Beyond the Ariel

[BudP]

Lynn,

I think I provided you with a pair of early electron pools, back when I sent you the Litz wire speaker cables, the ones with the twist lock banana plugs. These would be 5 inch long pieces of cotton tube, with a small length of tinned wire sticking out of one end, just beyond some shrink tubing.

Bud

[Lynn Olson]

I'm a little embarrassed to say that the big 2-channel rig still remains disconnected, while the Marantz AV8003/MM8003 5-channel setup is up and running - but with pretty mundane Blue Jeans XLR interconnects and speaker wire (Belden & Mogami twisted-pair wiring). Sound quality is listenable, which is more than I can say of the Denon 2905 I had previously, but not what I'd call high-end.

No reason I can't try the Electron Pools right away, letting it dangle from the speaker inputs of the Ariels. The resolution of the Marantz is so much lower than the all-triode 2-channel system, I wonder if it will be audible. No harm in trying.

[Lynn Olson]

Originally Posted by soongsc

Just this evening, we were testing interconnects on a friend's system that has an Pass Labs Aleph 0 connected to a Focal-JMlab Micro Utopia Be, to make it short, I never hear this combination so good. My friend was asking to keep the prototype to play around with more, but I could not leave it out of my sight. So I promised once I had everything in place and the samples in, I'd let him borrow the samples.[/FONT]

Hmm - interesting. Back in my Positive Feedback reviewing days, I had a pair of Alephs for review (with the Ariels of course). Unfortunately, I couldn't write a review - those amps were the most sensitive to speaker cables of any amplifier I had ever tried (and by then I'd gone through about 40 different amplifiers with the Ariels). Four different cables, four different amps, and no way to know what the thing was supposed to sound like.

What little I could tell was the Aleph was designed for a speaker very different than the Ariels, so for reasons unknown, the two were not a good match. I passed it on to another member of the PF reviewer group, and they went on to get a writeup in the magazine. (At that time, all of the PF reviewers had very different systems, so what worked in one system would not sound as good in another.)

[BudP]

Let me see about getting you some commercial spec parts, and maybe something for the Marantz too. Had I thought about it, I would have had you go by the Ground Control room at the RMAF and meet with Byron Collett.... ah well.

Bud

[soongsc]

Originally Posted by Lynn Olson

Hmm - interesting. Back in my Positive Feedback reviewing days, I had a pair of Alephs for review (with the Ariels of course). Unfortunately, I couldn't write a review - those amps were the most sensitive to speaker cables of any amplifier I had ever tried (and by then I'd gone through about 40 different amplifiers with the Ariels). Four different cables, four different amps, and no way to know what the thing was supposed to sound like.

What little I could tell was the Aleph was designed for a speaker very different than the Ariels, so for reasons unknown, the two were not a good match. I passed it on to another member of the PF reviewer group, and they went on to get a writeup in the magazine. (At that time, all of the PF reviewers had very different systems, so what worked in one system would not sound as good in another.)

Maybe the Aleph used lots of feeback? I think the Aleph and the Micro Utopia Be would have sounded even better if the Micro Utopia Be were tuned to a lower frequency, really a pity.

[dantheman]

Originally Posted by soongsc

The real skill is what criteria you tune it to. There is a very narrow band of optimum performance. If I told you the space between the lines, it would be almost giving the whole know-how away. My intent is to inspire more data than just talk.

Just this evening, we were testing interconnects on a friend's system that has an Pass Labs Aleph 0 connected to a Focal-JMlab Micro Utopia Be, to make it short, I never hear this combination so good. My friend was asking to keep the prototype to play around with more, but I could not leave it out of my sight. So I promised once I had everything in place and the samples in, I'd let him borrow the samples.

Somehow I knew it. Thanks for the data though.

much appreciated,

Dan

[soongsc]

I wish I had students that were so enthusiastic. Sometimes I think it's funny why they don't want to get the most out of what they paid for.

[Pano]

Originally Posted by Lynn Olson

Connect device A and B (say, a linestage and a power amplifier) to the AC wall socket. DO NOT interconnect them. Turn them both on. Measure the AC Volts potential between the two chassis.

Now that we've done that with the linear supplies, were does that leave us with the SMPS? Do they have a side that radiates into the chassis more than another?

[soongsc]

Originally Posted by TerryO

Sorry, wrong Olson I guess. I won't offer any more opinions.


Best Regards,
TerryO

I've heard lots of good things from Bud's ideas.
If Lynn had used them, the next I would ask about is EnABL

[Lynn Olson]

Originally Posted by panomaniac

Now that we've done that with the linear supplies, were does that leave us with the SMPS? Do they have a side that radiates into the chassis more than another?

Now that's a good question. I don't how the transformers in these things work - but they certainly do have transformers, and there's probably a leakage pathway from the incoming AC to the chassis (which is what causes all the trouble).

I haven't even mentioned switching noise, which is common to both linear and switchmode power supplies. It's not generally recognized that the very sharp current pulses radiate energy in several directions: back down the power cord (which then becomes an antenna that affects all nearby linear electronics), the percentage of noise that is not removed by the power supply electronics (no lowpass filter or regulator has 100% removal of noise), and magnetic induction into the space inside the chassis of the electronics. Think of it as a captive AM transmitter and you'll be getting the right idea.

That's why swapping power cords makes a difference - they are antennas that radiate wideband noise into nearby linear electronics, so differences in cable shielding and twisting the wire makes a difference. The power transformer makes a difference too: toroids typically have much more bandwidth (say, 60 to 100 kHz) compared to EI core power transformers (typical rolloff around 5 kHz). By contrast, if the windings are electrostatically shielded (have a foil wrap), that reduces capacitive winding-to-winding coupling substantially (20 to 40 dB reduction).

The "loop area" of the power supply circuit (the area occupied by supply and return wires between the rectifier and first capacitive shunt element) controls the effective size of the "antenna" that is radiating into the rest of chassis - thus, twisting wire and reducing the physical size of the first portion of the PS is a good idea.

The best approach of all is to carefully control the rate of charge between the rectifiers and the first shunt capacitor. The rate of current flow can be very high when a partially discharged capacitor is "topped up" by the next AC peak waveform, so it can be very effective to use a small resistor or even a little inductance between the rectifier and the first shunt capacitor - this can slow down the rate of charge many times, and thus reduce emission of magnetic pulses into the power cord, amplifier chassis, and across the filter elements of the PS filter section.

This is also the reason that mindlessly increasing the size the capacitors in a conventional bridge-cap power supply actually makes the noise worse, not better, since the charging pulses have a faster rise time and greater current flow. Think about it: if the cap were infinitely large and the voltage drop across the rectifier, transformer, and power cords were zero, than the current would be infinite at the moment the rectifier first starts conducting. Is that what you want? Better to control the rate of charge of the capacitor, rather than leave it to value of ESR in the capacitor and voltage drop across the rectifier and power transformer.

P.S. Bud's Ground Control gizmos are on their way, suitable for loudspeakers and electronics-chassis applications. One RCA-format GC "tail" can apparently do its magic on an entire chassis, since it is common practice in a lot of consumer gear to have all grounds in common. The Ariels have fully isolated high and lowpass sections of the crossover, so each driver gets its own GC device, connected to the low-side terminal. I'm kind of curious if this will be audible with upper-mid-fi electronics.