HI guys, a quick question please:
Till now I have been using a stereo LDR volume pot using 4 LDR's (each channel has 1 LDR in series and one to ground).
I now would like to create a four input volume pot (for balanced operation), but I only have 4 LDR's.
What is the latest thinking on creating LDR volume pots with a resistor shunt to ground?
Thanks for advice.
Till now I have been using a stereo LDR volume pot using 4 LDR's (each channel has 1 LDR in series and one to ground).
I now would like to create a four input volume pot (for balanced operation), but I only have 4 LDR's.
What is the latest thinking on creating LDR volume pots with a resistor shunt to ground?
Thanks for advice.
Thanks for the schematics Maximus. I can make DS1802 and IR control in one microcontroller chip. I am waiting for DS1802 and MCP42100 chips, and as soon as they arrive, I will post something useful. Also I have to order the LDRs, I am looking for someone who can sell me matched pairs!
What OPAMPs did you use in the VCCS?
Last edited:
Here is the circuit I have been using with MCP42100.
I found this on the internet, but havn't been able to re-locate it since. All I know is that it's a current regulator by Mike Mladejovsky.
Using the attached spreadsheet (borrowed from the JT thread IIRC), I calculate the ADC values I need to read to acheive the required taper.
The arduino calibration program then goes through trying different permutations of series/shunt to get these readings.
The program is a work-in-progress and somewhat messy, but is a reasonable starting point.
Feel free to use/critique/improve
Mike
I found this on the internet, but havn't been able to re-locate it since. All I know is that it's a current regulator by Mike Mladejovsky.
Using the attached spreadsheet (borrowed from the JT thread IIRC), I calculate the ADC values I need to read to acheive the required taper.
The arduino calibration program then goes through trying different permutations of series/shunt to get these readings.
The program is a work-in-progress and somewhat messy, but is a reasonable starting point.
Feel free to use/critique/improve
Mike
Attachments
Sorry, but your friend told you a crock of ---- . If anything passives done right give the best sound staging and depth.
Nelson did the simple unity gain buffer for guys who needed lower a output impedance after the Lightspeed for driving power amps with low input impedances of less than<47kohms, and some of his amps which are 20kohms or lower can then be used with the Lightspeed also.
Cheers George
Here is the circuit I have been using with MCP42100.
I found this on the internet, but havn't been able to re-locate it since. All I know is that it's a current regulator by Mike Mladejovsky.
Using the attached spreadsheet (borrowed from the JT thread IIRC), I calculate the ADC values I need to read to acheive the required taper.
The arduino calibration program then goes through trying different permutations of series/shunt to get these readings.
The program is a work-in-progress and somewhat messy, but is a reasonable starting point.
Feel free to use/critique/improve
Mike
This is similar to this circuit:
Voltage Controlled Current Source - Current Servo
Look at my earlier post
This is what I was thinking about few days ago with PGA231X, but when I read about temperature drifts and accuracy as well as the starting point of 0.1 mA I thought I should ask for the VCCS and Maximus replied with his schematics. Try it and let us know what results you get, I use AVR too, if you need help on codes I am here. Also why do u want to read the ADC?
Last edited:
yes, that circuit is the same, I must have missed your earlier post.
The program I posted is the calibration one used to determine the best settings for a particular pair of opto-couplers.
The ADC is being used to determine the attenuation that a particular series/shunt setting on the digital-pot will give.
I put 5v across the opto-couplers and read the resultant voltage.
The final goal is to get a pair of arrays like this:
static const byte volume_series[127] = {46, 30, 22, 18, 14, 12, 11, 9, 8, 74, etc.. };
static const byte volume_shunt[127] = {1, 2, 1, 1, 2, 1, 1, 1, 1, 2, etc.. };
Then in the actual pre-amp, when setting a volume, you would use something like this (single channel):
current_volume++;
spi_out(cmd_pot_series, volume_series[current_volume]);
spi_out(cmd_pot_shunt, volume_shunt[current_volume]);
obviously, you can use smaller arrays if required.
Using the spreadsheet, I have calculated a series of ADC values I need to get to achieve the desired taper.
The spreadsheet (some columns hidden) shows the optimum attenuation (dB) and then the equivalent ADC value needed.
The Blue line on the graph shows the optimum, the red line the achieved and the yellow shows the deviation (not too good at low volumes).
Hope this explains better where I'm trying to get.
I'll let you know how I get on.....
Last edited:
the best way to volume control a balanced signal is to connect Hot to Cold with a variable resistor. A single LDR can do that.More importantly it does not destroy your balanced signal.
The output impedance of this would not be able to drive cables so it must either have a pair of buffers or be located right at/on the balanced receiver.
The slight disadvantage to using an LDR rather than a variable resistor is the limited range of volume adjustment. The minimum resistance of the LDR cannot approach zero ohms.
Thanks Andrew,
Would you prefer your scheme to a series resistor with LDT shunt to ground? Have you heard both?
I was running my stereo LDR (series shunt) setup into an AD815, but then tried passive and all I can say is WOW! The passive is so transparent!
I'd prefer to have the volume control range, so perhaps the series resistor/shunt LDr is the way to go?
Would you prefer your scheme to a series resistor with LDT shunt to ground? Have you heard both?
I was running my stereo LDR (series shunt) setup into an AD815, but then tried passive and all I can say is WOW! The passive is so transparent!
I'd prefer to have the volume control range, so perhaps the series resistor/shunt LDr is the way to go?
the two series resistors are necessary to allow the shunt resistor to operate as a volume control in a balanced signal.
The two series resistors must be matched to better than 0.1% to maintain the benefit of the balanced signal.
The two series resistors cannot be replaced with LDRs because this will destroy the balanced nature of the signal.
They MUST be exactly matching fixed or exactly matching switched resistors.
A signal relay introducing a parallel bypass resistor across a fixed resistor say 20k and [1k+relay] would allow the series resistor to be either 952r or 20k. That would reintroduce a reasonable range of attenuation for the LDR volume control. The 20k//1k would be the "loud" setting and the 20k alone would be the "quieter" setting.
If you do a pair of LDRs as shunts to ground then again the difference in the shunt to ground values will ruin the balanced nature of the signal. You would be better converting the balanced signal to unbalanced and then using a standard unbalanced lightspeed to control volume. If you really need balanced signal after that then convert the lightspeed output back to balanced for the last part of the audio route.
The two series resistors must be matched to better than 0.1% to maintain the benefit of the balanced signal.
The two series resistors cannot be replaced with LDRs because this will destroy the balanced nature of the signal.
They MUST be exactly matching fixed or exactly matching switched resistors.
A signal relay introducing a parallel bypass resistor across a fixed resistor say 20k and [1k+relay] would allow the series resistor to be either 952r or 20k. That would reintroduce a reasonable range of attenuation for the LDR volume control. The 20k//1k would be the "loud" setting and the 20k alone would be the "quieter" setting.
If you do a pair of LDRs as shunts to ground then again the difference in the shunt to ground values will ruin the balanced nature of the signal. You would be better converting the balanced signal to unbalanced and then using a standard unbalanced lightspeed to control volume. If you really need balanced signal after that then convert the lightspeed output back to balanced for the last part of the audio route.
Last edited:
I see, then no need to get matched LDRS. Nice idea thu
Thanx - I knew there was something important I had forgotten to say....
Called Maxim, they "promised" that they would ship this weekend and would be on my "dock" early next week.
Coming from the Philippines...
Thanx - I knew there was something important I had forgotten to say....
I think I'll order 4 from farnell on Saturday and when I receive 'em I'll build my own tables, this is much cheaper for me
I was thinking about an equalizer (EQ) using those, but no use as long as I use my PC as it has an EQ already, so pointless to build one too.
BTW, have u tried that current source schematic? What results did u get?
The schematic appears to be giving good results (I'm no expert tho').
I'm still getting slightly inconsistent results in the calibration program (probably due to the optocouplers and the accuracy of the ADC sampling) and am coming to the conclusion that perhaps I'm expecting too much in terms of absolute accuracy wrt acheiving a perfect attenuation curve.
Anyway, I have decided to push on and am doing my own home-made layout as per the attachment.
Attachments
In contrast to this recommendation, I built a balanced Lightspeed attenuator according to a schematic shown early in the thread. It consists of a series and shunt LDR for each phase of each channel, eight LDRs altogether. Uriah Dailey provided well matched LDRs and I would be surprised if they were worse-matched than the usual four-gang potentiometer. At any rate I have never had music sound so good, and I am not using second-rate equipment, nor am I hard-of-hearing. My CD player and phono stage are fed directly to the amplifiers without a buffer through about 6 meters of cable. I detect no channel imbalance, no noise penalty, and soundstaging is great. If I wanted precise matching of hot and cold phases then I would probably look first to replacing the amplifiers which use twin triodes to handle the hot and cold signals. And yes I'm aware that is a reason people prefer SE. What I'm saying is I don't think the balanced nature of the signal will be ruined by using well-matched LDRs.
My opinion is that Dr H would be better off to get more LDRs to match up to the ones he has, or get eight that have been matched already by someone else. That would be two groups of LDRs, four matched LDRs per group.
Hi Barmy,
you need to read more of the literature on why balanced does it's job well and what causes it to not do it's job well.
The matching of impedances both at the source and at the receiver are crucial to maintain the advantage that balanced offers. If you don't maintain that balanced advantage then a balanced feed is a waste of resources
you need to read more of the literature on why balanced does it's job well and what causes it to not do it's job well.
The matching of impedances both at the source and at the receiver are crucial to maintain the advantage that balanced offers. If you don't maintain that balanced advantage then a balanced feed is a waste of resources