Bgt said:george, where to find your lightspeed att. on the net. URL please
Sorry Bert, but it's not anywhere on the net, but here on this forum, I do it as a hobby in small volume by word of mouth. The moment it becomes hard work, I'll probally lose interest. You can always PM me, if you need other info.
Cheers George
Badge said:
Badge, mine in my system is driving a bi-amped setup into a pair of Martin Logan Monolith III's with no problems, each stereo amp is 68k input, this means the Lightspeed is driving 32k. But I will state that my Musical Fidelity A3-24 Dac has a very strong output stage, being discrete class A at 100ohms, and can give very strong drive (attached is the circuit). When I stated that input impedence of the amp should be >68k I was was being very conservative, because it comes back to how strong you source is.
In your case 100k amp and probally 50k crossover, which becomes total 33k that to me still is ok so long as your source has a good strong and low output impedence (<100ohms) I don't see a problem.
Cheers George
Attachments
Questions related to building my own
These questions should maybe be asked in a pm, since there are commercial interests to protect. I ask in public to help myself and maybe others understand better these interesting devices.
I use active linestages. This looks like a nice way to attenuate volume. So hopefully the driving issues are not important.
Both current linestages have mono volume pots, and there are plenty of unused Alpha stereo log pots laying around here to make a voltage controller. So outfitting a low voltage supply and a small breadboard to hold the LDR's is a very simple mod.
Several tech questions. You mention using mid pot rotation for most listening levels. Is this a function of using log pots or of the linearity of the LDR"s with respect to applied voltage?
Looking your 5 volt supply, the 100K pot the mid rotation level is in the 1 ma amp range. Would a lower value pot give more usefull range of rotation?
This is asked because most log pots are not really log or audio taper. Just two linear taper sections in series. This gives a close fake log. The mono pots used now have 330 degree rotaion, and about 180 degrees is usable depending on source and level desired.
One linestage needs balanced mono controls, so this looks like two stereo pots and a total of 8 LDR's. The LDR must need to be paralleled as far as series and shunt lelement to each voltage control pot. That is sweet. Does making balanced controller mean maybe a 50K pot to deliver twice the current?
My LDR's are sorted, my plan to match is apply 6.2 volts with a 1K resistor in series. Then measure the voltage across the resistor. This current should give an idea of on resistance. Is this close enough? Hope to match four series elements and four shunt close enough to allow fine tuning of tracking with voltage controller pots. Most stereo pots really do not track real well anyway. Try measuring a stereo pot as it is rotated. Some might measure 2 -5 % in overall resistance, and vary by 6 dB at 25, 50, and 75% of rotation.
Sorry to babble on about pots, this is the most difficult part here it seems. Feeding the proper voltage to the LDR's.
George
These questions should maybe be asked in a pm, since there are commercial interests to protect. I ask in public to help myself and maybe others understand better these interesting devices.
I use active linestages. This looks like a nice way to attenuate volume. So hopefully the driving issues are not important.
Both current linestages have mono volume pots, and there are plenty of unused Alpha stereo log pots laying around here to make a voltage controller. So outfitting a low voltage supply and a small breadboard to hold the LDR's is a very simple mod.
Several tech questions. You mention using mid pot rotation for most listening levels. Is this a function of using log pots or of the linearity of the LDR"s with respect to applied voltage?
Looking your 5 volt supply, the 100K pot the mid rotation level is in the 1 ma amp range. Would a lower value pot give more usefull range of rotation?
This is asked because most log pots are not really log or audio taper. Just two linear taper sections in series. This gives a close fake log. The mono pots used now have 330 degree rotaion, and about 180 degrees is usable depending on source and level desired.
One linestage needs balanced mono controls, so this looks like two stereo pots and a total of 8 LDR's. The LDR must need to be paralleled as far as series and shunt lelement to each voltage control pot. That is sweet. Does making balanced controller mean maybe a 50K pot to deliver twice the current?
My LDR's are sorted, my plan to match is apply 6.2 volts with a 1K resistor in series. Then measure the voltage across the resistor. This current should give an idea of on resistance. Is this close enough? Hope to match four series elements and four shunt close enough to allow fine tuning of tracking with voltage controller pots. Most stereo pots really do not track real well anyway. Try measuring a stereo pot as it is rotated. Some might measure 2 -5 % in overall resistance, and vary by 6 dB at 25, 50, and 75% of rotation.
Sorry to babble on about pots, this is the most difficult part here it seems. Feeding the proper voltage to the LDR's.
George
I tried a lower value pot but the gain comes on too strong. I would have prefered to use a linear dual 100k but here in Australia they are rare for some reason, the log dual 100k is in abundance though.
If you wanted to make a fully balanced form of Lightspeed you could use 8 x of the ldr's and 2 x 100k dual lin or log that would give you individual channal gain (balance control) or a single quad 100k lin or log but here in Australia thank god there is'nt much call for a full balanced Lightspeed. I do'nt think I would make it anyway as the matching of all 8 ldr's for channel balance would be a P.I.T.A.
As for the led current, the led in the Silonex NSL-32SRS (sorted) works perfect with +5vdc 100ohm current resistor, as for what led you are using this will change just use the led current resitor formula.
Cheers George
If you wanted to make a fully balanced form of Lightspeed you could use 8 x of the ldr's and 2 x 100k dual lin or log that would give you individual channal gain (balance control) or a single quad 100k lin or log but here in Australia thank god there is'nt much call for a full balanced Lightspeed. I do'nt think I would make it anyway as the matching of all 8 ldr's for channel balance would be a P.I.T.A.
As for the led current, the led in the Silonex NSL-32SRS (sorted) works perfect with +5vdc 100ohm current resistor, as for what led you are using this will change just use the led current resitor formula.
Cheers George
Attachments
Burt, you will have to spend a bit of time and do some measuring, even with my so called matched ones I still have to re-match it is time consuming, this is why it can take me one day to build Lightspeed from begining to end to complete, once you have exprienced this you will understand why the price is $500aus.
You can cut down on this build time by being satisfied only having good tracking at the mid point of the volume control and out by 4-6db at all other settings this is up to you, but I match mine so they are within 1db from minimum to maximum volume.
Cheers George
You can cut down on this build time by being satisfied only having good tracking at the mid point of the volume control and out by 4-6db at all other settings this is up to you, but I match mine so they are within 1db from minimum to maximum volume.
Cheers George
I hope not!
I planned on only matching my selected at one operating point. Having two stereo switches will allow channel balancing by ear. Sounds like a good plan anyway.
The operation of the voltage control still is cloudy in my mind. I guess these pull 2 ma or so when fully on. With only 5 volts and a 100K pot, it seems that the right amount of current can be delivered when the pot is almost open all the way. With a linear pot, half the resistance is in series with the diode when set in the middle, so 5 volts/50K is 0.1 ma.
Guess I need to get off my, and try it to see how it works. Picked up a couple 50K dual linears today at lunch. These are Alps 50K B taper, 1/4 watt, carbon. Measured them end to end and all sections measured in the 44K - 46k range. Figured that since each section will have two diodes, the lower value resistance may work better. Heck with a linear pot, 10K may give more range of control. These are easy to find.
Might get this thing going this weekend. Really looking foreward to hearing these compared to a nice quality molded carbon log pot.
Thanks for sharing your ideas George!
George
I planned on only matching my selected at one operating point. Having two stereo switches will allow channel balancing by ear. Sounds like a good plan anyway.
The operation of the voltage control still is cloudy in my mind. I guess these pull 2 ma or so when fully on. With only 5 volts and a 100K pot, it seems that the right amount of current can be delivered when the pot is almost open all the way. With a linear pot, half the resistance is in series with the diode when set in the middle, so 5 volts/50K is 0.1 ma.
Guess I need to get off my, and try it to see how it works. Picked up a couple 50K dual linears today at lunch. These are Alps 50K B taper, 1/4 watt, carbon. Measured them end to end and all sections measured in the 44K - 46k range. Figured that since each section will have two diodes, the lower value resistance may work better. Heck with a linear pot, 10K may give more range of control. These are easy to find.
Might get this thing going this weekend. Really looking foreward to hearing these compared to a nice quality molded carbon log pot.
Thanks for sharing your ideas George!
George
Diodes matched up
Recieved my 10 "sorted" diodes this week. Matched them up last night.
First, there is a mistake on the datasheet. The pinout for the diode side shows that the narrower lead is the anode. After trying it for a couple minutes, figured out the narrower pin is the cathode. There is a white dot and it is always on the cathode of the LED.
Only planned on trying a normal attenuator when ordering, did not get enough to match up a balanced volume control. There were two out of range, the other eight are close enough to be usable.
The testing was done at 1.5 ma and 0.5 ma. The eight read 890 ohm +/- 60 at 0.5 ma and 350 ohms +/- 20 ohms at 1.5 ma.
The diodes were factory sorted at 164 - 181 at an unspecified current. Something like 1 ma +/- 1 ma.
Waiting on some more, hopw they are f range also.
Slowly coming together.
George
Recieved my 10 "sorted" diodes this week. Matched them up last night.
First, there is a mistake on the datasheet. The pinout for the diode side shows that the narrower lead is the anode. After trying it for a couple minutes, figured out the narrower pin is the cathode. There is a white dot and it is always on the cathode of the LED.
Only planned on trying a normal attenuator when ordering, did not get enough to match up a balanced volume control. There were two out of range, the other eight are close enough to be usable.
The testing was done at 1.5 ma and 0.5 ma. The eight read 890 ohm +/- 60 at 0.5 ma and 350 ohms +/- 20 ohms at 1.5 ma.
The diodes were factory sorted at 164 - 181 at an unspecified current. Something like 1 ma +/- 1 ma.
Waiting on some more, hopw they are f range also.
Slowly coming together.
George
Hi,
Interesting thread!!
I remember trying to do something years ago (for my A level technology, disturbingly) but with FETs being controlled by a single potentiometer.
I assume your solution would be far less noisy.
The main reasons I thought it was a good idea was that several channels could be controlled by a single pot (so they should have similar charateristics) and so that the signal could stay right at the back of the preamp, and not have to traverse to the front (or halfway with a long spindle) to the volume control.
In the end, I abandoned my idea for a quad volume control chip or something (I'd NEVER do that nowadays, but it was A level!), but it was a nice idea!
Anyway, your's is far more elegant and it works well!
Cheers,
Phil
Interesting thread!!
I remember trying to do something years ago (for my A level technology, disturbingly) but with FETs being controlled by a single potentiometer.
I assume your solution would be far less noisy.
The main reasons I thought it was a good idea was that several channels could be controlled by a single pot (so they should have similar charateristics) and so that the signal could stay right at the back of the preamp, and not have to traverse to the front (or halfway with a long spindle) to the volume control.
In the end, I abandoned my idea for a quad volume control chip or something (I'd NEVER do that nowadays, but it was A level!), but it was a nice idea!
Anyway, your's is far more elegant and it works well!
Cheers,
Phil
Hello. I recently purchased the Lightspeed attenuator from George and am extremely pleased with it. Its compact, very light and exhibits no sound of its own. Bass is stronger and deeper. Treble is clear and has increased with bigger slam factor on drums and bass, better ambience, improved stereo imaging with better detail. Also my stereo sounds better at low volumes. I will be trying to sell my Croft Epoch Elite preamp soon as sound is now easily bettered by the Lightspeed Attenuator. Its a bit of a secret as to what's in this device but if you have the opportunity to, go ahead and buy one from Georgehifi.
Banned
Joined 2002
Banned
Joined 2002
A buffer is like a pre-amp in a way. BUT it has way fewer parts. You might want to see what kind of impedance's you are dealing with and other variables. I don't use a buffer at all But i might build one to get more gain from my Aleph Mini's. They go pretty loud with the output of the cd-player but more is a little better. I have a few buffers i use that a friend has that i can play with maybe share with you.