AHA. George - You are correct -
I noticed on my connections there is a channel miss-wire - I had left and right channels switched - so the whole center imaging being "gone" could be explained by the channels not being wired correctly -
Me-so-we- todd-ed. Sofa-king-we-todd-ed.
Anyway, can't really do any critical listening since the kids are in bed...but tomorrow I will have a chance to listen in...taking the ENTIRE day off.
Every other Friday is "Tinker Day" for me.
I noticed on my connections there is a channel miss-wire - I had left and right channels switched - so the whole center imaging being "gone" could be explained by the channels not being wired correctly -
Me-so-we- todd-ed. Sofa-king-we-todd-ed.
Anyway, can't really do any critical listening since the kids are in bed...but tomorrow I will have a chance to listen in...taking the ENTIRE day off.
Every other Friday is "Tinker Day" for me.
from my childs for heavy diving into my tinker room I finished my Lightspeed the late afternoon.
LDRs for valve preamp
Hi George,
My name's Andy. I'm Melbourne-based, but just came back here from two years in Sydney, so I'm almost a local up there.
About two weeks ago I finally got my finger out and decided to try something I've been thinking about for years - adding an LDR attenuator to my valve pre-amp in place of the wire-wound linear pot that was there. This was inspired by memories of an old TAS or Stereofool review of a Melos preamp - common knowledge I now discover, but I never heard anyone else talk about it. At this stage, I knew nothing of your work. I also decided to use LEDs as the light source, because I could run them from batteries without lifespan, current-draw or heat limitations associated with incandescent light sources. Adding even more power-supplies to the plethora of un-matching boxes I already have is very unattractive to me. Also, remote volume control (wired, not wireless) was an integral part of the plan, as I figured I could just control the LEDs remotely via a 3metre cable and carbon pots.
So, off I went to Jaycar Electronics and got a handfull of LEDs and LDRs and set to work. Talk about rake's progress!! I discovered the hard way a whole lot of things that you have already discovered - finding matching units is a bear of a task, they drift if you get cheap ones, they're not all that well suited to the demands of a valve preamp attenuator, and they sound absolutely to die for wonderful. The Jaycar ones, at least, are also inclined to go noisy when dark, I don't know why.
The volume control in my preamp (a very souped-up dual-mono Trevor Lees) is unusual in that is inter-stage, between the last amplification stage and the cathode follower. This is a very nice idea, as it means it's the cathode follower stage directly driving the interconnects and power-amp, rather than through the attenuator pot. However, it adds extra complexity to implementing an LED/LDR attenuator. It requires a capacitor to block 200V dc going to earth and at least 100K Ohm constant(ish) overall attenuator impedance. I have found that I can get an impedance varying from 200 Ohms to about 500K Ohms from the best selected Jaycar LDRs lit by LEDs running 3.2V DC thru a 100K linear carbon pot and 300 Ohm fixed (current-limiting) resistor. But it's driving me mad trying to get them to track and stay stable. I can get much higher values (eg 10 M Ohms) in almost dark, but they are hopelessly variable unit-to-unit in this part of their curves.
Anyway, I recently became aware of your own work via the DIYAudio website. I realise you make complete Lightspeed Attenuators on a commercial basis, but they're really not suitable for my needs. So, anyway, if you would consider it, I'd be prepared to pay very fairly for two matched pairs of LED/LDR units selected by yourself (i.e. two channel's worth).
They'd have to provide more or less constant 100K combined impedance (higher is OK), because the 12AX7 valves I use won't drive less, and I will have to mess about myself with current-limiting resistors and/or trim pots to try to get them to work on AA batteries which will give me any multiple of 1.6 V (new batteries) e.g 1.6V 3.2V 4.8V 6.4V etc.
If you don't think the Silonex units will do the job, just let me know please and I'll forge on alone!! I have to say, I've heard pretty much every type of attenuator out there, and LDRs are the bees knees. Possibly Stevens and Billington tranformers are better at low volume, where they are uncanny, but everywhere else, LDRs rule.
Thanks for your time.
Andy
Melbourne
Australia
Hi George,
My name's Andy. I'm Melbourne-based, but just came back here from two years in Sydney, so I'm almost a local up there.
About two weeks ago I finally got my finger out and decided to try something I've been thinking about for years - adding an LDR attenuator to my valve pre-amp in place of the wire-wound linear pot that was there. This was inspired by memories of an old TAS or Stereofool review of a Melos preamp - common knowledge I now discover, but I never heard anyone else talk about it. At this stage, I knew nothing of your work. I also decided to use LEDs as the light source, because I could run them from batteries without lifespan, current-draw or heat limitations associated with incandescent light sources. Adding even more power-supplies to the plethora of un-matching boxes I already have is very unattractive to me. Also, remote volume control (wired, not wireless) was an integral part of the plan, as I figured I could just control the LEDs remotely via a 3metre cable and carbon pots.
So, off I went to Jaycar Electronics and got a handfull of LEDs and LDRs and set to work. Talk about rake's progress!! I discovered the hard way a whole lot of things that you have already discovered - finding matching units is a bear of a task, they drift if you get cheap ones, they're not all that well suited to the demands of a valve preamp attenuator, and they sound absolutely to die for wonderful. The Jaycar ones, at least, are also inclined to go noisy when dark, I don't know why.
The volume control in my preamp (a very souped-up dual-mono Trevor Lees) is unusual in that is inter-stage, between the last amplification stage and the cathode follower. This is a very nice idea, as it means it's the cathode follower stage directly driving the interconnects and power-amp, rather than through the attenuator pot. However, it adds extra complexity to implementing an LED/LDR attenuator. It requires a capacitor to block 200V dc going to earth and at least 100K Ohm constant(ish) overall attenuator impedance. I have found that I can get an impedance varying from 200 Ohms to about 500K Ohms from the best selected Jaycar LDRs lit by LEDs running 3.2V DC thru a 100K linear carbon pot and 300 Ohm fixed (current-limiting) resistor. But it's driving me mad trying to get them to track and stay stable. I can get much higher values (eg 10 M Ohms) in almost dark, but they are hopelessly variable unit-to-unit in this part of their curves.
Anyway, I recently became aware of your own work via the DIYAudio website. I realise you make complete Lightspeed Attenuators on a commercial basis, but they're really not suitable for my needs. So, anyway, if you would consider it, I'd be prepared to pay very fairly for two matched pairs of LED/LDR units selected by yourself (i.e. two channel's worth).
They'd have to provide more or less constant 100K combined impedance (higher is OK), because the 12AX7 valves I use won't drive less, and I will have to mess about myself with current-limiting resistors and/or trim pots to try to get them to work on AA batteries which will give me any multiple of 1.6 V (new batteries) e.g 1.6V 3.2V 4.8V 6.4V etc.
If you don't think the Silonex units will do the job, just let me know please and I'll forge on alone!! I have to say, I've heard pretty much every type of attenuator out there, and LDRs are the bees knees. Possibly Stevens and Billington tranformers are better at low volume, where they are uncanny, but everywhere else, LDRs rule.
Thanks for your time.
Andy
Melbourne
Australia
Hi,
could variable battery voltage into the LEDs be part of the longer term stabilty problem?
The LDRs seem to be becoming difficult to match at very high impedance.
Some builders are also using LDRs that are higher impedance than necessary for this project.
Has anyone looked at using a resistor bypass across each LDR to reduce the maximum impedance?
This is often used to help linearise analogue sensors (thermistors), it might help here. eg, we want 100k max in the LDRs and find they go to 500k. Add a 120k or 130k in parallel and the max is now near 100k.
Not only that but as the LDR impedance reduces the bypass resistor has less effect, gradually becoming insignificant when LDR is down to <10k. This would allow more precise matching at low impedance and accept a slightly wider tolerance at higher impedance where the bypass starts to swamp the matching error.
could variable battery voltage into the LEDs be part of the longer term stabilty problem?
The LDRs seem to be becoming difficult to match at very high impedance.
Some builders are also using LDRs that are higher impedance than necessary for this project.
Has anyone looked at using a resistor bypass across each LDR to reduce the maximum impedance?
This is often used to help linearise analogue sensors (thermistors), it might help here. eg, we want 100k max in the LDRs and find they go to 500k. Add a 120k or 130k in parallel and the max is now near 100k.
Not only that but as the LDR impedance reduces the bypass resistor has less effect, gradually becoming insignificant when LDR is down to <10k. This would allow more precise matching at low impedance and accept a slightly wider tolerance at higher impedance where the bypass starts to swamp the matching error.
Andy, great to see your trying to use other than the NSL32SR2S 's but your fighting a loosing battle, especially trying to drive it with a 12AX7, you need a source I say lower than 100R to drive this thing.
My first one I built was way back in the 70's and I got the LDR's out of camera lightmeters that were unrepairable and I put them, one in each end of a tube and one led in a hole drilled in the middle of that tube, so that one led controlled two LDR's but back then the LDR's were shocking, drift and matching, I have it easy now compared to back then.
Also I cannot sell any of my matched quads as I've been asked this countless times, matching the LDR's and calibrating each Lightspeed Attenuator is 80% of the labour involved in building each one. Would you pay 80% of $500 for 4 matched LDR's? I think not, and I wouldn't sell it.
BTW I got a quote from a suppier in Australia for the NSL32SR2S below is the return quote.
Quote>
Hello George,
Thanks you for the feedback.
I have spoken to management and below is what we can offer.
100pcs @ $2.50+GST ea.
We can post it out, but we have no means of tracking/tracing goods in case they get lost.
Our prefered method is Registered( <$5) or Express post ( ~$7)
As far as lead time, we initially quoted approx lead times given to us by the factory.
We courier these in by TNT/DHL, which take less than a week after they factory is ready to ship these out.
Even if Allied did not have stock, they will be facing identical lead times to us.
( If factory have stock, we would probably have these here in a weeks time.)
We accept Mastercard/Visa /Direct Deposit or cheque.
I hope this will help persuade you. If you have any further questions, please do not hesitate to contact me.
Regards
Tony
Quote<
Cheers George
My first one I built was way back in the 70's and I got the LDR's out of camera lightmeters that were unrepairable and I put them, one in each end of a tube and one led in a hole drilled in the middle of that tube, so that one led controlled two LDR's but back then the LDR's were shocking, drift and matching, I have it easy now compared to back then.
Also I cannot sell any of my matched quads as I've been asked this countless times, matching the LDR's and calibrating each Lightspeed Attenuator is 80% of the labour involved in building each one. Would you pay 80% of $500 for 4 matched LDR's? I think not, and I wouldn't sell it.
BTW I got a quote from a suppier in Australia for the NSL32SR2S below is the return quote.
Quote>
Hello George,
Thanks you for the feedback.
I have spoken to management and below is what we can offer.
100pcs @ $2.50+GST ea.
We can post it out, but we have no means of tracking/tracing goods in case they get lost.
Our prefered method is Registered( <$5) or Express post ( ~$7)
As far as lead time, we initially quoted approx lead times given to us by the factory.
We courier these in by TNT/DHL, which take less than a week after they factory is ready to ship these out.
Even if Allied did not have stock, they will be facing identical lead times to us.
( If factory have stock, we would probably have these here in a weeks time.)
We accept Mastercard/Visa /Direct Deposit or cheque.
I hope this will help persuade you. If you have any further questions, please do not hesitate to contact me.
Regards
Tony
Quote<
Cheers George
AndrewT said:
Definatly worth looking into this Andrew, good lateral thinking, I must find the time to R&D this.
Cheers George
Try R3S Andy
I am on vacation but can check if 100K is obtainable from a higher resistance of LDR. I did build one with 20K impedance in front of an active linestage. From fading memory, 100K pot with R3 LDR yielded 20 - 24K.
In fact, I think the R3 type most likely would hit a 100K with 500K controls. The R3 should also be a little less touchy at this extremely low LED current level. But all types are temperature sensistive and exhibit high hysteresis effects. That is the resistance is different depending if the current is stepped up to a level, or stepped down to that level. Overtime, the two come together.
Another Andy (A THIRD ONE) was using R3 LDR's. He may have tried to get higher impedances.
I got mine by accident. I do not know if Allied mis-shipped, or I misordered, but all the ones I started with were R3. Sorted R3 which I never saw listed at Allied.
But they might have applications also. Generally they run double or a little higher resistance compared to R2.
George
I am on vacation but can check if 100K is obtainable from a higher resistance of LDR. I did build one with 20K impedance in front of an active linestage. From fading memory, 100K pot with R3 LDR yielded 20 - 24K.
In fact, I think the R3 type most likely would hit a 100K with 500K controls. The R3 should also be a little less touchy at this extremely low LED current level. But all types are temperature sensistive and exhibit high hysteresis effects. That is the resistance is different depending if the current is stepped up to a level, or stepped down to that level. Overtime, the two come together.
Another Andy (A THIRD ONE) was using R3 LDR's. He may have tried to get higher impedances.
I got mine by accident. I do not know if Allied mis-shipped, or I misordered, but all the ones I started with were R3. Sorted R3 which I never saw listed at Allied.
But they might have applications also. Generally they run double or a little higher resistance compared to R2.
George
First report
This evening I had a small listening session with Diana Krall and Holly Cole.
The counterparts: Lightspeed MK II and Tolu Opamp Preamp
The amp: Atoll AM 100
The speakers: Tolero with Eton 9"-bass and 6"-mid and ER4 for tweeter
I am very confident with the Lightspeed because matching was a full evening job but I think I did it well.
Conclusion:
The LS played well at all. The spatiality was impressing. Mids and bass were good and appropriate. Voices are paramount.
The Tolu Preamp: He gave more details than the Lightspeed. If you know "the girl in the other room" from Diana Krall, there are many cymbals, brushed hi hats, bells etc. With the opamp pre every little detail was hearable. The well-known curtain was blown away. But these many details let sometimes miss the complete impression at all.
What I liked with the Lightspeed is the laid-back atmosphere. He isn't a nitpicker but a gentleman. He plays with maturity and dignity. I like that. But I cannot say he is the winner so far if it is even possible to judge black or white.
I would say for the moment: tie (with little advantage for opamp)
Well, wait for next game.
Regards
Thomas
This evening I had a small listening session with Diana Krall and Holly Cole.
The counterparts: Lightspeed MK II and Tolu Opamp Preamp
The amp: Atoll AM 100
The speakers: Tolero with Eton 9"-bass and 6"-mid and ER4 for tweeter
I am very confident with the Lightspeed because matching was a full evening job but I think I did it well.
Conclusion:
The LS played well at all. The spatiality was impressing. Mids and bass were good and appropriate. Voices are paramount.
The Tolu Preamp: He gave more details than the Lightspeed. If you know "the girl in the other room" from Diana Krall, there are many cymbals, brushed hi hats, bells etc. With the opamp pre every little detail was hearable. The well-known curtain was blown away. But these many details let sometimes miss the complete impression at all.
What I liked with the Lightspeed is the laid-back atmosphere. He isn't a nitpicker but a gentleman. He plays with maturity and dignity. I like that. But I cannot say he is the winner so far if it is even possible to judge black or white.
I would say for the moment: tie (with little advantage for opamp)
Well, wait for next game.
Regards
Thomas
Thanks for help
Thanks everyone for the replies. I've now spent quite a lot more time reading the many, many pages of this thread and realise most of my questions were answered in there somewhere already. Thanks for not flaming me. Maybe some of the really often-asked questions (like, can I purchase just LDRs?) could go into a FAQ sticky at the beginning of the thread - just to save you regular contributors having to field the same questions again and again.
Anyway, I'll get hold of some appropriate LDRs and continue, I'm confident I can make this work to my satisfaction. I have read some very clever ideas here for sophisticated methods of manging non-linearities and tracking, but it seems that I am making my own life comparatively easy because I have no qualms about using a dual-mono setup (i.e. two single-gang volume pots to manage volume and balance simultaneously). I'm used to this set up and it seems to make matching LDRs less of a big deal.
Andy
Thanks everyone for the replies. I've now spent quite a lot more time reading the many, many pages of this thread and realise most of my questions were answered in there somewhere already. Thanks for not flaming me. Maybe some of the really often-asked questions (like, can I purchase just LDRs?) could go into a FAQ sticky at the beginning of the thread - just to save you regular contributors having to field the same questions again and again.
Anyway, I'll get hold of some appropriate LDRs and continue, I'm confident I can make this work to my satisfaction. I have read some very clever ideas here for sophisticated methods of manging non-linearities and tracking, but it seems that I am making my own life comparatively easy because I have no qualms about using a dual-mono setup (i.e. two single-gang volume pots to manage volume and balance simultaneously). I'm used to this set up and it seems to make matching LDRs less of a big deal.
Andy
Tolu said:
do you have a recent circuit diagram for your op amp pre amp ?
what op amp are you using ?
mike
Hi Tolu,
how about reporting the effect of adding 100nF from pin7 to pin4?
and whether lowK ceramic (NP0 C0G) works as well as film in it's various guises?
how about reporting the effect of adding 100nF from pin7 to pin4?
and whether lowK ceramic (NP0 C0G) works as well as film in it's various guises?
AndrewT said:
What effect on sound would you expect with the C between V+ and V-? It's a little bit inconvenient, isn't it?
I haven't tried various part qualities so far. Do you have some good proposals for replacement? (Manufacturer, part #)AndrewT said:
The PSU has much room for improvement as well (7815, 7915, cheap Cs). The attenuation is at the moment a cheap Alps RK09.. because I burned down my blue Alps pot. Relay control would be next step because I like R/C!
One research goal is the improvement of spatiality. Perhaps a relay attenuator with 1% or 0,1% Rs will get it.
Tolu said:Hi Mike
here it is. I am still in research state but I am almost very content with it. It hasn't to be always discrete...
Regards
Thomas
Are you inserting the LS at the input or output of the opamp?
Will said:
Whether...nor!
I tried them separately. The LS as passive preamp versus my opamp pre! The opamp pre has an own pot! But perhaps it is a good idea to try the LS as attneuator for the opamp pre!
Higher resistance vol control...
Read here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=97944&highlight=
Gajanan Phadte
Read here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=97944&highlight=
Gajanan Phadte
AndrewT said:
mount the rail to rail cap directly on the pin7 & pin4 on the back of the chip i.e. on the copper side and lay the cap flat against the PCB with the body of the cap as near the middle of the chip as possible, not off to one side and NOT outside the profile of the DIL pins.Tolu said:
It should help to remove glitches in the power supply from affecting nearby chips/circuits and it should help meet transient current demand inside the chip. This could help clean up transients and help treble clarity. i.e. it should sound nice with no sibilance (we hope).
Tolu :
Have you tried the LDRs as volume control for the opamp ? (in replacement of your Alps)
By the way, why series/shunt instead of plain shunt, which from the Silonex paper seems to give less distortion ?
Have you tried the LDRs as volume control for the opamp ? (in replacement of your Alps)
By the way, why series/shunt instead of plain shunt, which from the Silonex paper seems to give less distortion ?
peufeu said:
I've A/B'ed both with freinds peufeu, and series/shunt blows shunt away, it's more dynamic (don't ask me why) and sounds tighter and more extended, it does have a much more constant input/output impedeance compared to just shunt, maybe, who knows?
Cheers George