The ST70 is about the easiest load imaginable, so the problem is elsewhere.
So if the other preamp didn't work for whatever reason, would the LDR present the same impedances?
No, ideal input impedance to design ( die for
) for NSL32SR3 LDR;s is approx 14k-20kThere is no high frequency losses with LDR;s ,...that you presently attribute to the other
passive device.
Cheers / Chris
The 13A control board is proving very popular, and offers exceptional audio quality
StereoCoffee LDR Passive Preamp Upgrade Preview - Product Reviews & Opinions - StereoNET
The new free regulator board that suits 50k potentiometers, provides opportunity to use the kit with a remote control. http://stores.ebay.com.au/stereo-coffee
Cheers / Chris
StereoCoffee LDR Passive Preamp Upgrade Preview - Product Reviews & Opinions - StereoNET
The new free regulator board that suits 50k potentiometers, provides opportunity to use the kit with a remote control. http://stores.ebay.com.au/stereo-coffee
Cheers / Chris
Hi Chris,
I have been through this entire thread - very interesting, thanks for sharing.
I am missing some info though. I was wondering if you could help...
1- What is the input impedance of this passive pre?
2- Do you have a graph or at least the max value of the output impedance? That is important to make sure all is OK for the next stage...
3- What is the max attenuation you can use while still having a linear volume controller? I understand you can go to "nearly infinite" attenuation when using the LDRs to switch sources (relais function), but would like to know the usuable range of the volume control (so from 0 to -??? dB).
4- Any more info on how to control volume with a remote control? Can the sources be switched per remote aswell in your latest kit?
Many thanks for your kind reply
Claude
I have been through this entire thread - very interesting, thanks for sharing.
I am missing some info though. I was wondering if you could help...
1- What is the input impedance of this passive pre?
2- Do you have a graph or at least the max value of the output impedance? That is important to make sure all is OK for the next stage...
3- What is the max attenuation you can use while still having a linear volume controller? I understand you can go to "nearly infinite" attenuation when using the LDRs to switch sources (relais function), but would like to know the usuable range of the volume control (so from 0 to -??? dB).
4- Any more info on how to control volume with a remote control? Can the sources be switched per remote aswell in your latest kit?
Many thanks for your kind reply
Claude
Hi Claude
Being a variable passive attenuator, it mimics approximately a 12k potentiometer:
Some typical figures I measure on input 3 this evening
At zero volume 220k in the series LDR between input and output
At zero volume 300 ohms in the shunt LDr between output and ground
At mid volume 10.3k ohms in the series LDr between input and output
At mid volume 11.4k ohms in the shunt LDR between output and ground
At full volume 300 ohms in the series LDR between input and output
At full volume 220k in the shunt LDR between output and ground
This is achieved with a extremely low overall voltage leaving the regulator approaching ideal characteristics namely 2.7v to 2.5v - hence the slightly higher figures of 300 ohms, than normally seen with the NSL32SR3 which can achieve 60 ohms when run at 20ma. The current expended at vol min/max is just 5 ma, so extremely kind on the LDR.
I use a quad 306 power amp with 330mv for full output loaded by 20k and can zero volume Normally full quieting is achieved where the series LDR input to output resistance at zero volume is above 240k. So a small trade off, given the voltage and current are so low and the far better audio that results, and effectively full quieting with less sensitive power amps .
Your question 3 I read further as asking about input switching, I do not use relays rather use the R off characteristics of the NSL32SR3 to be a switch where it achieves above 20 mega ohms,
there is no bleed between inputs, and the LDR forms a natural switch in its off state ie no anode current.
The remote volume can be achieved with many of the available motorised types using a usually a dual 50k Alps. Here the normal pots Rin Lin becomes series cathode return, Rout Lout is wiper and Signal gnd is shunt cathode. Because the signal ground is often also the body metal of the potentiometer, effort has to be made to either isolate on the circuit board or insulate the shaft, as kits do not ground cathodes rather see cathodes as a potential difference to be used.
Yes sources can be switched - if the remote controller has this sophistication, switching is achieved on the anode side rather than cathode which is volume..
I hope that helps
Cheers / Chris
Being a variable passive attenuator, it mimics approximately a 12k potentiometer:
Some typical figures I measure on input 3 this evening
At zero volume 220k in the series LDR between input and output
At zero volume 300 ohms in the shunt LDr between output and ground
At mid volume 10.3k ohms in the series LDr between input and output
At mid volume 11.4k ohms in the shunt LDR between output and ground
At full volume 300 ohms in the series LDR between input and output
At full volume 220k in the shunt LDR between output and ground
This is achieved with a extremely low overall voltage leaving the regulator approaching ideal characteristics namely 2.7v to 2.5v - hence the slightly higher figures of 300 ohms, than normally seen with the NSL32SR3 which can achieve 60 ohms when run at 20ma. The current expended at vol min/max is just 5 ma, so extremely kind on the LDR.
I use a quad 306 power amp with 330mv for full output loaded by 20k and can zero volume Normally full quieting is achieved where the series LDR input to output resistance at zero volume is above 240k. So a small trade off, given the voltage and current are so low and the far better audio that results, and effectively full quieting with less sensitive power amps .
Your question 3 I read further as asking about input switching, I do not use relays rather use the R off characteristics of the NSL32SR3 to be a switch where it achieves above 20 mega ohms,
there is no bleed between inputs, and the LDR forms a natural switch in its off state ie no anode current.
The remote volume can be achieved with many of the available motorised types using a usually a dual 50k Alps. Here the normal pots Rin Lin becomes series cathode return, Rout Lout is wiper and Signal gnd is shunt cathode. Because the signal ground is often also the body metal of the potentiometer, effort has to be made to either isolate on the circuit board or insulate the shaft, as kits do not ground cathodes rather see cathodes as a potential difference to be used.
Yes sources can be switched - if the remote controller has this sophistication, switching is achieved on the anode side rather than cathode which is volume..
I hope that helps
Cheers / Chris
It varies with each volume setting.
Usually the maximum output impedance is @ the half volume setting (-6dB ref maximum)
Do what ChrisD has done. Measure the actual LDR resistances at the various settings you want/need to investigate.
You should be taking these measurments when selecting which LED/LDR best suit each other for a reasonably balanced stereo volume control.
Your LED/LDR maximum attenuation for a Selector Switch is limited by the minimum LDR resistance of the shunt element. This is nowhere near infinite attenuation.
There are Threads discussing remote controls for these LED/LDR volume controls
Last edited:
Thanks to both for the comprehensive answers.
I am still confused about the pre's output impedance. Say it is highest at mid volume, does that mean output impedance is then roughly 12k Ohm? (vs say "4k ish" for a conventional 12k potentiometer).
I understand they are others factors to take into account to calculate what is happening re energy transfer and frequency (cables, power amp input impedance across frequency etc.), but roughly said, does that mean that "worst case" the power amp "sees" 12k from the pre?
If so, even using short very low capacity cable, it would probably make it difficult to mate it with a power amp that has a 20k input impedance and that would probabaly like to "see" say 5k max?
Am I correct? I know 20k input is lowish and that normaly 50 to 100k are usual for power amp, but well...
Last but not least, regarding the variations of the pre's input impedances, if these are variable then only a test with my sources would reveal if it affects the sound at various volume settings... or not. I would though have hoped the pre's input never drops below 10k across all the usable volume setting range...
Thanks again
Claude
I am still confused about the pre's output impedance. Say it is highest at mid volume, does that mean output impedance is then roughly 12k Ohm? (vs say "4k ish" for a conventional 12k potentiometer).
I understand they are others factors to take into account to calculate what is happening re energy transfer and frequency (cables, power amp input impedance across frequency etc.), but roughly said, does that mean that "worst case" the power amp "sees" 12k from the pre?
If so, even using short very low capacity cable, it would probably make it difficult to mate it with a power amp that has a 20k input impedance and that would probabaly like to "see" say 5k max?
Am I correct? I know 20k input is lowish and that normaly 50 to 100k are usual for power amp, but well...
Last but not least, regarding the variations of the pre's input impedances, if these are variable then only a test with my sources would reveal if it affects the sound at various volume settings... or not. I would though have hoped the pre's input never drops below 10k across all the usable volume setting range...
Thanks again
Claude
Last edited:
Perhaps to make things easier...
Apart from the "real life volume adjustability", which depends on a lot of variable from all components (efficiency, sensitivity) etc.
A- Is this pre OK to use with reasonable cables (low capacity, 1m length max) with sources that have up to 2k output impedance?
B- Is this pre OK to use with power amps that have only 20k imput impedance (again using max 1m low capacity interconnects)
Or... is it a gamble and only testing can tell?
Many thanks and sorry again for down to earth questions
Claude
Apart from the "real life volume adjustability", which depends on a lot of variable from all components (efficiency, sensitivity) etc.
A- Is this pre OK to use with reasonable cables (low capacity, 1m length max) with sources that have up to 2k output impedance?
B- Is this pre OK to use with power amps that have only 20k imput impedance (again using max 1m low capacity interconnects)
Or... is it a gamble and only testing can tell?
Many thanks and sorry again for down to earth questions
Claude
I love using LDRs and LEDs as the perfect linear opto isolator, but a preamp must have gain. An input impedance of 100K and an output impedance as low as possible.
But why not use a 4066? Which is a quad analogue switch.
I have built a circuit for a manufacturer of juke boxes that automatically and seamlessly adjusts the volume, using a peak detector and four comparators
But why not use a 4066? Which is a quad analogue switch.
I have built a circuit for a manufacturer of juke boxes that automatically and seamlessly adjusts the volume, using a peak detector and four comparators
You should take account of the cable capacitance and the RF attenuation capacitance in the receiver when assessing the roll-off due to high source impedance.
I aim <<2k source impedance and generally use a Buffer to drive the cable and other capacitances.
If your LDR resistances end up @ 12k then the LDR has a maximum output impedance of 6k. But based on George's recommendation I think you might find that the max output impedance is a bit lower than that.
F-3dB = 1/{2 Pi R C}
Using 50pF for the cable and 47pF at the Receiver with Rs=6k you end up with 273kHz.
That's a full decade above the audio band and very few if any of us would hear that roll-off.
But if you have RF attenuation of 1k +1nF in the Receiver, then that changes to a 3pole roll-off with the lowest pole at 1/{2 Pi 7kohms 1097pF} = 21kHz
The cable effect has turned out to be relatively minor compared to the Rs effect combined with a fairly normal RF filter at the amplifier.
I aim <<2k source impedance and generally use a Buffer to drive the cable and other capacitances.
If your LDR resistances end up @ 12k then the LDR has a maximum output impedance of 6k. But based on George's recommendation I think you might find that the max output impedance is a bit lower than that.
F-3dB = 1/{2 Pi R C}
Using 50pF for the cable and 47pF at the Receiver with Rs=6k you end up with 273kHz.
That's a full decade above the audio band and very few if any of us would hear that roll-off.
But if you have RF attenuation of 1k +1nF in the Receiver, then that changes to a 3pole roll-off with the lowest pole at 1/{2 Pi 7kohms 1097pF} = 21kHz
The cable effect has turned out to be relatively minor compared to the Rs effect combined with a fairly normal RF filter at the amplifier.
Hi Claude
You are seeing problems, where really there are none, let me explain.
Attenuation you are understanding as I see as receiving something less, when this is not the case at all. Rather high quality attenuation is fully necessary. Todays audio components have prodigious outputs with very low output impedance, making attenuation entirely necessary and now fully capable of being solved by passive attenuation, rather than former active methods.
The exception is a phono stage that was formerly part of of an active preamp, but is easily catered for, by a stand a lone phono stage that can have its own power supply and dedicated circuitry.
Regarding 10k or less, unsure of what you base that on, but their is general agreement that a passive needs to be at mid volume about 15k, and some go as high as 20k .
With cables I have experienced absolutely no loss at all. My own system sees a 5 meter length of interconnect, because I locate my power amp close to my speakers. It delivers outstanding audio quality and clarity, I couldn't ask for more,
Cheers / Chris
Hi Claude
The 3 input LDR kit suits the majority of audio systems, it suits power amps with as much as 1.5v sensitivity loaded by 50k., or lower sensitivity and lower loading as exampled by the Quad 306 power amplifier which is 330mv loaded by 20k.
It also suits every known interconnect cable type, however my preference is for coax type cables.
A review here of Stereo Coffee vs Lightspeed
Comparison Lightspeed DIY LDR Pre VS Stereo Coffee DIY - DIY Audio Projects - StereoNET
However there have been many changes further improving the stereo coffee since that review,
"I find my jaw dropping in disbelief every time I press play. Hence, I tend to stay up too late frequently now. My perception of 'low cost' sound has been shattered and I now realise that less components in the signal chain can and does give more accurate reproduction with quality speakers."
StereoCoffee LDR Passive Preamp Upgrade Preview - Product Reviews & Opinions - StereoNET
Cheers / Chris
Thanks Chris.
Actually I am quite aware of that as my current set up uses a my old faithful DIY passive pre (Alps Black Beauty + source switch, 3 decades old). It is a (highish) 50k log pot so I am not really afraid of losing anything (not to mention that voltage is not something I am sensitive to
The problem is I know from experience that passive devices not always match with any kind of partners, obviously. Sound can either be distorted or dull. As engineer I should have known the answers but I got leasy over the decades
I designed ages ago my own devices and I am very happy with them but should they fail one day I would probabably be to leasy to start again from scratch and would rather look at current high quality DIY modules rather than trying to play with SMD & Co.
Your pre wouldn't be a problem in my current set up as replacement of my current 50k BB pot... Obviously (well, perhaps max attenuation could be a limitation). And that pot was a good choice as my sources weren't as low output impedance as today's, while having sides line 10V output (and enough sides phono) so really attenuation only is needed with reasonable amps to drive my LS as loud as needed. As I did it all myself, I could though easily reduce the source output gain (and probabaly even gain a bit of band/speed while at it, who knows). Today's power amp isn't a prob either as that's a 80k input 30 cm away from the pre. I am not focused on the 10x impedance rule, it is all a matter of low pass filter as Andrew posted and values so that exact calculation give the answer on OK or not.
My problem comes if I want to replace pre+amp one day to move to something that is at least as good sound wise (hence me looking at this promising thread). But then I am also looking into a promising amp module that has a nearly fixed 20k input (eg low), that can be somewhat extended at a negligeable cost of more noise but there is a limit to it (say 30k).
So my sole concern was: is your device capable of "driving / transmitting" to a 20k input impedance amp across all the adjustable volume range with "reasonable cables" without any sonic drawback. As ever calculating is key... and probably confirming/testing then aswell. That amp I am considering has already a reputation on this forum for not matching well with various passive pres due to impedance issues (not gain), hence posting
Thanks again
Claude
Actually I am quite aware of that as my current set up uses a my old faithful DIY passive pre (Alps Black Beauty + source switch, 3 decades old). It is a (highish) 50k log pot so I am not really afraid of losing anything (not to mention that voltage is not something I am sensitive to
The problem is I know from experience that passive devices not always match with any kind of partners, obviously. Sound can either be distorted or dull. As engineer I should have known the answers but I got leasy over the decades
I designed ages ago my own devices and I am very happy with them but should they fail one day I would probabably be to leasy to start again from scratch and would rather look at current high quality DIY modules rather than trying to play with SMD & Co.
Your pre wouldn't be a problem in my current set up as replacement of my current 50k BB pot... Obviously (well, perhaps max attenuation could be a limitation). And that pot was a good choice as my sources weren't as low output impedance as today's, while having sides line 10V output (and enough sides phono) so really attenuation only is needed with reasonable amps to drive my LS as loud as needed. As I did it all myself, I could though easily reduce the source output gain (and probabaly even gain a bit of band/speed while at it, who knows). Today's power amp isn't a prob either as that's a 80k input 30 cm away from the pre. I am not focused on the 10x impedance rule, it is all a matter of low pass filter as Andrew posted and values so that exact calculation give the answer on OK or not.
My problem comes if I want to replace pre+amp one day to move to something that is at least as good sound wise (hence me looking at this promising thread). But then I am also looking into a promising amp module that has a nearly fixed 20k input (eg low), that can be somewhat extended at a negligeable cost of more noise but there is a limit to it (say 30k).
So my sole concern was: is your device capable of "driving / transmitting" to a 20k input impedance amp across all the adjustable volume range with "reasonable cables" without any sonic drawback. As ever calculating is key... and probably confirming/testing then aswell. That amp I am considering has already a reputation on this forum for not matching well with various passive pres due to impedance issues (not gain), hence posting
Thanks again
Claude
Hi Claude
Where the sensitivity of the power amp is 500mv or lower , 20k loading will be fine, and Yes will achieve that without any sonic drawbacks. If sensitivity is higher circa 1v on your power amp ( given it is a DIY project ) then simply replace the 20k resistor fixed loading with a higher value like 100k... easy.
An asset sometimes not factored in of LDR's is to quieten the noise floor so that it is basically a silent background, so if that is allowed for, you are achieving a far better signal to noise ratio than active preamps , and hence more music.
Cheers / Chris
Where the sensitivity of the power amp is 500mv or lower , 20k loading will be fine, and Yes will achieve that without any sonic drawbacks. If sensitivity is higher circa 1v on your power amp ( given it is a DIY project ) then simply replace the 20k resistor fixed loading with a higher value like 100k... easy.
An asset sometimes not factored in of LDR's is to quieten the noise floor so that it is basically a silent background, so if that is allowed for, you are achieving a far better signal to noise ratio than active preamps , and hence more music.
Cheers / Chris
Thanks Chris
I asked the designer of the module and apparently going higher than 30k could be a problem. Sensivity is close to 2V.
The modules I am interested in are the FO 1.4, on this forum aswell. They are quite popular. A matching pre will be released by the designer for these- probably the intuitive choice, but probably active while I quite like the passive concept.
Your pre is pretty popular so perhaps I should just wait for current FO owners to test further pres to find out what works best.
Claude
I asked the designer of the module and apparently going higher than 30k could be a problem. Sensivity is close to 2V.
The modules I am interested in are the FO 1.4, on this forum aswell. They are quite popular. A matching pre will be released by the designer for these- probably the intuitive choice, but probably active while I quite like the passive concept.
Your pre is pretty popular so perhaps I should just wait for current FO owners to test further pres to find out what works best.
Claude
2020 Stereo Coffee Kit Review
The 2020 stereo coffee kit is reviewed here:
2020 StereoCoffee LDR passive preamp redesign. - DIY Audio Projects - StereoNET
Happy listening.
The 2020 stereo coffee kit is reviewed here:
2020 StereoCoffee LDR passive preamp redesign. - DIY Audio Projects - StereoNET
Happy listening.