An advantage of having a master analogue volume control for all 6 channels leaves you the option to change the output levels of the channels to suit the sensitivity of the speaker drivers.
There are two lcations where you can change the gain.
a.) at the diff amplifier, where the existing is 1k000:1k000 change to 2k000:1k000 for an extra -6dB of gain, or
b.) at the output inverting amp, where the existing is 2k00:4k00
The 4k could even be a 1k5+ 5kVR to give a range of gain from -2.5dB to +10.5dB
That gives a preset adjustment of -8.5dB to +10.5dB by changing a few (two per channel) resistors. The VR can then be replaced by a fixed parallel pair to give matching to the required gain to better than 1%.
Seems you guys are talking about my controller for the Behringer DCX2496.
It has 6 output channels and all have their own channel in a CS3318. The firmware is such that you can:
1 - get all channels up or down - vol mode;
2 - get all odd channels up and even channels down - balance mode (and other way around for other direction balance);
3 - get any pair of channels up or down - set channel per way (low, mid, high) to adjust for amp+ speaker sensitivity differences;
4 - keep the differences between channels set at (3) for all level and balance changes;
5 - store/recall up to three sets of settings for up to three installations.
Everything in 0.5dB steps.
It seems to me that the Maya (which has features my unit has not) could have a firmware revision to make it work like this. It is not a trivial change - it is not rocket science but will take some time to code it.
Really a different unit, but you wouldn't have to start from scratch. And it doesn't really make a big difference whether you control something like a CS3318 or a set of relays, the concept is the same, only difference is the decoding and driving in hardware.
Sorry to ramble on like that, but it is a nice area to discuss. 🙂
Jan
It has 6 output channels and all have their own channel in a CS3318. The firmware is such that you can:
1 - get all channels up or down - vol mode;
2 - get all odd channels up and even channels down - balance mode (and other way around for other direction balance);
3 - get any pair of channels up or down - set channel per way (low, mid, high) to adjust for amp+ speaker sensitivity differences;
4 - keep the differences between channels set at (3) for all level and balance changes;
5 - store/recall up to three sets of settings for up to three installations.
Everything in 0.5dB steps.
It seems to me that the Maya (which has features my unit has not) could have a firmware revision to make it work like this. It is not a trivial change - it is not rocket science but will take some time to code it.
Really a different unit, but you wouldn't have to start from scratch. And it doesn't really make a big difference whether you control something like a CS3318 or a set of relays, the concept is the same, only difference is the decoding and driving in hardware.
Sorry to ramble on like that, but it is a nice area to discuss. 🙂
Jan
Thanks for that. There was confusion regarding 2) due to the reply by the Maya guy (sorry, can't remember his handle) - seems we were both talking about different things. I asked about different levels for different inputs, but it seems he replied talking about different levels for different chanels (L + R) which he said would require twice as many relays. Hence the confusion.
For 1), independent variable outputs, what happens to the state of the relays when an input is de-selected? Do the relays not stay in the same position and therefore volume level as was previously used or do they close to a defult state e.g. minimum level?
When you leave the two PCF chips, you can connect 6 switches to operate the 6 relays.
Possible but not very practical.
Hans
You can interface the PCF chips with a micro controller over I2c but that will require you to implement your own logic in the uC.
After having considered several options with Tibi, the producer of MAYA, we have come to the following options where the BPBP can be equipped with a remote control.
1) A 2 channel Stereo setup.
For this setup, you will need one main MAYA board, one BPBP and one “Hans” board.
Further you will need a +/- 12 V to +/- 15 V 1A trafo for the BPBP and a +6 V to 8V 2A trafo for Maya.
2) A multi-channel LS system, driven from DAC’s.
For this setup, the same number of BPBP’s and “Hans” boards are needed as the amount of channels, without restriction. So for a 5 channel system you will need 5 of each.
This conglomerate will be driven by one main MAYA board. Control lines from MAYA have to be daisy chained to all boards. And of course you will also need the trafo’s as under 1).
Volume of the whole lot will be set all at the same time with the remote control, giving all “Hans” boards the same volume setting. *
3) A multi-channel surround system.
This setup has exactly the same requirement as under 2).
* In case of 2) and) 3) you may need different volumes to be set between channels, but only once.
Either the hardware sourcing the BPBP’s may have some facility to make these adjustments, or the needed difference in volume between channels has to be set with the BPBP.
In the case of the BPBP, this can be done by adjusting R14/R34 to the correct value.
Which value this should be, can be determined by temporarily soldering a 50K logarithmic pot in parallel to R14/R34, only for those channels that are too loud.
Once set, measure their value and replace them by fixed SMD resistors, to be soldered on top of R14/R34.
For those wanting to have a fixed output for recording purposes playing from the same source as the variable output, best thing to do is to make a break out at U2A/Net-Tie J4 and U7A/Net-Tie J10.
These signals are already available on the “Hans” board as P1/Gnd1 and P4/Gnd4 and can be directly taken from here. 50 Ohm resistors have to be placed in these lines, hot and cold, going to a balanced output to pins 2 and 3 of the XLR plug. Pin 1 must be connected to chassis ground.
Hans
1) A 2 channel Stereo setup.
For this setup, you will need one main MAYA board, one BPBP and one “Hans” board.
Further you will need a +/- 12 V to +/- 15 V 1A trafo for the BPBP and a +6 V to 8V 2A trafo for Maya.
2) A multi-channel LS system, driven from DAC’s.
For this setup, the same number of BPBP’s and “Hans” boards are needed as the amount of channels, without restriction. So for a 5 channel system you will need 5 of each.
This conglomerate will be driven by one main MAYA board. Control lines from MAYA have to be daisy chained to all boards. And of course you will also need the trafo’s as under 1).
Volume of the whole lot will be set all at the same time with the remote control, giving all “Hans” boards the same volume setting. *
3) A multi-channel surround system.
This setup has exactly the same requirement as under 2).
* In case of 2) and) 3) you may need different volumes to be set between channels, but only once.
Either the hardware sourcing the BPBP’s may have some facility to make these adjustments, or the needed difference in volume between channels has to be set with the BPBP.
In the case of the BPBP, this can be done by adjusting R14/R34 to the correct value.
Which value this should be, can be determined by temporarily soldering a 50K logarithmic pot in parallel to R14/R34, only for those channels that are too loud.
Once set, measure their value and replace them by fixed SMD resistors, to be soldered on top of R14/R34.
For those wanting to have a fixed output for recording purposes playing from the same source as the variable output, best thing to do is to make a break out at U2A/Net-Tie J4 and U7A/Net-Tie J10.
These signals are already available on the “Hans” board as P1/Gnd1 and P4/Gnd4 and can be directly taken from here. 50 Ohm resistors have to be placed in these lines, hot and cold, going to a balanced output to pins 2 and 3 of the XLR plug. Pin 1 must be connected to chassis ground.
Hans
Last edited:
Do I have a problem?
The output offset is 0.0mVdc on one channel and 0.1mVdc on the other channel when the vol pot is set to minimum (max attenuation).
The offset stays low till about 50% rotation where gain/attenuation is +0dB
But then increases slowly till about 95% rotation, when it jumps to 10.8Vdc of output offset.
Same for both channels. and does not alter if a dummy zero ohms is plugged into an input.
Is this an assembly error?
What could I have done wrong?
Or is it attributable to the inverting opamp seeing a virtual zero ohms at it's input and maginfiying any virtual ground error by the gain of near +infinity dB?
The output offset is 0.0mVdc on one channel and 0.1mVdc on the other channel when the vol pot is set to minimum (max attenuation).
The offset stays low till about 50% rotation where gain/attenuation is +0dB
But then increases slowly till about 95% rotation, when it jumps to 10.8Vdc of output offset.
Same for both channels. and does not alter if a dummy zero ohms is plugged into an input.
Is this an assembly error?
What could I have done wrong?
Or is it attributable to the inverting opamp seeing a virtual zero ohms at it's input and maginfiying any virtual ground error by the gain of near +infinity dB?
I'm referring to the picture, post #408: Are you using wires approved for 230 VAC for connecting the QRV09 headphone amp? I think the wires look a bit weak when it comes to insulation. The low voltage wires to Putzeys pcb is also very near the 230 VAC wires. You should have them well separated both from safety reasons and disturbance, EMC.
Last edited:
Andrew, two suggestions:
- that you don't see the offset change until above 0dB may have to do with the sensitivity of your multimeter;
- with open input, there may be an offset created because of the opamp input bias current.
Jan
Thanks JanD
the DMM shows a tiny offset growing from 0.0 or 0.1mVdc to ~1mVdc as the vol pot is turned in little increments.
Then it grows a bit faster to ~40mVdc of offset. Both are not quite identical, but similar.
Then suddenly both jump to 10.9Vdc
Changing from open input to shorted input does not change the offsets. I think this indicates a good choice of those 47k resistors at the input buffers.
I have now done some gain measurements.
With a 43mVac input from my sig gen I see a gain of ~+45dB to output ~22Vpp and that is where that signal would be clipped, if I turn it any higher.
My linear vol pot must be rubbish. Two intermediate attenuations checked. Channel balance @ ~0dB gives a good 0.15dB unbalance.
@ -10dB the imbalance is 0.51dB That is poor.
I will be adopting a stepped resistor vol pot copying H.Polak attenuation.
I just happen to have a cheap chinese 33step switch and have calculated the resistors to get ~1.5dB/step from +6dB to -45dB, or with one resistor change, ranging from +0dB to -51dB (I think I will do the second PCB with this resistor change to limit my maximum gain to +0dB).
the DMM shows a tiny offset growing from 0.0 or 0.1mVdc to ~1mVdc as the vol pot is turned in little increments.
Then it grows a bit faster to ~40mVdc of offset. Both are not quite identical, but similar.
Then suddenly both jump to 10.9Vdc
Changing from open input to shorted input does not change the offsets. I think this indicates a good choice of those 47k resistors at the input buffers.
I have now done some gain measurements.
With a 43mVac input from my sig gen I see a gain of ~+45dB to output ~22Vpp and that is where that signal would be clipped, if I turn it any higher.
My linear vol pot must be rubbish. Two intermediate attenuations checked. Channel balance @ ~0dB gives a good 0.15dB unbalance.
@ -10dB the imbalance is 0.51dB That is poor.
I will be adopting a stepped resistor vol pot copying H.Polak attenuation.
I just happen to have a cheap chinese 33step switch and have calculated the resistors to get ~1.5dB/step from +6dB to -45dB, or with one resistor change, ranging from +0dB to -51dB (I think I will do the second PCB with this resistor change to limit my maximum gain to +0dB).
Last edited:
Hi,
Anyone can update me with the latest BoM? Tried to find it but it got me to some broken link... Can it be added to first post of this thread?
Anyone interested in doing Hans Polak volume attenuator in a GB?
Ciao!
Do
Anyone can update me with the latest BoM? Tried to find it but it got me to some broken link... Can it be added to first post of this thread?
Anyone interested in doing Hans Polak volume attenuator in a GB?
Ciao!
Do
Thank you for your observations and recommendations, and also for the very fine design of the QRV-09 headphone amp.
The photo is a rather old one and shows the first version of the set. The mains cables have already been exchanged along with a few other changes in the PSU section.
The 12V wires are still the same as on the photo - they are for the most part on the bottom plate of the casing, and 10mm away from the QRV PCB.
In any case, the EMI is apparently very low, being inaudible in the loudspeakers after being amplified by 29dB in the power amp and with the BPBP preamp at full gain (+16dB in my case).
Regards,
Braca
There is already a GB for Hans' PCB!
It's not a lone GB though, it is as an add-on for the Maya controller GB but the Hans attenuator PCB can be ordered as seperate item.
See here: http://www.diyaudio.com/forums/grou...advanced-volume-controller-2.html#post4696177
A fully built board GB might be good though... but a bit of soldering SMD never hurt I guess - hones them skills and steady hands.
I do like your board, and TBH if I had more project funds would have bought the kit some time ago. I2C is a boon for the hobbyist with all the chips out there that support it making a lot of things a lot easier to do these days. I must de-rust my meagre programming abilities 🙂
How many people have built their boards from the BPPBP Group Buy and what are their opinions of it?
post453 gives a short report.
I'll add to that. Using a very cheap linear 10k vol pot, the standard set up is nearly unusable in my system. Only the first 70°, or so, of rotation before the gain becomes too high compared to my DCB1 where I can use the first 100° for similar attenuations. But the Putzeys' has a lot more noise, both hiss and hum. That was measurable on the test bench and is noticable when connected to the audio system.
I really need that stepped attenuator version to see how a reduced gain version behaves. But I've been away for a week and done nothing except cruise in the sun.
I'll also experiment with a remote mains transformer to see if that has any effect.
I'll add to that. Using a very cheap linear 10k vol pot, the standard set up is nearly unusable in my system. Only the first 70°, or so, of rotation before the gain becomes too high compared to my DCB1 where I can use the first 100° for similar attenuations. But the Putzeys' has a lot more noise, both hiss and hum. That was measurable on the test bench and is noticable when connected to the audio system.
I really need that stepped attenuator version to see how a reduced gain version behaves. But I've been away for a week and done nothing except cruise in the sun.
I'll also experiment with a remote mains transformer to see if that has any effect.
Last edited:
A cheap pot is not the ultimate device, but it will not very likely introduce noise, hiss or hum.
Replace the pot for a test with a set of fixed resistors, like 1K/1K to get 0dB gain.
With open input this setting should be dead silent.
Hans