Vishay specify the track linearity as "Independent linearity (Typical)" and for the pot series P11A2/P11S2 from which my pot came it is +-5% (see my previous post). This is different from the pot resistance tolerance.
Vishay term for the tracking accuracy is "Special Interlinearity - Interconformity", but they do not quote a typical value for this parameter, at least not for this pot series. For the so-called "First orders", there is an option to specify a +-2% value.
However, deciphering the concrete value from the part specification is virtually impossible. It appears that it is encoded in some way in the part No. but the pertinent code is not to be found in the DS. I had an issue with Mouser last year over another parameter from the same pot series, and turned to Vishay for the code info. They did not bother to reply.
Regards,
Braca
5% linearity error in each track implicates that a maximum possible difference of 10% or the equivalent of +/- 1dB in attenuation may occur between both channels. In most cases it will be closer to +/- 0.5dB, but even that seems to be unacceptable for a high end preamp.
A discrete controlled stereo attenuator has several advantages.
1) When equipped with 0.1% resistors, it will have a tracking error in the order of 0.01dB.
2) One can set the minimum and the maximum gain as required, thereby completely preventing a potential accident because of turning the pot beyond safe levels.
3) There is never the danger of a wiper coming loose, potentially ruining the speakers.
4) When using proper resistors, possible causes of distortion such as thermal modulation will be kept to the minimum.
In practice 2 dB steps for a volume control seems to be adequate, 1dB steps are definitely too small for normal use.
Control range for a Preamp should be somewhere between 32dB and 48dB, depending on the variety of used sources, meaning 16 to 24 steps of 2dB each.
Implement this with a linear R/2R ladder, and you will need 5 to 6 relays plus a linear to dB converter to control the relays.
Doing the same with a logarithmic network results in 4 to 5 relays.
When no remote controller is used to change volume level, a BCD encoder switch can be used in both cases.
Because of the lower complexity a logarithmic network is to be preferred.
It is not easy to find BCD switches with more than 16 steps, that is why I have made the implementation for a 32dB Volume control, adequate for most cases but not for all.
That is why I have added in my own preamp a fifth relay plus resistor over the feedback resistor of the output OpAmp for an extra step of 16dB.
To my opinion every preamp should actually have remote control, so in my case this remote control takes care of the extra 16dB step giving me a control range from -40dB to +6dB in real gain.
In practice I never exceed 0dB or -32dB when used with a high output balanced DAC and a low output Phono preamp. This leaves me 6dB headroom on both ends.
However, in case of controlling the volume directly on the preamp, a separate switch can be installed to control the fifth relay when 32dB is not sufficient.
As a last word, a volume controller can all but "improve" the sound reproduction, it can only influence the sound in a negative way by adding unwanted components.
Using the volume control in the feedback chain prevents many of the possible negative influences as BP has tried to explain in his excellent design.
By using a discrete volume control, a further step can be made in reducing unwanted components being added to the signal, hence potentially resulting into a better sound.
A discrete controlled stereo attenuator has several advantages.
1) When equipped with 0.1% resistors, it will have a tracking error in the order of 0.01dB.
2) One can set the minimum and the maximum gain as required, thereby completely preventing a potential accident because of turning the pot beyond safe levels.
3) There is never the danger of a wiper coming loose, potentially ruining the speakers.
4) When using proper resistors, possible causes of distortion such as thermal modulation will be kept to the minimum.
In practice 2 dB steps for a volume control seems to be adequate, 1dB steps are definitely too small for normal use.
Control range for a Preamp should be somewhere between 32dB and 48dB, depending on the variety of used sources, meaning 16 to 24 steps of 2dB each.
Implement this with a linear R/2R ladder, and you will need 5 to 6 relays plus a linear to dB converter to control the relays.
Doing the same with a logarithmic network results in 4 to 5 relays.
When no remote controller is used to change volume level, a BCD encoder switch can be used in both cases.
Because of the lower complexity a logarithmic network is to be preferred.
It is not easy to find BCD switches with more than 16 steps, that is why I have made the implementation for a 32dB Volume control, adequate for most cases but not for all.
That is why I have added in my own preamp a fifth relay plus resistor over the feedback resistor of the output OpAmp for an extra step of 16dB.
To my opinion every preamp should actually have remote control, so in my case this remote control takes care of the extra 16dB step giving me a control range from -40dB to +6dB in real gain.
In practice I never exceed 0dB or -32dB when used with a high output balanced DAC and a low output Phono preamp. This leaves me 6dB headroom on both ends.
However, in case of controlling the volume directly on the preamp, a separate switch can be installed to control the fifth relay when 32dB is not sufficient.
As a last word, a volume controller can all but "improve" the sound reproduction, it can only influence the sound in a negative way by adding unwanted components.
Using the volume control in the feedback chain prevents many of the possible negative influences as BP has tried to explain in his excellent design.
By using a discrete volume control, a further step can be made in reducing unwanted components being added to the signal, hence potentially resulting into a better sound.
Consider the trace of a pot as a nonlinear resistance track, but consistent in it’s nonlinearity from start to end.
Nonlinear in this context meaning that resistance varies with the voltage applied.
As long as this pot is placed the way that Bruno advises, this non linearity will be fully cancelled.
When placing resistors either in parallel or in series with the pot, the audio signal will be modulated by the signal resulting in distortion.
The increased complexity with an R/2R network is even higher than I posted above.
One will need at least 7 relays to have a range of 32dB with (+/-0.3dB) 2dB steps, and for 48dB you will need at least 10 relays !!
A logarithmic network is more accurate and far less complex.
Nonlinear in this context meaning that resistance varies with the voltage applied.
As long as this pot is placed the way that Bruno advises, this non linearity will be fully cancelled.
When placing resistors either in parallel or in series with the pot, the audio signal will be modulated by the signal resulting in distortion.
The increased complexity with an R/2R network is even higher than I posted above.
One will need at least 7 relays to have a range of 32dB with (+/-0.3dB) 2dB steps, and for 48dB you will need at least 10 relays !!
A logarithmic network is more accurate and far less complex.
Hi Hans,
Was looking at your BCD controller the other day with respect to a multichannel setup I am building to drive active speakers. As I am irrationally against using the miniDSP to control overall levels I need something to do overall gain on 7 channels with the ability to have a fixed offset between them to handle sensitivity differences. MiniDSP outputs 14dBu so unlikely to ever need gain will have them set to 0dB max, or possibly less. Have you found any issues with reducing R12 much below 4K? It's one of those 'shouldn't matter' cases but thought I would ask.
Was looking at your BCD controller the other day with respect to a multichannel setup I am building to drive active speakers. As I am irrationally against using the miniDSP to control overall levels I need something to do overall gain on 7 channels with the ability to have a fixed offset between them to handle sensitivity differences. MiniDSP outputs 14dBu so unlikely to ever need gain will have them set to 0dB max, or possibly less. Have you found any issues with reducing R12 much below 4K? It's one of those 'shouldn't matter' cases but thought I would ask.
TI states that the LM4562 "easily drives loads up to 600 Ohm", which seems to be confirmed when looking at the distortion figures.
So R12 can be safely reduced to a a lower value that suits you, up to a minimum of 600 Ohm.
I've got a couple of the Haksford, Essex Audio untis as "gamble purchases" on ebay (as in I didn't know what the were, nor did the seller). Sadly, I can't change the crossovers (filters are on an Eprom) nor do I know what any of the room correction programs do on that unit either (and Malcolm has no info either..).
All I know is the the digital crossover is set up for TAD units - think I'll have to sell it on to a TAD user.
Great quality internals though I must say.
Last edited:
@ Jan Didden + Hans Polak.
Hi guys, the group buy for these boards has taken on new momentum with a change of CEO Bertel has handed over the reins to tfboy.
There appears to be some discrepancy over the resistors to be used.
Apparently the PCB pad spacing is set for 1% SMD but is has been said that 0.1% MELF
resistors are essential.
Which did you guys use and is there a workaround.
Many thanks for your input.
http://www.diyaudio.com/forums/group-buys/279981-bruno-putzeys-balanced-preamp-group-buy-20.html
Hi guys, the group buy for these boards has taken on new momentum with a change of CEO Bertel has handed over the reins to tfboy.
There appears to be some discrepancy over the resistors to be used.
Apparently the PCB pad spacing is set for 1% SMD but is has been said that 0.1% MELF
resistors are essential.
Which did you guys use and is there a workaround.
Many thanks for your input.
http://www.diyaudio.com/forums/group-buys/279981-bruno-putzeys-balanced-preamp-group-buy-20.html
In addition, I'm interested in the change from TL072 to another LM4562. The original design uses the TL072 but the Excel BOM changed to the same LM4562 opamp used elsewhere with the comment that the TL072 has a too large DC offset and adds more of its own, which seems strange considering it was designed in the circuit to remove any incoming DC unless I've misunderstood something.
A DC servo requires low input offset voltage and low drift of input offset voltage.
Vos for the 4562 is ±0.1mV typical and ±0.7mV max
Vos/deltaT is 0.2uV/C typical and no max specified.
No Vos drift with time in the table.
These are very good. But a waste of an exceptional opamp, just for servo duty.
071/2 is the wrong device for servo duty.
Vos for the 4562 is ±0.1mV typical and ±0.7mV max
Vos/deltaT is 0.2uV/C typical and no max specified.
No Vos drift with time in the table.
These are very good. But a waste of an exceptional opamp, just for servo duty.
071/2 is the wrong device for servo duty.
I was thinking of replacing the volume pot by a digital potentiometer. The idea is to use an i2c controlled pot connnected to an arduino in order to be able to build a remote controlled preamp. What I want to do is a small breakboard for the digital pot that you can put in place of the original pot. This will be a way for me to learn some pcb design and some electronics (I'am good at soldering and following an already made project but I have yet to "create" something). I was thinking of using the max5487 pot, do you think it's will be a good choice for what I want to do ?
Why don't you do the same with the Stepped Attenuator by Hans Polak ....search the Forum for it.
He designed it specifically for this preamp and looks like the perfect choice.
You can control the atenuator with the arduino and do whatever format you wish for atenuation display.
Would be a nice arduino project.
I have seen Hans Polak attenuator and while it looks very good, it's a bit more complex (using relays, and a "lot" of components) than what I was thinking about. Also by using a dual digital pot I would be able to allow a balance setting and I will have more than 16 different steps of attenuation (I also want to do a something along : the faster you turn the button the faster the volume change). My question was more on whether this kind of digital pot will be suitable as a replacement of a classic pot.
Thanks a lot.
Thanks a lot.
For this quality preamp i would sugest you look at the MUSES 72320 volume control chip then ( at Mouser ).
And write an arduino sketch to do that.
If it's good for PASS LABS top of the line preamp.....