I do not get it. You tell us that it does not matter but in practice you have found big differences in sound. Isn't it time to change your believes ?
I would think that all the noise from the power supply is modulated on the lightsource and thus is transferred to audio.
What he means is that the quality of the supply matters. The cost of the resistors and the caps does not matter. Its the design of the supply not the quality of the parts that comes into play here. If you design a terrible supply with Black Gates its not going to help your cause, but if you build a very good supply with regular 50 cent caps and 10 cent resistors you can have the best sound.
Uriah
hi Patrick,
Most of the pots available in Australia form the local electronics stores are 0.5w, so you need to use a dual pot to get effetively a 1w pot.
If I recall correctly, George tried using single pots for a short while then reverted back to the dual pots.
regards
I don't understand the argument.
In normal circumstances, either both arms are allowing the same current through (which would be 50k:50k), or one arm (the shunt) will have significantly higher current.
Assuming that you will drop something like 2.5V across the pot at 20mA max (limit of the LDR), you would be heating the shunt part of the pot to 50mW (way below 500mW) while there is hardly any power dissipated on the other part. That does not change at all whether you use a dual pot or not.
The pot spends most of its time at listening level, which I would say around 1-5mA. At that level, the dissipation is more like 5mW in total at most. And since the current is more or less constant (I hope), there is no issue for cross coupling between the two arm.
Patrick
In normal circumstances, either both arms are allowing the same current through (which would be 50k:50k), or one arm (the shunt) will have significantly higher current.
Assuming that you will drop something like 2.5V across the pot at 20mA max (limit of the LDR), you would be heating the shunt part of the pot to 50mW (way below 500mW) while there is hardly any power dissipated on the other part. That does not change at all whether you use a dual pot or not.
The pot spends most of its time at listening level, which I would say around 1-5mA. At that level, the dissipation is more like 5mW in total at most. And since the current is more or less constant (I hope), there is no issue for cross coupling between the two arm.
Patrick
I've gone down that path Patrick and the total current draw and being dc in nature was too much for single pots, in the end they started to misbehave.
Cheers George
I do not get it. You tell us that it does not matter but in practice you have found big differences in sound. Isn't it time to change your believes ?
I would think that all the noise from the power supply is modulated on the lightsource and thus is transferred to audio.
Hi Jaap,
Trust me, I have no problem using top parts if it'll make a difference.
In this case, I very much doubt a $10 vishay bulk metal foil resistor on the input of the LDR would make any difference. The problem with the wallwart I used is that it is likely a switch mode, not a linear regulated one like George recommends. As for the batteries, perhaps the 7805 prefers to be fed more than 9V.
Hi soongsc,
It sounds like white noise, audible from 1m or so. Usually with the lightspeed switched powered up I can put my ear on the speaker membrane and hear nothing. Interestingly, if the lightspeed is not powered there is some minimal noise, but less than what I have with the cheap wallwart.
Hi George,
I will certainly have to get one and compare. Although the SLA battery lasts for several days, charging it is still an extra chore.
When the Lightspeed is powered up there is less resistance to ground because the shunt LDR is on... so more of the junk on the signal line gets to ground rather than the amp.
Uriah
is there any other source for LDRs than silonex?
pardon my language but they are utter crap.
just over this weekend i've heard of two LDR attenuators crashing because of crap silonex photo resistors that broke down after a few months of use.
is there an alternative? someone mentioned a military stuff from the 80s?
pardon my language but they are utter crap.
just over this weekend i've heard of two LDR attenuators crashing because of crap silonex photo resistors that broke down after a few months of use.
is there an alternative? someone mentioned a military stuff from the 80s?
You are either heating them up too much when soldering or not getting the correct ones (NSL32SR2S) and matching properly, or not building to my circuit design, as I have now built hundreds of production Lightspeed Attenuators over the years and not one has gone faulty, drifted or changed it's characteristics. Only one came back when someone put a car battery on it in reverse polarity, smart man, blew to hell and back and tried to claim warranty.
Cheers George
Rocosr you seem to be the only one talking about drift in the 6 only posts you have contributed, it sounds thou like you have not even built one.
If you punish them, drive them over their rated 20mA you will do damage and yes they will change, also if they are overheated (soldered) this will also change them, and also if they are not kept uniformly temperature coupled by close proximity potting as I do, this may also cause drift on different days. Something does smell of fish.
Cheers George
If you punish them, drive them over their rated 20mA you will do damage and yes they will change, also if they are overheated (soldered) this will also change them, and also if they are not kept uniformly temperature coupled by close proximity potting as I do, this may also cause drift on different days. Something does smell of fish.
Cheers George
Tinitus,
No. An individual LDR could only drift up or down. Obviously not left or right. So when we have 4 LDRs that are matched to all track the same, all made out of the same material with the same propensity to drift being part of physics then matched LDRs would drift in the same way.
For instance the resistance will change about 4% with a 1F degree change in the room. But if all 4 change 4% in the same direction... then our voltage divider pretty much gives us the same voltage so impedance would have changed a tiny amount but there would be no problem with channel balance.
Also, if I use the word drift, my definition is a change in resistance. This is what they do. The whole idea is for them to change in resistance. If we use the word drift to mean that they would change their resistance at a given voltage and current to a different resistance at the same voltage and current then we can talk about something undesireable. So considering this is probably what you guys are talking about lets consider this. What exactly could happen for one to drift without its couterpart drifting? Its, again, a voltage divider. So if a shunt LDR's value at one measured point drifts it would have to happen because of the environment and usage. If the environment and usage are the same for both shunt LDRs then why would one change and the other not change? Circuit is the same for them both. So if one changes and the other changes, why would they change differently. So if they dont change differently but they change the same, then what will change for the audience? Its a voltage divider and both shunt LDRs are always equal to each other so if both shunt LDRs change equally there is no change for the audience.
Uriah
No. An individual LDR could only drift up or down. Obviously not left or right. So when we have 4 LDRs that are matched to all track the same, all made out of the same material with the same propensity to drift being part of physics then matched LDRs would drift in the same way.
For instance the resistance will change about 4% with a 1F degree change in the room. But if all 4 change 4% in the same direction... then our voltage divider pretty much gives us the same voltage so impedance would have changed a tiny amount but there would be no problem with channel balance.
Also, if I use the word drift, my definition is a change in resistance. This is what they do. The whole idea is for them to change in resistance. If we use the word drift to mean that they would change their resistance at a given voltage and current to a different resistance at the same voltage and current then we can talk about something undesireable. So considering this is probably what you guys are talking about lets consider this. What exactly could happen for one to drift without its couterpart drifting? Its, again, a voltage divider. So if a shunt LDR's value at one measured point drifts it would have to happen because of the environment and usage. If the environment and usage are the same for both shunt LDRs then why would one change and the other not change? Circuit is the same for them both. So if one changes and the other changes, why would they change differently. So if they dont change differently but they change the same, then what will change for the audience? Its a voltage divider and both shunt LDRs are always equal to each other so if both shunt LDRs change equally there is no change for the audience.
Uriah
Gentlemen, please dont make me consider this as intentionally obstructive 
You contribute nothing positive
Its free to build or not
It is about building Lightspeed
To debate whether it works or not makes no sense
I am willing to go very far to protect a popular thread from going down, make no mistake about it
I really dont like to interfere, but if you push it I will
If you want to take this further, I suggest to do it in http://www.diyaudio.com/forums/anal...-attenuator-preamp-general-2.html#post2060163
You contribute nothing positive
Its free to build or not
It is about building Lightspeed
To debate whether it works or not makes no sense
I am willing to go very far to protect a popular thread from going down, make no mistake about it
I really dont like to interfere, but if you push it I will
If you want to take this further, I suggest to do it in http://www.diyaudio.com/forums/anal...-attenuator-preamp-general-2.html#post2060163