A precision LED/LDR-based Attenuator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Fine-tuning the website is ongoing, it'll probably never stop -- I'm an engineer who needs a manager to tell him when to stop, and I don't have a manager. :)

I am NOT saying the LDRs will never need calibration, I am saying that I myself have never had to recalibrate my LDRs after the first calibration. YMMV, and it probably will.

However, I do believe that my system has advantages over Lightspeed-type circuits that make mine more stable both in the short term and long term.

Short term stability is enhanced because I control LED current, not voltage, and LEDs are current-driven devices. Any voltage-control method is bound to be less effective because temperature changes at the LED junction affect the amount of current passing through the LED even if voltage is held constant. This changes brightness and that changes LDR resistance even if voltage is held steady.

Long term, my circuit will remain more stable because current through the LDRs is strictly limited to 10 milliamps or less, which is half of the device's published rating of 20~25 milliamps maximum. BTW, it has been recently suggested in another thread that PIC-controlled systems can damage LDRs by delivering excessive current to make unmatched LDRs match up, but the reverse is true -- I control the current down to the micro-amp level, and maximum current is never more than 10.005 milliamps, and almost always far, far less! Anyway, a 10 milliamp drive to a 20 milliamp-rated LED is not going to hurt, or even stress it.

As an aside, for folks who have an LDR preamp based on the voltage-controlled circuit with 100 ohm trimmer I've seen published in various threads, you might want to do the math. The LDR LED is rated at 25 milliamps at 2.5V drive, and that means with a 5V supply the trimmer must drop 2.5 volts. Dropping 2.5 volts through 100 ohms will deliver 25 milliamps which I believe is too much and, further, if you reduce the resistance of the 100 ohm trimmer to balance the channels as the instructions direct, the trimmer will now pass more than 25 milliamps at 2.5 volts and that seems to me to be a real no-no. This circuit should have something like a 125~150 ohm fixed resistor and a trimmer that is set to zero and then increased to balance channels. This approach will actually protect your LDRs. However, driving the LDRs conservatively in this circuit may not give you much attenuation, and probably that's why the max current is set so high, and this may be why LDRs have the reputation of changing over time.

In any case, I believe that because my circuit allows the LDRs to loaf along at 10ma maximum, there isn't much change in the LDR responses over time.

I am not in doubt about your system, was a believer since you started this thread, and increasingly impressed with the care you have built this module. I guess I was just hoping I could grab this to test if I liked it. Now I must be very sure, because for this kind of money I need to be very sure, as I have to save up for months. I currently run a DCB1 homebuilt on veroboard that has cost me perhaps 50 bucks in total ( yes I was using some scrap parts ) and I like it very much. But hey, there is always something better to persue.

Did not mention it before, but love your choice of display. We have OLED displays in our weighing terminals and the visibility is awesome. Also less noise compared to a VFD.

Kind regards,

Jslot
 
I am not in doubt about your system, was a believer since you started this thread, and increasingly impressed with the care you have built this module. I guess I was just hoping I could grab this to test if I liked it. Now I must be very sure, because for this kind of money I need to be very sure, as I have to save up for months. I currently run a DCB1 homebuilt on veroboard that has cost me perhaps 50 bucks in total ( yes I was using some scrap parts ) and I like it very much. But hey, there is always something better to persue.

Did not mention it before, but love your choice of display. We have OLED displays in our weighing terminals and the visibility is awesome. Also less noise compared to a VFD.

Kind regards,

Jslot

You made me curious and I did some reading about the DCB1, finding Nelson Pass's pdf on building the original b1 and how it sounds. I guess it sounds pretty marvelous? So simple, and no NFB. . .

I completely understand your questioning the cost of my system. Perhaps you'll waint to wait until a few people have received a copy and have commented favorably here before you decide to commit?

I designed the encoder and IR receiver and the display because they were capabilities that I would want for my own preamp, but they are not necessary to enjoy the sound. Much of the cost of my system is in those optional parts, none of which is required to operate the LDR control board so my LDR system doesn't need to be that expensive.

If the sound of LDRs is your main goal, then consider buying only the LDR control board, or just build yourself a Lightspeed circuit on perf board, maybe using sockets for the LDRs as I have done, and a slightly more conservative current limiting arrangement than the original Lightspeed circuit.

I myself have an application in mind that will require only the LDR board: I recently bought a second-hand Monsoon computer speaker system which has planar magnetic flat-panel satellites and a dynamic subwoofer. They sound terrific for computer speakers, and are designed to play with all computer volume levels set to maximum and volume controlled with a potentiometer and mute button wired into the cable between the soundcard output and the amplifiers in the subwoofer box.

A few months ago I learned my lesson regarding leaving computer volume controls at maximum and controlling volume after the DAC for best sound, so I'm going to replace that potentiometer with one of my LDR boards. I expect that little update to make a big difference to the sound of these planar speakers. This is a long-winded way of saying that there are a number of applications where the LDR board can be used by itself and maybe your dcb1 would benefit from just the LDR board?
 
Since you're calling this attenuator a passive preamp on your website, are you suggesting using only this setup between the source and power amp and no active preamp?

Thanks...

That's the way I'm using it in my system:

netbook -> USB -> Schiit DAC -> precisionLDR -> chipamp -> modified Klipsch Forte.

Right now, I don't use multiple sources. When I pulled my old preamp (high end Yamaha) and inserted the precisionLDR, I was truly stunned by the improvement in sound -- it was as though I'd stepped my whole system up several notches.

You can use the precisionLDR as a potentiometer in any other setup, maybe with DCB1 buffer or another buffer that James Hill mentioned a while ago (James, what was that?), or inserted in a commercial preamp or amp, it's really up to you -- this is a pure potentiometer but with amazing sound. You can use it any way you'd like.

I have come to believe that most audio systems have too much gain, and removing the gain of a conventional premap was a good thing for me -- I now listen with volume control at 9 o'clock for soft and 3 o'clock for blow-my-eardrums-out sound with LM3886 chipamp and fairly efficient speakers whereas in the past my volume control never got past 10 o'clock.

The purpose of the RE/IR board and display is to provide all the tools of a preamp (including, eventually, source switching via an LDR-based input selector).

So, I believe calling it a passive preamp is reasonable, but also it can be thought of as a plain potentiometer (with balance control built in) that you can insert in any configuration of active preamp.
 
I've tried the DC B1 buffers (with the Salas shunt reg 'family'), 'keantoken's' transistor based buffer and the Aikido valve buffer (with & without the gain stage) and a few other valve buffers.

Each adds it's own particular characteristic that varies with the power amp - the 'Kuartlotron' is the most transparent - as you can imagine, some amps seem to work really well with no buffer at all, like the 'loaner' 845 valve amp but the DAO headamp works best with the ***** Aikido buffer - horses for courses and all that ...
 
I think I've solved the question of how to integrate Left & Right Balance control into an inexpensive remote control -- any old Sony-compatible, or very simple universal, TV/VCR remote should work.

I have programmed the Left - Right balance as the Sony specification requires for actual amplifier/receiver remotes, but also added the codes for VCR Rewind for Balance Left and Fast Forward for Balance Right.

That way, the ultra-cheap 3-device universal remote I'm using as a testbed ($5.50 from Amazon, an RCA unit) can be programmed for Sony TV and Sony VCR, and the TV controls Volume Up and Down and Mute and, eventually, will control source select when that option becomes available. By programming the lower VCR keys for Sony also, the left arrow becomes Balance Left and the right arrow becomes Balance Right. So now a complex Sony amplifier remote is no longer required to control balance from the listening position.

When the RE/IR is set to Volume, pressing a Balance button will automatically switch it over to Balance mode and, conversely, pressing a Volume button or Mute will cause the control to revert to Volume mode.

I am now comfortable that $6 will get you a remote that will control all functions of my system. If you're using my LDR preamp with a Sony VCR, you'll have a problem, but does anybody use a VCR anymore? And if you do, hopefully it is not a Sony.
 
Last edited:
But keep in mind that the RE/IR board will work with any remote that is a dedicated Sony or a universal remote that can be programmed for Sony TV and VCR codes.

I chose the one I did because it was inexpensive, small, and simple. You could easily integrate this LDR control using a sophisticated universal remote that could make it play nice with all of the other components in your system as the main volume control for your entire system.
 
You made me curious and I did some reading about the DCB1, finding Nelson Pass's pdf on building the original b1 and how it sounds. I guess it sounds pretty marvelous? So simple, and no NFB. . .

I completely understand your questioning the cost of my system. Perhaps you'll waint to wait until a few people have received a copy and have commented favorably here before you decide to commit?

I designed the encoder and IR receiver and the display because they were capabilities that I would want for my own preamp, but they are not necessary to enjoy the sound. Much of the cost of my system is in those optional parts, none of which is required to operate the LDR control board so my LDR system doesn't need to be that expensive.

If the sound of LDRs is your main goal, then consider buying only the LDR control board, or just build yourself a Lightspeed circuit on perf board, maybe using sockets for the LDRs as I have done, and a slightly more conservative current limiting arrangement than the original Lightspeed circuit.

I myself have an application in mind that will require only the LDR board: I recently bought a second-hand Monsoon computer speaker system which has planar magnetic flat-panel satellites and a dynamic subwoofer. They sound terrific for computer speakers, and are designed to play with all computer volume levels set to maximum and volume controlled with a potentiometer and mute button wired into the cable between the soundcard output and the amplifiers in the subwoofer box.

A few months ago I learned my lesson regarding leaving computer volume controls at maximum and controlling volume after the DAC for best sound, so I'm going to replace that potentiometer with one of my LDR boards. I expect that little update to make a big difference to the sound of these planar speakers. This is a long-winded way of saying that there are a number of applications where the LDR board can be used by itself and maybe your dcb1 would benefit from just the LDR board?

I built the Salas version DCB1 with shunt regulator, running at 400mA. It is very sensitive to the potentiometer used, more than any preamp I've used before, so that is why I think this is the point the amp can even be improved by using NO potentiometer. I use a Cosmos Tocos 10 pot that sounds fabulous, but does not track very well at low volume positions. My Alps 10 tracks very well, but sounds not quite as good as the Tocos. Compromises...

Using only the volume control board for starters would be an option I would certainly consider! What would the price be for this board? And do I need more ( rotary encoder, something to read remote control codes ) to really work with it? In short, what would my starting price be?

Those monsoon speakers, are they using NXT? Was always curious to hear that technique. Mission and Magnat had some speakers based on that technology, but they are no longer available. So there must have been something not selling well.
 
I built the Salas version DCB1 with shunt regulator, running at 400mA. It is very sensitive to the potentiometer used, more than any preamp I've used before, so that is why I think this is the point the amp can even be improved by using NO potentiometer. I use a Cosmos Tocos 10 pot that sounds fabulous, but does not track very well at low volume positions. My Alps 10 tracks very well, but sounds not quite as good as the Tocos. Compromises...

Using only the volume control board for starters would be an option I would certainly consider! What would the price be for this board? And do I need more ( rotary encoder, something to read remote control codes ) to really work with it? In short, what would my starting price be?

Those monsoon speakers, are they using NXT? Was always curious to hear that technique. Mission and Magnat had some speakers based on that technology, but they are no longer available. So there must have been something not selling well.


LDR respons is logarithmic, so very hard to control resistance at 100K where current draw is about 5 microamps, and very easy at 50 ohms where current draw is maybe 5 milliamps. Thus I can control my LDRs to about .2 ohms at 50 ohms which is where they will be at low levels. Subsequently, low level tracking should be very good. I really haven't noticed any audible difference, I think the tracking is good at any level. The LDR board has only potentiometer control. You can do other things like get a different remote control board because all you need to do is deliver 0~5VDC to the pot terminal to drive the board from muted up through almost zero attenuation. Such a solution would not allow you to use the OLED display which needs the RE/IR/Display board to function.

With regard to how you want to proceed with buying something, please wait for a short while longer until the website has that information, it should come pretty clear. This is not a commercial thread and it needs to stay in the technical realm.

The Monsoons are planar magnetics, like Magnepans, I don't think NXT technology applies.
 
Last edited:
Did some final fixes --

1. Fixed the 'screensaver' so that instead of going to the smaller size number and jumping randomly around the screen, the Volume number now stays at the full two-line height and moves sedately back and forth between the left and right edges of the screen. I like this much more than the earlier version because now I can read the Volume from a good distance even when the screen-saver and dimming are doing their thing.

2. Made the milliamp readout values slightly more accurate. Not a big improvement but, hey, any improved accuracy is good, right?

I think I'm done with engineering changes! :D
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.