The Lightspeed has an input output and ground just like your pot does. If you have room to house the whole circuit in your preamp then you can easily just plug the signal in the same way it was plugged to your pot. Like Andrew said there is a lot more circuit to it than just a pot but the circuit is rather small and is very simple. If you have 6 square inches of space you have enough room. I suppose you would need around an inch of height to that.
Uriah
Uriah
Banned
Joined 2002
Hendrik
Its simple and its a wonderful circuit to start learning . The other great thing about it is that the guys on this thread are very very helpful. Cant wait to give you a hand with it. You may even decide to remove your preamp and use only the Lightspeed. We will use pictures and you will need a basic DMM aka mutimeter to know what voltage and resistances you have.
So yes you need a soldering iron, some perfboard, 24AWG wire, solder, a DMM and the components for the job.
You will need a 5VDC power supply which we will work on first. Once you get the 5VDC working we can go further from there.
URiah
Its simple and its a wonderful circuit to start learning . The other great thing about it is that the guys on this thread are very very helpful. Cant wait to give you a hand with it. You may even decide to remove your preamp and use only the Lightspeed. We will use pictures and you will need a basic DMM aka mutimeter to know what voltage and resistances you have.
So yes you need a soldering iron, some perfboard, 24AWG wire, solder, a DMM and the components for the job.
You will need a 5VDC power supply which we will work on first. Once you get the 5VDC working we can go further from there.
URiah
Relationship between LDR & "ideal" potentiometer
Over the course of the last week or so, I have waded through the 3288 posts (so far) in this thread. Whew! Fascinating subject, and I would like to build an LDR volume control and play with the concept a little. My existing system is very average – M Audio 2496 sound card, gainclone amp, BBC Monitor speakers – but I’d love to see what this control can do for me, especially in conjunction with a new outboard DAC for my computer.
I find that after all of this reading I understand quite a bit about the LDR system -- the power requirement and the characteristics of the LDR, etc – but have little fundamental understanding of the basic characteristics and specifications of a best-performance, standalone volume control. I want to know what the ideal characteristics would be if there were no technical limitations, and then try to get as close as I can with the LDR.
Is there a section in diyaudo where I could go to read more about what the “ideal” volume control would look like? Or somewhere on the web where it is discussed? Any suggestions appreciated.
Specific questions I would like answered somehow:
1. Given the typical source output and amplifier input impedances of current systems, what is the ideal potentiometer value that would best cover the most source/amplifier combinations well? 50K? 100K? 250K? Some other value?
2. If the output impedance of the volume control is important to the quality of the sound and should ideally remain constant (I’m gathering that this is the case from the reading I’ve done here), can that be accomplished somehow in theory? I’m trying to understand the underlying principles here.
3. I’ve read the Elliot article on volume controls. It appears that commercially-available log (audio taper) controls are not log at all, just a combination of two linear tracks that roughly approximate the log taper. Is this really acceptable, or would a real log taper be noticeably more pleasant to deal with?
4. Between the recommended log taper control and the nonlinear LDM, what does the actual curve of the current LDM system look like? Is it anything like a real log taper?
Any help in understanding this would be truly appreciated.
Karl
Over the course of the last week or so, I have waded through the 3288 posts (so far) in this thread. Whew! Fascinating subject, and I would like to build an LDR volume control and play with the concept a little. My existing system is very average – M Audio 2496 sound card, gainclone amp, BBC Monitor speakers – but I’d love to see what this control can do for me, especially in conjunction with a new outboard DAC for my computer.
I find that after all of this reading I understand quite a bit about the LDR system -- the power requirement and the characteristics of the LDR, etc – but have little fundamental understanding of the basic characteristics and specifications of a best-performance, standalone volume control. I want to know what the ideal characteristics would be if there were no technical limitations, and then try to get as close as I can with the LDR.
Is there a section in diyaudo where I could go to read more about what the “ideal” volume control would look like? Or somewhere on the web where it is discussed? Any suggestions appreciated.
Specific questions I would like answered somehow:
1. Given the typical source output and amplifier input impedances of current systems, what is the ideal potentiometer value that would best cover the most source/amplifier combinations well? 50K? 100K? 250K? Some other value?
2. If the output impedance of the volume control is important to the quality of the sound and should ideally remain constant (I’m gathering that this is the case from the reading I’ve done here), can that be accomplished somehow in theory? I’m trying to understand the underlying principles here.
3. I’ve read the Elliot article on volume controls. It appears that commercially-available log (audio taper) controls are not log at all, just a combination of two linear tracks that roughly approximate the log taper. Is this really acceptable, or would a real log taper be noticeably more pleasant to deal with?
4. Between the recommended log taper control and the nonlinear LDM, what does the actual curve of the current LDM system look like? Is it anything like a real log taper?
Any help in understanding this would be truly appreciated.
Karl
Hi Karl
1: If using a standard log pot (like an Alps Blue Velvet) as a passive attenuator then a 10k log would be IMO the value to use, into those impedances, and your source would need to be less than 200ohms output impedance.
2:The commercial Lightspeed Attenuator as I build them to my circuit uses quad matched NSL32SR2S this mimics if there were such a thing a 7k Alps Blue Velvet log pot, the output impedance remains far more consistent than series resistor shunt LDR's that some are trying to make only because they easier to match and cheaper to make but don't sound nearly as good and cannot get good minimum volume level and have variable output impedances according on where the volume control is set.
Cheers George
Also beside what Andrew has stated, the environment inside a pre or poweramp is not conducive to good temperature stability for the LDR's, I go to great lengths in the production Lightspeed Attenuator to thermally couple all 4 matched NSL32SR2S's together, then for good measure I then pot them in hard wax to keep them that way.
Cheers George
George, thank you for your response, and the information. I've been looking on the web for information on designing attenuators, and have built a spreadsheet that lets me play around with "T" and "Pi" configurations with constant Zo. I just got down to looking at the various possibilities late last night, but I can see that there must be compromises to get attenuation anywhere below about 50db. I'll keep looking.
Karl
Modify attenuator level control
I remember reading some posts by people who were not satisfied with the LDR because the sound was still too loud with the level set at minimum. I came across these techniques for either lowering or raising the attenuation range, and am wondering if sound quality would be affected:
Modified Attenuator Tapers
I remember reading some posts by people who were not satisfied with the LDR because the sound was still too loud with the level set at minimum. I came across these techniques for either lowering or raising the attenuation range, and am wondering if sound quality would be affected:
Modified Attenuator Tapers
Its true that zero volume is not possible with this configuration. Best sound quality possible is possible though so its a give and take. Its frankly to loud with my gainclone and 96db efficient speakers. But think about that. 96db with a 'gain'clone. Its just asking for trouble. If you have 90db speakers or less you will be fine. Tube amp... almost in all cases you will be fine. An amp that is not a gain monster and you will be fine. Any ONE of those or a combination of those options and you will have very low min volume.
That Goldpoint thing will work but then you have regular resistors in the attenuator and a big part of the sound is the fact that these LDRs are the best sounding resistor out there. Anything less is something less at the output. If you did your experiment using LDRs well then that would work real well. It would mess with the impedance of course. But again as George said if you have a source with very low output impedance than it wont matter.
Uriah
That Goldpoint thing will work but then you have regular resistors in the attenuator and a big part of the sound is the fact that these LDRs are the best sounding resistor out there. Anything less is something less at the output. If you did your experiment using LDRs well then that would work real well. It would mess with the impedance of course. But again as George said if you have a source with very low output impedance than it wont matter.
Uriah
Although the resistor divider ahead of the LDR is indeed not an LDR, I had understood from reading the posts that the real issue is to avoid any mechanical contacts and any mechanical potentiometer. Two high quality resistors soldered in place will not harm the sound, will they? The signal path is chock-a-block with resistors, after all.
Regarding the impedance matching with the source, are there no resistor values you could put in there that would allow any reasonable source impedance to work fine while also playing nice with the LDR range of resistances?
Not challenging your remarks here -- I'm on the steep part of the learning curve.
You are correct there Wapo, if a voltage divider is placed before the Lightspeed Attenuator circuit you will indeed change the input impedance the source is seeing, and as well the output impedance of the Lightspeed itself. Impedance matching the system together then may well be thrown out the window.
I have a customer that has a cdp source that is 4v output, and speakers of 105db!!, he has no trouble, at minimum you can bearly hear it and his normal listening level is about 9o'clock and loud level is around 11o'clock. About the same as an active pre would be.
My system cdp 2v, speakers 86db, my normal listening level is about 11o'clock and 2 o'clock for loud level listening
Cheers George
Last edited:
Unsure the Vout of my CDP but the gain of my amp has to have a lot to do with it. Also I dont run the LDRs lower than 45-50R. I run them at 6k for Rtot so when we have 6k/50R with a bunch of gain, a 50R out CDP and 96db this is probably the problem. Maybe if I ran the LDRs at 25R but I dont like running them that low. Its to close to destruction for me.
Uriah
Most amps sensitvity I've seen on the market are 1v in for maximum output just before clipping, some odd ones like Quads are .5v
CDP red book standard is 2v output on the loudest possible transient 0db, a few are a bit over this, hardly any are under 2v
I run my NSL32SR2S's at 20mA max this is 5mA below their rating, never had one blow yet over 6 years.
Cheers George
CDP red book standard is 2v output on the loudest possible transient 0db, a few are a bit over this, hardly any are under 2v
I run my NSL32SR2S's at 20mA max this is 5mA below their rating, never had one blow yet over 6 years.
Cheers George
uC controlled VCCS
..... my VCCS (Paul Hynes) is now running and I´m able to control the DS1802 by a microcontroller, that reads the serial port of my networkplayer.
I wonder if there is still room for improvement, the sound (especially with 24bit/96khz files) is very "life", e.g. with jazz-trio recordings.
Maybe, sometime later I´ll try a buffer between the LDRs and the (only 10k) input of the active crossover (Orion ASP). But I´m not expecting too much of it.
Really great. Thanks George, Uriah, Paul and all the others, contributing to this George´s forum.
Cheers, Frank
..... my VCCS (Paul Hynes) is now running and I´m able to control the DS1802 by a microcontroller, that reads the serial port of my networkplayer.
I wonder if there is still room for improvement, the sound (especially with 24bit/96khz files) is very "life", e.g. with jazz-trio recordings.
Maybe, sometime later I´ll try a buffer between the LDRs and the (only 10k) input of the active crossover (Orion ASP). But I´m not expecting too much of it.
Really great. Thanks George, Uriah, Paul and all the others, contributing to this George´s forum.
Cheers, Frank
I have been spreadsheeting the "T" attenuator numbers using the 7k that George said his system operates at. In Uriah's case (for the T attenuator), the attenuation for Zo=6k and R2=50R is 49dB. By taking R2 down to the theoretical 40R for one LDR, attenuation is 51dB -- only 2dB better.
Using my goal of Zo=6000, at 40R, attenuation is about 47.5dB, but I want to parallel two LDRs for R2 and run R2 down to 20~24R, which will give me a best case max attenuation of 57dB. I believe that this is a reasonable design goal provided you have a stiff, well-regulated supply for the LDRs and precision control of the inline variable resistance.
While you are at max current and it will be hard on the R2 LDRs at this level, the fact is that you won't be at this level while actually listening, and you don't have to use this level for turning the sound way down, especially if you include in the design provisions that unload and remove the R2 LDRs from operation at lower than 57db attenuation.
I think for real-world operation you can build a system where you will almost never be at 20R for two paralleled R2 LDRs in a Zo=7000 T attenuator design (or solve everything by using three LDRs for an R2=13R -- 60dB --system!). Uriah has piqued my curiosity -- I'm going to run the numbers for a Zo=6k pot and see if that would also work well. That's my goal at this point, we'll see if I can make it work. Skepticism is definitely appropriate at this stage!
Karl