Any problems with sound balance due to the heat from the Aikido tubes? How did you layout the LSA board with respect to the Aikido preamp board?
Thanks
Joey
I was one of the first to bring up the issue of power supply regulation for better sound. I heard my own version of the improvement a good regulated supply can bring. There must be a technical definition that gives this audible result. Wouldn't those low-frequency signals intermodulate? I mean, you have a supply jumping around with these semi-chaotic but quite identifiable spikes. Each of those spikes will create its own lower-frequency voltage signal in the LED, which will then emit a lightwave with a virtual sea of VLF modulations. It seems those would combine in nonlinear ways as audio frequency modulations appearing and disappearing.
If you use pure dc as the Lithium Ion battery pack I have recommended and do a blind A/B as I have done with 20+ audiophiles, they could not reliably detect any change from it and the standard regulated 5vdc supply that is in the production Lightspeed Attenuator, even Sam Tellig an experienced reviewer said he cannot tell. If there were any "spikes" which there aren't, the CdS cell (ldr) is far too slow to be able to react to them anyway.
Cheers George
Cheers George
I replaced the stock Shallco series attenuator in my Atma-Sphere MP-1 integrated preamp with a 100K balanced shunt attenuator comprised of TX2575s in the series positions and matched LDR shunts between phase and anti-phase. This works great and easily surpasses Shallco. The matched L & R LDR shunts track fairly well in summer and less so in the cool of winter. You need a balance control or trim control pots for this. In a high-impedance volume control as required for an active preamp, even carefully matched LDRs won't track acceptably in the Lightspeed Mk II configuration. However a fixed series resistor with a variable LDR shunt works great. You can select the control pot and balance/trim pot values to set whatever range of useful attenuation you want to give to the volume control. Ideally for a shunt volume control this is toward the attenuation side of the range.
I don't see why "too slow" should necessarily mean "inaudible magnitude" or some such. Those "lagged" LDR responses are surely there, I mean, objectively, theoretically. And surely lag effects, to call them by some name that gives a workable image, will combine in nonmusical fashion---those lag-changes will combine to create lag+lag+lag-here, lag+lag-there, etc. Apart from that, it as a fact that many posters with only the goal of reproducing music suitable to their ears have claimed audible benefits in regulating the LED supply, and by various means with reported differences between those means.
Fwiw, commercial versions of the Lightspeed (DarTZeel, Constellation) regulate LED supplies and smooth the effect of any resistance changes, small as those may be, by parallel LDR operation (100 and 48 LDRs respectively).
Any external power spikes if any would be ignored by the ldr's as they cannot react to them fast enough, so in fact they provide their own way of regulating mains spikes, as they are too slow to react to them.
It's all a mute point as the noise that's on the output of the Lightspeed Attenuator with a shorted input (the way it should be measured) is a few uV (microvolts, "probably the crows test leads") not mV (millivolts) as most preamps give out, this is why it is very hard to detect any difference between pure battery and the regulated supply that in the Lightspeed Attenuator as so many A/B's have proved. We're only powering led's not the national grid. You can almost rub two sticks together to light up a led.
Cheers George
A possible reason why batteries may sound better from John Swenson:
Two source of noise in normal DC powersupply:
One is the RF noise generated by the reverse charge of the diode. The other is the resonance of the transformer secondary. In many cases this resonance falls right in the range of noise generated by the diode. The diode noise excites the resonance which can amplify it to a large degree. Not only is this an issue for the box the PS is in, but the signal from the resonance passes right through the transformer and onto the power cord where it can go to other boxes and be radiated.
RE: Low RFI power supply design - John Swenson - Tweakers' Asylum
Two source of noise in normal DC powersupply:
One is the RF noise generated by the reverse charge of the diode. The other is the resonance of the transformer secondary. In many cases this resonance falls right in the range of noise generated by the diode. The diode noise excites the resonance which can amplify it to a large degree. Not only is this an issue for the box the PS is in, but the signal from the resonance passes right through the transformer and onto the power cord where it can go to other boxes and be radiated.
RE: Low RFI power supply design - John Swenson - Tweakers' Asylum
I see that you integrated preamp is tube based. Did the tube heat have any effect on the tracking? How was it laid out inside the preamp?
JoeyGS, The preamp has 14 tubes. The LDRs drift up to 10% depending on room and chassis ambient. I use the Lightspeed Mk I configuration-- an LDR in the shunt position only, in a balanced application shunting between phase and anti-phase. With this arrangement any LDR mismatch or temp drift does not affect CMRR. It requires only an adjustment to L/R channel balance. For this purpose I added L and R 50K trim pots. A 500K dual-log pot controls master volume. This three-pot arrangement drops into the three front panel positions of the stock Atma volume controls.
I power the LDRs with a Twisted Pair Placid shunt regulator placed inside a small project case connected by umbilical to the preamp chassis. The PS could have been placed inside the preamp, but the Atma is a two-chassis preamp with only DC in the top unit. I didn't want to bring AC in there. I added .1uf and 5uf film bypass caps close to the LDRs to compensate for the long power supply umbilical.
For LP and CD sources I needed a useful range of -70db to -30db attenuation. This required that the resistance of the shunt LDR go from full bright (around 40R) to 3K2. A 500K dual log control pot obtains this range from the LDRs. Being able to use the full range of the control pot is a nice improvement in granularity over the stock Shallco. I had only been using the first 6-7 steps of the series attenuator.
To prevent the preamp from jumping to full volume in the event of an LDR failure I added a 3K2 fixed resistor in parallel with each LDR shunt.
I power the LDRs with a Twisted Pair Placid shunt regulator placed inside a small project case connected by umbilical to the preamp chassis. The PS could have been placed inside the preamp, but the Atma is a two-chassis preamp with only DC in the top unit. I didn't want to bring AC in there. I added .1uf and 5uf film bypass caps close to the LDRs to compensate for the long power supply umbilical.
For LP and CD sources I needed a useful range of -70db to -30db attenuation. This required that the resistance of the shunt LDR go from full bright (around 40R) to 3K2. A 500K dual log control pot obtains this range from the LDRs. Being able to use the full range of the control pot is a nice improvement in granularity over the stock Shallco. I had only been using the first 6-7 steps of the series attenuator.
To prevent the preamp from jumping to full volume in the event of an LDR failure I added a 3K2 fixed resistor in parallel with each LDR shunt.
Thanks George.
So what do people use as output capacitor values in their source equipment preceeding the Lightspeed?
I don't feel like putting another buffer between the source and the lightspeed.
Hi, if you have to have an output coupling cap on your source because of DC offset from it, you would be ok with a 5uf cap, that will give you a -3db @ 3.2hz.
If you don't have any DC offset from your source before the cap, you should not need the cap, then you will be dc coupled.
Cheers George
If you don't have any DC offset from your source before the cap, you should not need the cap, then you will be dc coupled.
Cheers George
For less than the cost of an audiophile 5uF cap, you can add a follower output, either emitter or cathode, and put a servo below it to DC couple the output. Berning, Yaniger, Broskie and Rankin have versions. I've used the Berning/Rankin variants, but prefer the Broskie (a little different than John's, but pretty close) version. A friend actually made up PCBs for Broskie's version. Makes adding DC coupled output to a tube stage a cinch. Also, with the right tube & operating point, it makes driving the Lightspeed much easier.
Stuart
Stuart
What source do you have where you need an output cap, yet don't have one already? And if so what value is it now?
As the Lightspeed does not dictate that you need a cap on your source, only your source dictates this, because it has some serious DC offset or it is a tube output stage.
BTW a 4.7uf quality polyprop cap can be got for around ten bucks, that is if you need one.
Cheers George
George, on your site you say: "It is believed that any light weight contacts (switches and especially pots and their wipers) have a diode rectification effect on the AC music signal, trying to rectify it which is very bad. The Lightspeed Attenuator (LDR) system totally addresses is diode rectification effect. This is why it is so transparent and totally honest to the source, not adding any colorations or distortions to the source, it is just like being able to plug your cd players fixed output directly into your poweramp yet having control over the volume."
My own preamp has a conventional capacitor-coupled analogue volume pot and has worked perfectly for 20 plus years. There is no scope-visible channel drift, distortion, noise (above source), nor thermal instability and I hear nice music.
So although I have no motivation to implement your circuit in any particular project (especially since it passes the DC current for the LEDs to ground across the wiper contact of the logarithmic/fragile control pot) and am not one to automatically assume that a diode rectification effect in potentiometers is at all specious, I wonder if you haven't done so already, would you care to substantiate the basis of the belief and/or direct your followers to an authoritative technical paper describing how the phenomena manifests itself in conventional volume pots. My web searches (limited by the word "potentiometer") have all directed back to your attenuator.
Thank you.
My own preamp has a conventional capacitor-coupled analogue volume pot and has worked perfectly for 20 plus years. There is no scope-visible channel drift, distortion, noise (above source), nor thermal instability and I hear nice music.
So although I have no motivation to implement your circuit in any particular project (especially since it passes the DC current for the LEDs to ground across the wiper contact of the logarithmic/fragile control pot) and am not one to automatically assume that a diode rectification effect in potentiometers is at all specious, I wonder if you haven't done so already, would you care to substantiate the basis of the belief and/or direct your followers to an authoritative technical paper describing how the phenomena manifests itself in conventional volume pots. My web searches (limited by the word "potentiometer") have all directed back to your attenuator.
Thank you.