Remote
I don't mean how you physically cut the traces. But I mean it does not appear you are using the pot as a dual gang. It looks like you are possibly running the 5V to the wiper, and using the other two pins to get the variable resistance. Am I right? Is it possible to use this board in true dual gang mode, like George has shown in his original lightspeed schematic? You only have 3 wires shown connected in your photos.
I don't mean how you physically cut the traces. But I mean it does not appear you are using the pot as a dual gang. It looks like you are possibly running the 5V to the wiper, and using the other two pins to get the variable resistance. Am I right? Is it possible to use this board in true dual gang mode, like George has shown in his original lightspeed schematic? You only have 3 wires shown connected in your photos.
Yes, I AM using it as a dual pot, and you only need THREE wires to do so, as each wiper & one side of each pot are the same point electrically.
It just so happens that's the way the pot is connected to the board. You only have to connect 3 wires, and it works perfectly. I verified all this by taking measurements with 2 meters to see one side's R increase while the other side decreases.
I just built the lightspeed attenuator for my B1 buffer. The result is awesome to me. Compared to the TKD potentiometer (It is already a good pot made of conductive plastics), the optical attenuator is even much better. It has a soft, warm, transparent, and natural sounding.
However, I have a problem using the optical attenuator that its attenuation curve is like a linear type pot. The loudness changes greatly at the beginning steps (7 o'clock to 9 o'clock). For almost CD, the loudness is too high when the knob is set over than 8 o'clock so that I can only adjust the volume within a small range (7 o'clock to 8 o'clock). The 100K dual pot used for LED control is already a log type one. IIRC, a linear type pot may be worse in my case.
Currently, a 50K resistor is added in series of the input signal and the loudness can be adjusted in a reasonable range (8 o'clock to 13 o'clock). However, the volume is limited by the 50K resistor so that the loudness is somewhat low for low level CDs.
In short, how can I get the attenuation curve of the lightspeed attenuator as well as a traditional log type pot? Hope someone here could help me.
However, I have a problem using the optical attenuator that its attenuation curve is like a linear type pot. The loudness changes greatly at the beginning steps (7 o'clock to 9 o'clock). For almost CD, the loudness is too high when the knob is set over than 8 o'clock so that I can only adjust the volume within a small range (7 o'clock to 8 o'clock). The 100K dual pot used for LED control is already a log type one. IIRC, a linear type pot may be worse in my case.
Currently, a 50K resistor is added in series of the input signal and the loudness can be adjusted in a reasonable range (8 o'clock to 13 o'clock). However, the volume is limited by the 50K resistor so that the loudness is somewhat low for low level CDs.
In short, how can I get the attenuation curve of the lightspeed attenuator as well as a traditional log type pot? Hope someone here could help me.
Are you using quad matched NSL32SR2S and a 100k log pot as per my circuit? as you will get what you have if not.
If you only have matched series and matched shunts you could try swapping the series for shunts and visa versa, this should make it behave the opposite of what you have, then you can take the "lesser of two weevils" Russell Crowe from Master & Commander great movie.
Cheers George
If you only have matched series and matched shunts you could try swapping the series for shunts and visa versa, this should make it behave the opposite of what you have, then you can take the "lesser of two weevils" Russell Crowe from Master & Commander great movie.
Cheers George
Hi George,
The NSL-32SR2 are only matched for two pairs (one for series and the other for shunt.)
I will try to swap them. Thanks for your reply 🙂
The NSL-32SR2 are only matched for two pairs (one for series and the other for shunt.)
I will try to swap them. Thanks for your reply 🙂
how have people got on with dvb-projekt's PCB.
My amp is nice and quiet but with this circuit in place i'm now getting buzz 🙁
My amp is nice and quiet but with this circuit in place i'm now getting buzz 🙁
Hi Ted,
this is normally impossible because there is only the passive resistor part of the LDR in the signal line...
Have you checked the GND connections?
Balanced
I am looking for anyone who has implemented an ldr in balanced configuration with success. I have two ugs modules I would like to use an ldr solution with. Any help much appreciated.
dave
I am looking for anyone who has implemented an ldr in balanced configuration with success. I have two ugs modules I would like to use an ldr solution with. Any help much appreciated.
dave
Bringing this thread back to the top.
I've not sorted the GNDing issues however it seems to not allow full volume. For example if i put a 2v signal in i can only get 1.7 out.
can anyone recommend what may be causing this ?
I've not sorted the GNDing issues however it seems to not allow full volume. For example if i put a 2v signal in i can only get 1.7 out.
can anyone recommend what may be causing this ?
disconnect all the audio in out interconnects.
Set the LED/LDR to max attenuation.
Measure the input resistance (of both channels) from +IN to +OUT and from +IN to Audio Ground.
Measure the output resistance (again both channels) from +OUT to Audio Ground.
Set the LED/LDR to max volume.
Repeat all six measurements.
Report back.
Set the LED/LDR to max attenuation.
Measure the input resistance (of both channels) from +IN to +OUT and from +IN to Audio Ground.
Measure the output resistance (again both channels) from +OUT to Audio Ground.
Set the LED/LDR to max volume.
Repeat all six measurements.
Report back.
None of the LDRs will go to 0 ohms. Its a voltage divider. You cant get X out if you put X in but you get a fraction of X since you can not go to 0 Ohms.
Uriah
Uriah
paralleling LDRs
a friend of mine was thinking to parallel many LDRs (5-10 in parallel) in order to lower the LDR´s resistance. The LEDs can be in series to assure the same intensity.
Another advantage of paralleling might be the much more relaxed situation with the matching of the LDRs.
Anybody have tried this before? Any comments?
Regards
Mamal
a friend of mine was thinking to parallel many LDRs (5-10 in parallel) in order to lower the LDR´s resistance. The LEDs can be in series to assure the same intensity.
Another advantage of paralleling might be the much more relaxed situation with the matching of the LDRs.
Anybody have tried this before? Any comments?
Regards
Mamal
Yes,
paralleling of the shunt LDRs, but series connecting the LEDs will average out some of the variations.
The series LDRs must be series connected to increase the series resistance to give a higher attenuation.
Doubling the LED/LDR from 4 to 8 will increase the range of attenuation by ~11.9dB.
Doubling again to 16 LED/LDR will increase the attenuation by a further ~11.9dB gaining ~24dB of extra attenuation. With 4devices in either series or parallel then averaging is starting to have an effect.
I have never tried it and I don't have the maths to predict the probabilities of good channel matching over the range of attenuation. Using 5 to 10 devices in each group must give better averaging, but, is that economic? That's 20 to 40 LED/LDR for an unbalanced two channel attenuator.
paralleling of the shunt LDRs, but series connecting the LEDs will average out some of the variations.
The series LDRs must be series connected to increase the series resistance to give a higher attenuation.
Doubling the LED/LDR from 4 to 8 will increase the range of attenuation by ~11.9dB.
Doubling again to 16 LED/LDR will increase the attenuation by a further ~11.9dB gaining ~24dB of extra attenuation. With 4devices in either series or parallel then averaging is starting to have an effect.
I have never tried it and I don't have the maths to predict the probabilities of good channel matching over the range of attenuation. Using 5 to 10 devices in each group must give better averaging, but, is that economic? That's 20 to 40 LED/LDR for an unbalanced two channel attenuator.
Last edited:
It is something I have been considering and will do either later this year or start of next year. I have redesigned the power supply for it and definitely the regular LSA power supply must be examined to see if it will handle the extra load and no longer will the regular volume pot do the job. When we get more than the normal 4 LDRs the volume pot will smoke and die at low volume. If the power supply is noisy with just 4 it will be so much more noisy with more LDRs so the power supply is extremely important here. Any noise at all will be amplified by injecting it multiple times into the LEDs of each shunt LDRs.
Uriah
Uriah
Yes. I have wired them parallel but with a different power supply where a pot is not used for volume control. I have used 3 per shunt so far and 2 per series.
When you have a .25W to maybe .5W pot and you push minimum 80mA through it at 5V the wiper, as you have often pointed out, will not last long.
Uriah
When you have a .25W to maybe .5W pot and you push minimum 80mA through it at 5V the wiper, as you have often pointed out, will not last long.
Uriah
Wire the shunt LDRs in parallel to obtain a greater attenuation when at minimum volume.
Wire the series LDRs in series to obtain a greater attenuation when at minimum volume.
Wire the shunt LEDs in series.
Wire the series LEDs in series.
The control pots see the same currents as they would if single LED/LDRs were used. Over current through the pot wiper is no worse with multiple sets of LED/LDRs
Wire the series LDRs in series to obtain a greater attenuation when at minimum volume.
Wire the shunt LEDs in series.
Wire the series LEDs in series.
The control pots see the same currents as they would if single LED/LDRs were used. Over current through the pot wiper is no worse with multiple sets of LED/LDRs
thanks for the feedbacks. it seems that it is worth a try to parallel/serie the LDRs.
by using many LDRs, do you think that it still needs to sellect/match the DLRs or the effect of "averraging" relaxs the whole issue of matching? 🙂
cheers
mamal
by using many LDRs, do you think that it still needs to sellect/match the DLRs or the effect of "averraging" relaxs the whole issue of matching? 🙂
cheers
mamal
Andrew, I must be missing what you are saying. Probably.
When you wire a total of 4 LDRs into the shunt position you will get 80mA draw at min volume in a Lightspeed configuration. In the Lightspeed only 2 LDRs are ever drawing any significant amount of current at one time, never the 4 of them. So if we double the shunt LDRs then we double the current draw at min volume.
Uriah
When you wire a total of 4 LDRs into the shunt position you will get 80mA draw at min volume in a Lightspeed configuration. In the Lightspeed only 2 LDRs are ever drawing any significant amount of current at one time, never the 4 of them. So if we double the shunt LDRs then we double the current draw at min volume.
Uriah
- Status
- Not open for further replies.