Badge, your problem is the Marchand, at 25k together with your amp 100k you have a total load going back to the Lightspeed of 20k, and with the Lightspeeds approx 7k your cd player is seeing 5.2k this is OK so long as your cd player or dac has a very good output stage, and judging by your remarks about the bass it maybe hasn't. What is your cd player/dac?
Also there are a couple of fixes for this.
One is to raise the input impedence of the Marchand (the easiest) and will help, I don't know why Phil Marchand said it wouldn't.
Two is to get a lower impedence/stronger output on the cd player/dac.
Three is to put an opamp buffer on the input of the Marchand.
Also the interconects between.
Source and Lightspeed ,
Lightspeed and poweramp,
Lightspeed and Marchand,
Marchand and poweramp
must be kept low in capacitance and resistance as you have lots of them in your setup.
In my own bi-amped setup I'm driving two stereo amps of 68k each (34k) total with no problems at all with stunning impactfull tight bass down to 20hz, but I have a dac with a strong output.
Cheers George
Also there are a couple of fixes for this.
One is to raise the input impedence of the Marchand (the easiest) and will help, I don't know why Phil Marchand said it wouldn't.
Two is to get a lower impedence/stronger output on the cd player/dac.
Three is to put an opamp buffer on the input of the Marchand.
Also the interconects between.
Source and Lightspeed ,
Lightspeed and poweramp,
Lightspeed and Marchand,
Marchand and poweramp
must be kept low in capacitance and resistance as you have lots of them in your setup.
In my own bi-amped setup I'm driving two stereo amps of 68k each (34k) total with no problems at all with stunning impactfull tight bass down to 20hz, but I have a dac with a strong output.
Cheers George
George, again thanks for all your help. I will raise the input impedance on the crossover. I am also using a back up SACD player which is a modified transformer output stage Sony SACD 775. My regular unit is a Sony SACD 9000 ES with Modwright signature truth modifications which includes a tube output stage and more drive output than the 775. The wire umbilical cord that ties the tube output stage to its own power supply is off to cryogenics at the moment.
Badge, I maybe wrong with yours, but a source that has a transformer or a tube coupled output stage, will not have a low enough impedence and strong output to drive into such a low load as you have. I hope I'm wrong for your sake.
Also the tube output stage will have capacitor coupling to the outside world, make sure that cap is at least 10uf.
Cheers George
Also the tube output stage will have capacitor coupling to the outside world, make sure that cap is at least 10uf.
Cheers George
Same here
Badge,
I get about the same overall loading on my sources, 8K ll 100K. The caps in my tuner are a little too small, the ones in my phono and SA-8260's are fine.
Waiting on a box of 100K limear dual pots to arrive. This should reiple the impedance of my Lightspeed clone from what it is now.
Also while apart I need to jack up my current limiting resistors. Used 121 ohm. But George is using a single control on his. Using two controls each section of the pot is only controlling one LDR. The resistors in mine need to be 200 ohms.
Attracted to this design for use as an attenuator in front of a buffer. But thought it should be tried first as designed. Learned quite a lot about how to rig it. Plus it really sounds nice.
Sometime soon hope one installed in my preamp in front of an AD815 line driver circuit. This will take of any loading issues in my system.
George
Badge,
I get about the same overall loading on my sources, 8K ll 100K. The caps in my tuner are a little too small, the ones in my phono and SA-8260's are fine.
Waiting on a box of 100K limear dual pots to arrive. This should reiple the impedance of my Lightspeed clone from what it is now.
Also while apart I need to jack up my current limiting resistors. Used 121 ohm. But George is using a single control on his. Using two controls each section of the pot is only controlling one LDR. The resistors in mine need to be 200 ohms.
Attracted to this design for use as an attenuator in front of a buffer. But thought it should be tried first as designed. Learned quite a lot about how to rig it. Plus it really sounds nice.
Sometime soon hope one installed in my preamp in front of an AD815 line driver circuit. This will take of any loading issues in my system.
George
One step forward, big one backwards
Tried a couple more things to optimze the unit. I read early in the thread about contact points in the signal path. Since I use multiple sources regularly single input is not an issue.
The case already had a nice vintage ceramic wafer CRL rotary 2 pole. Had a four pole identical switch in the closet so replaced the dual with a quad switch. This switch is used ala Croft, five position, with pos 2 and 4 as mutes. This puts a ground between each input and resduces channel leakthough. Also doubled up the contacts so that each input and the outputs have two high quality contacts.
The switch change is audible. Seemed to even improve bass weight, but cannot see how. Anyway one forward.
Then replaced the dual 50K linear pots with 100K linear pots. This changed my measured input impedance from 8 -9K to 22K - 30K depending on pot position. The impedance is lower in the middle, higher at the extremes of rotation.
This made the Lightspeed clone sound more like the passives tried before. Colored the sound, rounded the transients. It made it smoother, but did not add the fullness to the bottom expected.
It did really help with volume adjustment. With the 50K all listening was in first 20 -30 degrees of rotation. Now normal is 30%, insame levels will hasve to wait for beer!
So after several hours of solderind: the 5 position, 4 pole input switch is good. The 100K dual linear pots are getting removed. Good to look for the 100
Tried a couple more things to optimze the unit. I read early in the thread about contact points in the signal path. Since I use multiple sources regularly single input is not an issue.
The case already had a nice vintage ceramic wafer CRL rotary 2 pole. Had a four pole identical switch in the closet so replaced the dual with a quad switch. This switch is used ala Croft, five position, with pos 2 and 4 as mutes. This puts a ground between each input and resduces channel leakthough. Also doubled up the contacts so that each input and the outputs have two high quality contacts.
The switch change is audible. Seemed to even improve bass weight, but cannot see how. Anyway one forward.
Then replaced the dual 50K linear pots with 100K linear pots. This changed my measured input impedance from 8 -9K to 22K - 30K depending on pot position. The impedance is lower in the middle, higher at the extremes of rotation.
This made the Lightspeed clone sound more like the passives tried before. Colored the sound, rounded the transients. It made it smoother, but did not add the fullness to the bottom expected.
It did really help with volume adjustment. With the 50K all listening was in first 20 -30 degrees of rotation. Now normal is 30%, insame levels will hasve to wait for beer!
So after several hours of solderind: the 5 position, 4 pole input switch is good. The 100K dual linear pots are getting removed. Good to look for the 100
vatofale said:
For memory I think the first breakup price was 10, that should see you out of trouble, you save a dollar each maybe, i always buy in lots of 50 or 100, you get a much better price that way.
Which just reminds me we have company RS Components here in Australia, and I wanted to get some quarter watt Vishay resistors from them, get this, the price was $20 each if you buy singles, but if you bought 50 it was $30 for the 50, go figure.
Have fun building it, if you follow my instructions to the letter, you will love the sound also, get this a member of DIY flew to Australia from Singapore to buy one in person then few back that day, talk about on a mission, I hope he posts his findings in this thread soon.
Cheers George
The real thing
Thanks George,
I have my new lightspeed installed and having change my Naim NAP-160 power amp to 100K input impedance as suggested by you the results are simply stunning with the Meridian G-08 > lightspeed > NAP-160 > Naim SBL speakers every thing has opened up with great bass, well balance mids and silky smooth highs not something I would have ever believed possible using Naim products. Just goes to show how much pre-amps alter the sound from the CD Players.
Taking into account the number of LDR's needed to match 4 and the time it must take to match and assemble the finished product its a steal at the price your charging.
Cheers
Alan
Thanks George,
I have my new lightspeed installed and having change my Naim NAP-160 power amp to 100K input impedance as suggested by you the results are simply stunning with the Meridian G-08 > lightspeed > NAP-160 > Naim SBL speakers every thing has opened up with great bass, well balance mids and silky smooth highs not something I would have ever believed possible using Naim products. Just goes to show how much pre-amps alter the sound from the CD Players.
Taking into account the number of LDR's needed to match 4 and the time it must take to match and assemble the finished product its a steal at the price your charging.
Cheers
Alan
Hi George,
Early on in this thread there was a lot of talk about LDR distortion
and how audible, if at all, it would be.
Well I was thinking deeply about LDR's today and came to the
conclusion that at exactly 6dB gain reduction, ie; same resistance
value for series and shunt LDR, the distortion of each should
cancel perfectly at the OP.
Since we can look at whatever distortion mechanism that exists
in the LDR as a variation of it's impedance WRT signal level, then
at 6dB attenuation the signal level through each is identical.
As such, any voltage induced impedance change of each should
result in the same exact 6dB attenuation, ie; near perfect
linearity.
Soooo, the big question, provided my assumption is correct, is
do these volume controls sound different from 6dB attenuation
to anywhere else in the range?
It's worth having a listen to see if you can actually pick
by ear any linearity change in its attenuation range.
Cheers
Terry
PS can you get the ones you recommend locally (austr)
Early on in this thread there was a lot of talk about LDR distortion
and how audible, if at all, it would be.
Well I was thinking deeply about LDR's today and came to the
conclusion that at exactly 6dB gain reduction, ie; same resistance
value for series and shunt LDR, the distortion of each should
cancel perfectly at the OP.
Since we can look at whatever distortion mechanism that exists
in the LDR as a variation of it's impedance WRT signal level, then
at 6dB attenuation the signal level through each is identical.
As such, any voltage induced impedance change of each should
result in the same exact 6dB attenuation, ie; near perfect
linearity.
Soooo, the big question, provided my assumption is correct, is
do these volume controls sound different from 6dB attenuation
to anywhere else in the range?
It's worth having a listen to see if you can actually pick
by ear any linearity change in its attenuation range.
Cheers
Terry
PS can you get the ones you recommend locally (austr)
Availability
Vatofile,
I was able to matched up four for a stereo control and eight for a balanced control from a batch of 22 sorted. This was eaiser than it looks, used tighter matching for the series pairs, and shunt pairs so that stereo balance was achieved versus theoretical tracking.
I only matched mine at three currents, 1.5 ma, 0.5 ma, and 0.1 ma. Then plotted them out. The resulting curve is exponetial, but slow decay. The slopes and intercepts were not consistant, but this did allow better matching.
I have only used the stereo version as a clone. Measuring the impedance as the pots are rotated, the matching for series and shunt are very close side to side. The units have a very high tempco. Leaving on all the time should deal with this. Just touching them causes the impedance to change a little.
Like others have posted the results are great. I have no plans on going back to using the carbon PEC pots used before for attenuation. These are much better. And track just as good or better.
Allied Electronics here has the sorted for about 2.00 each. 10 or 12 should be enough for a control. 12 might allow enough for 2 controls.
George
Vatofile,
I was able to matched up four for a stereo control and eight for a balanced control from a batch of 22 sorted. This was eaiser than it looks, used tighter matching for the series pairs, and shunt pairs so that stereo balance was achieved versus theoretical tracking.
I only matched mine at three currents, 1.5 ma, 0.5 ma, and 0.1 ma. Then plotted them out. The resulting curve is exponetial, but slow decay. The slopes and intercepts were not consistant, but this did allow better matching.
I have only used the stereo version as a clone. Measuring the impedance as the pots are rotated, the matching for series and shunt are very close side to side. The units have a very high tempco. Leaving on all the time should deal with this. Just touching them causes the impedance to change a little.
Like others have posted the results are great. I have no plans on going back to using the carbon PEC pots used before for attenuation. These are much better. And track just as good or better.
Allied Electronics here has the sorted for about 2.00 each. 10 or 12 should be enough for a control. 12 might allow enough for 2 controls.
George
Terry Demol said:
Terry if you connect your cd player or dac direct to the poweramp, and had a -6db series/shut resistor network SOLDERED at the input of that poweramp, no the sound would be identical, trouble is you have no way of varying the volume. Add a pot anywhere and the sound will be degraded.
And Allen glad your liking yours, a lot of people have expressed to me that the source and annoyance of their in your face (hard) sound was fixed by removing their preamps and using the Lightspeed, they addmited they were wrongly accusing their cd players for that sound instead of the pre's.
Cheers George
Hi George,
A Lightspeed Attenuator would also be fine for 5.1 systems. 1 Stereo pot to control 6 channels.
In a 6 channel configuration the current on the pot is 3 times higher than in a stereo Lightspeed. Do you know the current values from your Lightspeed and is it possible to use a normal Alps Pot for such currents without shorten endurance?
Arne
A Lightspeed Attenuator would also be fine for 5.1 systems. 1 Stereo pot to control 6 channels.
In a 6 channel configuration the current on the pot is 3 times higher than in a stereo Lightspeed. Do you know the current values from your Lightspeed and is it possible to use a normal Alps Pot for such currents without shorten endurance?
Arne
George,
My Naim NAP 135 clones have an input impedance of 18k according to the Naim manual, presumably from the 100k and 22k resistors in the attached schematic (100k||22k = 18k) ? What would you recommend to increase the input impedance sufficiently to use your lightspeed?
Thanks,
Graeme
My Naim NAP 135 clones have an input impedance of 18k according to the Naim manual, presumably from the 100k and 22k resistors in the attached schematic (100k||22k = 18k) ? What would you recommend to increase the input impedance sufficiently to use your lightspeed?
Thanks,
Graeme
An externally hosted image should be here but it was not working when we last tested it.
gfinlayson said:
Graeme, looking at all the circuits I have for the Naim amps they are all bipolar input, this would be hard to get up to 100k, if they were fet input then this would be a lot easier, maybe the Naim 160 is fet, I'm not sure as it is the only circuit I don't have, or maybe Allen has had his modded.
Anyhow I've revised the minimum input of a poweramp to be used with a Lightspeed, down to 60k or so, as at home I am bi-amped driving 2 x stereo poweramps at 68k each with one stereo Lightspeed, and that becomes 34K the Lightspeed is seeing, and it's doing it on it's ear, with no troubles, great tight low bass still and extened clean highs.
I think I was being TOO conservitive when i quoted 80k-100k.
Cheers George
Lightspeed impedance
Hello George,
Interesting stuff!
Did I understand correctly, that the Lightspeed is similar to a 7kohm pot?
Any way to get the impedance up a lot?
I'm thinking perhaps 7k is the highest possible with "selected" devices, while there are higher that would require more work with matching or so?
/Ronnie
Hello George,
Interesting stuff!
Did I understand correctly, that the Lightspeed is similar to a 7kohm pot?
Any way to get the impedance up a lot?
I'm thinking perhaps 7k is the highest possible with "selected" devices, while there are higher that would require more work with matching or so?
/Ronnie