pid%3D100167%26rk%3D2%26rkt%3D2%26sd%3D192129156303
Last edited:
I have read the article before my initial post. I read the exact section you quote. I don't think you understood what he wrote there. He *does* refer to distortion in the wiper contact. That is where nonlinear distortion arises.
You, however, are suggesting that the circuit model for the carbon/conductive plastic track element of a potentiometer is not a resistor. Show it please. Measurements for the L and C values would assist.
As for "read the main thread", internet chatter is a good way to get oneself confused.
Last edited:
Sorry to say, but you have misread the article.
It is not the wiper causing the distortion but the nonlinear properties of the track material.
In this design nonlinearity is cancelled because the track has the same non linear properties on both sides of the wiper.
One part A of the Linear Pot is divided by the other part B, like A*Z / B*Z = A / B where Z represents some nonlinear track resistance per distance.
No matter how non linear Z is, Z is simply removed from the equation.
Non linear is this context means that twice the voltage does not result in twice the current.
Hans
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.