Hi
I have noticed a few circuits in the forum still using 5 volts into the Silonex NSL32SR2. Silonex themselves say: "The applied voltage across the cell: the higher the voltage, the worse the distortion. Thus, when the device is turned hard ON and cell voltage is small, THD will be very low."
The LT 1118 -2.5 , or even a LM317 with R1 and R2 both at 220 ohms will achieve 2.5 volts, don't forget to use 1uf caps on input and output. It seems a pity that 2.5 volts or lower, has not been accepted as mandatory for high fidelity use.
The NSL32SR2 is a current device so be aware to use either series resistors, or current regulation to not exceed 25ma on each device - lower is better to prolong life of the NSL32SR2. Any such circuit controlling current invites individual control of each channels series and shunt element.
Cheers Chris
I have noticed a few circuits in the forum still using 5 volts into the Silonex NSL32SR2. Silonex themselves say: "The applied voltage across the cell: the higher the voltage, the worse the distortion. Thus, when the device is turned hard ON and cell voltage is small, THD will be very low."
The LT 1118 -2.5 , or even a LM317 with R1 and R2 both at 220 ohms will achieve 2.5 volts, don't forget to use 1uf caps on input and output. It seems a pity that 2.5 volts or lower, has not been accepted as mandatory for high fidelity use.
The NSL32SR2 is a current device so be aware to use either series resistors, or current regulation to not exceed 25ma on each device - lower is better to prolong life of the NSL32SR2. Any such circuit controlling current invites individual control of each channels series and shunt element.
Cheers Chris
Is this being used as an optical switch? If it is then there is no distortion, just the time taken in changing from the On and Off position, as in slew rate. This is why microprocessors run commonly at 1.8volts as the slew rate is faster as opposed to 5volt supply. The voltage swing is less and the time taken to move from 0volts to +rail is reduced with the lower voltage swing.
Last edited:
Hi Harleyjon
Nice point with switches, but the circuit i am referring to is the use of NSL32SR2 in a audio L Pad configuation.
Cheers / Chris
Even though we use an overall 5 VDC supply in our LDR based passive preamps, the max absolute voltage seen by the attenuating LDRs is ~3 VDC with the differential voltage varying between ~1.4 VDC and ~1.8 VDC. We also use LDRs as input switches (no relays) and full on they also see about 3.1 VDC incoming with ~1.82 differential. We use resistors and diodes in series with the LDRs to limit the total drop across the LDRs. Distortion has been a non-issue.
I think you will find that the reason micros use lower voltage supplies these days is that the transistor technology has much finer geometry so lower voltages are needed. This also reduces power consumption so heat dissipation is less of a problem. The finer geometry gives the speedup. The lower voltage actually slows things down, other things being equal.JonSnell Electronic said:
Hi
I have noticed a few circuits in the forum still using 5 volts into the Silonex NSL32SR2. Silonex themselves say: "The applied voltage across the cell: the higher the voltage, the worse the distortion. Thus, when the device is turned hard ON and cell voltage is small, THD will be very low."
The LT 1118 -2.5 , or even a LM317 with R1 and R2 both at 220 ohms will achieve 2.5 volts, don't forget to use 1uf caps on input and output. It seems a pity that 2.5 volts or lower, has not been accepted as mandatory for high fidelity use.
The NSL32SR2 is a current device so be aware to use either series resistors, or current regulation to not exceed 25ma on each device - lower is better to prolong life of the NSL32SR2. Any such circuit controlling current invites individual control of each channels series and shunt element.
Cheers Chris
Chris, a few things you may be misinterpreting..
1: in the datasheet they say up to 60V across the cell. I dont recommend it and it would be very high distortion, but its possible. What you quoted had a key word in it. CELL. The cell is the resistive side and is not connected to the power supply positive voltage. The LED side of the LDR is connected to the power supply.
To clear up the voltage across the cell and distortion issue.. I have used my Audio Precision to test signal voltage distortion when passing through a Lightspeed style circuit. 2.5V of signal or lower is best in my measurements.
2: You said 25mA or less. Please less. 20mA, even though the datasheet says its okay, is really pushing it. If you can get 40 ohms on the LDR with only 10mA then dont push it any further. If you need lower than 40 ohms then I advocate parallel LDRs rather than pushing them harder.
I wanted to add that I used to run LDRs at higher voltages into the LED side. This works and sounds great. I have had 4 failures reported running them at 13V. I think lower voltage would have let them live longer but mainly minimum current needed to hit 40 ohms is your best bet for long LDR life. I will now be running them at lower voltages.
I am really really done matching LDRs. I detest the drudgery of it but it had been fun in the past. Not any longer. If a guy wants to build a Lightspeed he has to spend big bucks to buy at least 20 LDRs that then he only uses 4 of. If two years down the road one LDR bites the dust then he has to match again and good luck finding one that matches the other 3. Probably better off matching for a new 4 LDRs. Lets say this guy sells preamps with LDRs inside. Man can you imagine being the customer when "this guy" says "I'm not matching LDRs anymore.?" I can. Not fun for either party.
I am really becoming a fan of shunt attenuators with an LDR as series and discrete resistors as shunt. Now you match inexpensive resistors, or just buy 1% and dont match. Make a switched attenuator and stick a LDR in there as shunt. No more matching. Put the LDR at about 6k and go to the switched attenuator calculator site and get your shunt resistances. Let me be the first to say it sounds great and the distortion is .01% max to much lower depending on volume setting and input voltage.
I have compared this with a Lightspeed, a Lighter Note and basically with itself. By "with itself" I mean that I removed the LDRs and replaced them with the same value Caddocks. The LDR of course smokes it. The Caddock was lifeless. The Lightspeed and Lighter Note both sounded great and the new circuit I was trying, the shunt attenuator, held its own with them.
I advocate this method for a few reasons. We still get he LDR sound and yet we can now have perfect channel balance. In fact we can get such great perfection in matching resistors for shunt and then we can easily get 1% on our LDRs and hopefully better depending on the stability of your power supply and the batch of LDRs you are dealing with that we can begin to use LDR attenuators in balanced circuits and still reap some of the rewards of balanced operation. We can realistically hope for a 40db drop in common mode noise. If we could get ourselves better matching of LDRs (you match them by just dialing them in so you have a lot of control) then we could get better noise rejection but even if we got 20dB of noise rejection it would be beneficial and now we can use LDRs in balanced lines.
If we had used hand matched LDRs in balanced lines and varied their value like in the Lightspeed, man there would be almost no common mode noise rejection at all. It would still work and sound wonderful, but the noise would be there for sure.
Well, I only meant to respond to your note about 2.5V for the power supply but I really got a bit carried away.
One thing I dont understand...I am looking at http://media.digikey.com/pdf/Data S...c PDFs/OptoCouplers - Audio Level Control.pdf specially figure 11.
So at 2.3V through the cell... basically the THD is very low no matter what the attenuation ?
So at 2.3V through the cell... basically the THD is very low no matter what the attenuation ?
No. 2.3V on the control input corresponds to equal LDR resistance in that circuit i.e. attenuation of -6dB. This minimises distortion (assuming the two LDR are identical) because each LDR has the same signal voltage across it so their distortion cancels. The same would be true for any driving arrangement at all; it is the equal LDR resistance (therefore giving -6db attenuation) which matters. Distortion rises for all other values of attenuation because LDRs are nonlinear.
Don't fall into the Chris Daly trap of confusing control voltage with signal voltage.
Don't fall into the Chris Daly trap of confusing control voltage with signal voltage.
Reducing the drop across each LDR reduces their distortion. To your point, if you keep the total drop low enough distortion is not an issue to the listener. I'm being practical here and not trying to incite another endless LDR debate.
On the other hand, if you have a source with a very hot line stage output of say 4-5 volts or even higher, an LDR attenuator probably won't be your best choice.
In my view the voltage drop across the LED (control) end of the LDR isn't relevant. Use only enough voltage differential (current) to get the resistance level you need from the photoresistor end and keep the absolute control voltage as low as you can while doing this.
OK I got it...DCout is actually the V control from figure 10.
"Reducing the drop across each LDR reduces their distortion. To your point, if you keep the total drop low enough distortion is not an issue to the listener. I'm being practical here and not trying to incite another endless LDR debate. "
In my simulations (using Mosfet ) and Vgs (to control Vds) I get a dropout voltage from LDR (led) betwen 1V to 4.8V... below I see this is not important.
"In my view the voltage drop across the LED (control) end of the LDR isn't relevant. Use only enough voltage differential (current) to get the resistance level you need from the photoresistor end and keep the absolute control voltage as low as you can while doing this. "
I am using 5V to feed the LED...any advantages on going lower ?
In the end, I want to have the lowest distortion possible...using current control . What should I concentrate on ? (my source is a 2.1Vrms max DAC)
"Reducing the drop across each LDR reduces their distortion. To your point, if you keep the total drop low enough distortion is not an issue to the listener. I'm being practical here and not trying to incite another endless LDR debate. "
In my simulations (using Mosfet ) and Vgs (to control Vds) I get a dropout voltage from LDR (led) betwen 1V to 4.8V... below I see this is not important.
"In my view the voltage drop across the LED (control) end of the LDR isn't relevant. Use only enough voltage differential (current) to get the resistance level you need from the photoresistor end and keep the absolute control voltage as low as you can while doing this. "
I am using 5V to feed the LED...any advantages on going lower ?
In the end, I want to have the lowest distortion possible...using current control . What should I concentrate on ? (my source is a 2.1Vrms max DAC)
There is nothing to concentrate on regarding distortion. You will get whatever distortion level your LDRs give with the signal voltage you put across them.
Most LDR fans concentrate on achieving a smooth volume control range, or near-constant input or output impedances, as they are things they can improve. Whatever control arrangements you use will eventually simply put a current through some LEDS, which will then shine some light onto some LDRs to set their resistances. The same current will give the same light and the same LDR resistance (and the same LDR distortion) however you arrived at the current.
Most LDR fans concentrate on achieving a smooth volume control range, or near-constant input or output impedances, as they are things they can improve. Whatever control arrangements you use will eventually simply put a current through some LEDS, which will then shine some light onto some LDRs to set their resistances. The same current will give the same light and the same LDR resistance (and the same LDR distortion) however you arrived at the current.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.