Some updates:
RESISTANCE IS NOT FUTILE!
First I must apologize for the big build-up this week. This announcement, while it does result in a great sonic advance, is not, as some have speculated, an announcement of an entirely new UpTone/Swenson product. No UpTone optical gadget, no magical power supply.
Actually, the news is better—and I have nothing new to sell you (well maybe some of you).
I will cut right to the chase:
We have added $0.40 worth of tiny resistors to the circuit board of the USB REGEN and increased its musical performance by 40-50%—most especially in the bass! Anyone who has a REGEN will be astonished by the improvement, and anyone who hears the new one for the first time will be even more amazed. Given the already unanimous rave reviews being heaped on the REGEN—by the 95 people around the world who got one from the mid-April first run—it is hard to fathom how the thing could get that much better. But it has. And John and I are feeling rather gleeful about it.
[If you wish to skip the below story and just find out about what this means for the REGEN you either have in your hands or on order, just jump to the very bottom of this post.]
Actually, some of you have heard a bit of exactly the improvement made to REGEN, and there is a short story behind that. It of course involves the recently popular Corning 3.OpticalUSB cable which some of you have bought (and some struggle with for compatibility/reliability reasons).
The Corning is interesting, and is a more complicated device that it might seem. Corning had to design a small ASIC circuit to translate the semi-differential USB signaling to a format that could work with the timing of their optical transceivers. And the receiving end of the Corning cable needs a small voltage to run (which may in fact be 12 volts converted with a switching regulator at the send end, as the USB 5V from the computer would not survive their shortest 10 meters length), so there is a pair of 28awg wires run alongside the optical fiber.
So not only does the Corning not provide galvanic isolation from the computer, but there are some other things going on with its power and USB ground wires. (As most of you know, the REGEN ignores the pin 1 +5VBUS wire from the computer or the Corning and does not connect it to anything since the REGEN has its own power and also makes clean 5V for DACs that need it.) Still, aside from the REGEN as hub often helping the Corning work with computers and DACs, having the Corning feeding the REGEN sounds good—especially in the bass as many have reported.
Now John Swenson had speculated that cutting the Corning’s wires and feeding cleaner power to the receiver near the DAC end might help it further, so he bought one to listen to and planned to then dissect it. When I was at his place last weekend (before and after the fun ComputerAudiophile Berkeley lunch get-together), we were listening to the Corning with the REGEN (I brought a thumb drive will all my favorite test tracks), and indeed the Corning seems to significantly increase bass detail. Quite a nice combination.
Then John began speculating about the extra resistance that 10 meters of 28awg wire presents to the USB ground wire. The typical ground path in an audio system measures between 0.3-0.5 ohms. But that 33 feet of thin wire is going to be more like 3 ohms, and that forces some of the USB ground currents to take other paths. Maybe that is a good thing? So he carefully cut open one of the 6-inch USB cables I supply with the REGEN kit (for those who can't use the solid adapter), disconnected the shield entirely, and inserted a resistor in series with the black ground wire.
Now we had just listened to that same, unmodified, 6-inch cable feeding the REGEN (straight from John’s Squeezebox Touch just inches away—so it was Touch>6” cable>REGEN>solid adapter>Bottlehead DAC), and as I have posted elsewhere, I don’t like the sound of that 6” cable at all—even before the REGEN.
But now we go and put the same short cable—with resistor in series with the ground wire—back between the Squeezebox Touch and the REGEN. Oh my! Exactly the same sort of benefit we heard from having the Corning cable feeding the REGEN.Wow! No doubts at all. So we turned to each other and then raced to his engineering bench room to do the logical thing: LiftUSB input pin 4 of the REGEN and put the resistor between it and ground. (The data +/- pair of USB really does need a ground reference to work reliably.) Back to the studio to listen to it fed by the much smoother Supra USB cable instead of that nasty 6-incher. Nice.
After that, we of course wanted to hear what if any benefit the Corning might still have feeding the REGEN. And guess what, to our ears it sounded a little worse. We think that the Corning’s benefit previously (before modifying the REGEN) overshadowed whatever effects its complex circuitry has.
So as far as we can tell, the Corning benefit was all about that extra resistance in the USB ground. The resistor we chose to use is greater than the 3 ohms John calculated that the Corning inserts, but I’d prefer not to say what value we settled on.
On my drive home I realized that while we cut the shield with the 6-inch cable, when we first modified the REGEN we did not lift or insert resistance on its input jack shield contacts (virtually all USBhosts and devices end up connecting pin 4 ground to shield at some point), thus some ground currents were finding a path over the shield. I asked John why we still heard such a big difference, and he thinks it is because USB shield wires are far more symmetrical around the +/- data wires and thus they are more equally affected by whatever ground-noise currents and cancel out. I think very few USB cables shield the data wires separate from the power and ground, so the ground wire is not symmetrical about the data lines.
So that night John drilled out the ground vias for the jack, and listened with and without resistors there too. The board we had modified with the one resistor was one of the pre-production prototypes (not a production REGEN board with the layer stack-up he optimized impedance matching for), and I took that one home with me to listen to. Thus the REGEN John further modified and listened to was the production unit I gave him (his first with a case!), and he said it sounded even better when he was all done.
So that’s the long story. Maybe you will or will not continue to like the Corning cable with the updated REGEN. But you WILL be absolutely delighted by the sound of the new version.
__________________________________________________ ______________
Okay, now you all want to know what this means both for theUSB REGENs you have on order, and for those 95 units already in use by the earliest adopters.
Well I scrapped 150 bare boards that were about to be populated at my board house. The new version, with the 3 added ultra-low noise resistors, is at the PCB maker now for a 500 unit run. And I just mailed a 5-figure check to the assembly house to cover those 500 bare boards and parts and labor to stuff 350.
Previously, if your web order number was between 1091 and 1223, the promised shipping dates were from May 29 to June 3.Orders in that range will now be shipped by June 10th.Only a 1-week slip. Worth the wait I promise!
If your order number is 1224 or above, then the hold-up is not the new circuit boards, rather it is the aluminum cases. Japanese “Golden-week” holiday ends this coming Monday and so next week we will have an answer regarding delivery of the case order that will cover these latest orders. People in this group,beginning with order 1224, already saw—at the time they ordered—a promised ship date of June 18th. So for this group there is no further delay at all. In fact, your REGENs may ship 2-3 days earlier than that.
So once again: All unshipped REGENs will be of the awesome new version!
Some of you who already have a REGEN have ordered another (or several!) more—thank you. How will you tell the difference between the first version and the new one? Well, sonically you will hear it right away. But just to make it easy, and to not have to change the case markings (too late for my enclosure vendor), I selected an amber LED to replace the green one. Maybe that will match the amber dot LED of our JS-2 and in the “p” of UpTone on the web site. I am trying to avoid calling the new version “REGEN Amber,” but I know some of you might and that’s okay.
This post is already ridiculously long, so I am going to make another explaining how existing, “REGEN Green” owners can update to “Amber.”
