Uptone Audio Regen

[Superdad]

Wow guys, thanks for your enthusiasm! Gives me motivation to move the project along quickly. And I know this early interest will please John too.

John has done his part (I think I mentioned that I just got back from him a final prototype with big cap to improve PDN, and he has already rearranged the board layout), so I could start PCB production any time.

I'll get going on quotes for the milling and printing of the required openings for the end caps, and will start to add up costs so a price can be set.

 

I promise to keep you all updated.

Have a great holiday!

--Alex C.

[Superdad]

Alright gents, better tone it down before I get into trouble with the management.

 

We are not much into crowd-funding/design (unless you want to design a camel), but I'll take input on color combo for the little case. Here are the choices. No promises about going along with the crowd, and in the end there will be only one color combo offered.

 

Remember, this thing will measure just 55mm x 45mm x 18mm

 

[Superdad]

Gentleman: Thank you all for your color input. To be honest, I did it just for fun, to see who is hanging out here, and to give you something to talk about while we move the Regen towards production. Glad to see that black for the extruded part of the case is acceptable to most, since with the samples I have here the black looks nicer. This thing will be hidden behind your DACs anyway, so the color combo is mostly about marketing appeal.

 

Yet this is a perfect--and harmless--small example of what happens with any crowd-design project. I only asked about color preference and got more than a half dozen different responses (admittedly from among 14 options). I now have much more empathy--and pity--for firms that attempt to take input from the "crowd" on the functional and technical design of their products.

While I am not entirely in the Steve Jobs camp of "No focus groups, we tell the customers what they want," I know that the other extreme of community/committee design often results in madness and delay.

 

 

Gary asked a good question with regards to power. The answers are:

 

a) No, it is not powered from the USB bus. Even with the ultra-low noise regulators being used, it would kind of defeat part of the purpose of the device to do so;

 

b) Since the device uses two (of the somewhat pricey T.I. TPS7A4700 ; $4.60 each at my quantities) regulators, one for delivering better 5V bus power, the supply needs to be greater than 5V for that regulation to work. 6V or 7V will be required. I'll check with John as to if 9V would result in too much heat from the drop.

 

c) I plan to provide with the Regen a small, 6V, 3 watt, regulated linear wall wart (probably this one from Jameco unless I find a better for the price or less expensive equivalent.

 

d) Of course everyone is free to use a better PS with the Regen. A spare output from a JS-2 (set at 7V) certainly works nicely!

 

Thanks again for your interest. I will report more in few weeks as it progresses towards production.

 

Happy Holidays and New Year to all!

 

--Alex C.

[Superdad]

Just to be clear 6Vdc right? 6Vdc @ (3W) = .5A Correct?

 

Correct. 6-7V DC, 0.5A is fine. That little Jameco I linked to should be okay since the power really gets cleaned on our board as well. But I will of course listen to it versus my JS-2.

 

 

Hi Alex,

Add one to my JS-2 order please.

Brian

 

Thanks Brian, but your JS-2 will be coming to you in January and I expect these little Regens are a couple months out. Postage for the little fellas will be almost nothing though, so don't worry about that.

[Superdad]

Also, will you have an early adopter program before the official release?

 

Thanks again for your interest everyone.

 

As for an "early adopter program" Jason:

Since all parts are surface mount and the little chassis have to be machined/printed in a good size batch to be anywhere near affordable, we will be going straight to production. So the price will be the price (I think special discounts pervert business models, force the real retail price to be higher later, and make later buyers feel badly). Early adopters will be those of you who buy one from the first run (I'm guessing 50 or 100 is all I'll be able to capitalize to start).

 

Ciao!

 

ALEX

[Superdad]

Alex is putting a lot of his time into this, investing a lot of money, for sth which is still a hobby, not a real business. ... Support him.

 

Thanks Adam. But don't tell that to my wife! All this IS supposed to be a real business and contribute to our household and help get our kids through college (or at least out the damn door).

 

And I already do feel supported by the great members here. You guys are awesome with your enthusiasm.

 

My primary goal is to bring John Swenson's best and most practical designs to market. Each one will hopefully help capitalize the next so we can then get the more ambitious (and groundbreaking) projects out to the world. The Regen is pretty simple, but the long gestating follow-on, the UpTone/Swenson USB>Ethernet Audio Bridge is what will really put us on the map. (OEMs, DIY, and then retail is how that plan goes.)

 

Again, thanks to you all!

--Alex

[Superdad]

Hi guys:

 

Thanks for all the input, ideas, concerns, etc. We'll get it all sorted out soon. The nice thing is that with a female USB 'A' jack as the output--though our intention is to avoid the use on another cable since John went to the trouble of producing accurate output impedance and the idea is to have the signal recreated as close to the input as possible. Still, a short 4" cable is fine.

 

For fun I just weighed the Regen with case on my postage scale: less than 2 ounces; the A>B adapter (22mm center or 48mm end-to-end with the plugs) I have adds another 1/2 ounce. But as John pointed out to me last night, a heavy USB cable (and to a lesser extent the DC cable to the Regen) will exert some pressure.

 

Since very few DACs have their USB inputs with the 'D' facing sideways (unless their input PCB is vertical inside the DAC), use of an adapter is going to places the Regen in a horizontal position, with its 57mm x 18 mm oval face about 30mm from the back of the DAC's chassis. A small block of wood or foam underneath (maybe secured by a rubber band?) could support it for those who are concerned about forces.

Remember though, this device is going to live unseen behind your DAC, and will not be plugged and unplugged frequently like a dongle-DAC/headphone amp would be.

 

[Of course John--being John--suggested that we create and offer a 3D model file of a little cradle block with a formula to calculate its height based on user measurement of the lip of their DAC's USB jack; Then give them a link to one of the 3D printer service vendors for them to order their custom cradle block--in any color they wish--to snap in their Regen. I think I could cut a piece of foam and dig up a rubber band in the time it took me to describe his idea ]

 

I could include any sort of cable or adapter, but of course that just gets added to the price. And if I stock and include a bunch of pieces that cost me $6, they will have a $10 impact on the price. Just math and basic business.

 

Almost HAPPY NEW YEAR to you all!

 

--Alex C.

[Superdad]

So even with the adapter inserted into the DAC, that would be over an inch between the DAC and the face of the Regen.

 

I like the idea of a cradle, though I think to go along with the general aesthetic and design ideas behind my DAC, I'll likely have my own made out of wood.

 

Right, and if there are no other cables going into jacks closer than 1" to the left or right of the DAC's USB input, then there should be no fit/interference issues at all. If there are interference issues, then just use a very short cable.

 

And just to be clear, I will not be offering cradles!

 

--Alex

 

P.S. Jud: Have you listened to that Joe Henry CD I sent you yet?

[Superdad]

Hopefully there has been progress made on this product

 

Yes Jason, things are moving along nicely. Have added a power-on LED, and John improved the power distribution network with a 220uF cap. Have been listening to it with a Geek Pulse on my desktop system (after I replaced the tweeter the Pulse fried due to volume control/firmware bug). It is amazing the difference the Regen makes even with my AudioEngine P$ speakers two-feet away from me.

I am really excited about this device! Now that I am done getting a bunch more JS-2 units built and shipped, I can turn towards moving the Regen into production.

We promise it won't be long. You guys are going to LOVE this thing.

 

Have a great weekend,

 

--Alex C.

[Superdad]

We cant wait. Any idea of price ballpark?

 

Sorry to be slow in reply. I hope to be able to announce pricing within a couple of weeks. Waiting on vendor quotes.

 

On Monday I finally tested out in my main system the Geek Pulse DAC (delivered to me late December). Was fun to at long last be able to experiment with HQ Player converting Redbook to DSD128. Very nice, but very processor intensive.

Anyway, I tried the pre-production prototype USB Regen with the Geek Pulse. Oh yeah!!

 

Its funny, because the Regen cleans up so much that at first my reaction is "Can that be right?" It kind of takes away false "air" and "bloom" and that is initially confusing, as if you are loosing something. Plus it pretty radically alters perspective of the stage in a way that I can't describe but that you hear right away--like putting in a different DAC! In fact, for my system I had to overcome this by sitting about 1/2 meter closer to the speakers (I may want to reduce my almost flat toe-in of the speakers by a few millimeters to spread the stage). But instruments are still so right and intonation so clear and coherency increased--that you just can't point to anything wrong. And the minute you remove the Regen you realize how much it helps.

 

USB cables still make a difference, but with the Regen in the differences are not as stark. For example, I found the Lightspeed 1G that came with the Geek Pulse to be almost unbearable without the Regen after it. (The cheap 20awg Belkin Pro--I used it for years before getting the terrific Supra USB--is far more pleasantly balanced than the $59 Lightspeed 1G.) But with the Regen in I could make it through a couple of tracks before taking the Lightspeed back out.

 

Sadly, the Regen really does sound better powered by my JS-2 LPS than by the $10 Jameco linear wall wart we have been planning to include with the Regen. It is fine with the Jameco, but I got a big grin on my face as soon as I plugged the JS-2 into it.

(John and I are going to talk about alternatives--either pre-made or something small I could build to sell as a upgrade to the Regen. It is tricky, because as soon as I go to put something in a chassis with parts and soldering, etc. the practical price goes way up as I have no economies of scale--even at a couple hundred pieces. So I may just buy and try some other low amperage LPS brand--like a Teradak for $44--that might be worth recommending as an in-between. But it needs to be 6-9V.)

 

Anyway, that's the update for this week.

 

Cheers,

--Alex C.

[Superdad]

Is there an input voltage range for the Regen or is it only 6Vdc?

 

6V is about the minimum (because the regulators need to drop that to 5V to pass along--so supplying 5V is not going to cut it), but the regulators we use will handle upwards of 20V. However, I think we would prefer to see it supplied with 6-12V since the regs will get pretty warm dropping a higher voltage--especially if there is an actual load from a DAC that runs on USB VBUS (though I still think most of those are not supposed to draw more than about 800mA--it is just USB power after all).

 

For those just jumping in here, I offer the reminder that while the Regen does provide clean 5V USB power to DACs (or other devices) that need it, USB power is not its primary raison d'être.

My DAC does not use ANY USB power (I can cut/tape the power legs of my USB cable), yet I still enjoy the main function of the Regen. Which is to completely "regenerate" the USB signal right near the DAC--using the most minimal function USB hub chip, a good clock, and ultra-low-noise regulators combined with John Swenson's power network techniques--and to output it with at truly proper an controlled impedance (John makes a big deal out of the impedance aspect and why he uses a 4-layer board; I can't say I understand all of that part of it).

 

That said, the Regen does indeed provide very clean USB power and may deserved to be compared to devices that do just that.

John and I were talking last night about my experience with the Regen via little Jameco wall wart LPS I have here versus the UpTone JS-2. He said that although the Jameco is a linear supply, it of course does not have the 66,000uF of reservoir capacitance that the JS-2 does, and that is likely why I get a nice kick from the bigger supply. (The expensive regulators and on-board caps in the Regen doing a fine job of cleaning.) So he suggested that I experiment with some robust SMPS bricks that I might have around here. Some of those are rated for 3-5 amps or more and might actually sound better.

 

Still, the differences I heard between the Regen powered by the 6V/0.5A Jameco versus the JS-2 were not huge--the device still does its main magic in a big way from either supply. And I am loath to suggest people plug an SMPS into their audio system AC mains circuits--mainly because some of them pollute the power lines.

I'll keep everyone posted with regards further experiments.

 

--Alex C.

[Superdad]

I'm pretty excited about this...just wish it was ready!

 

Me too!

[Superdad]

How does the Regen compare to the AudioQuest Jitterbug?

 

Technically quite different. The Regen--with a USB hub chip, ultra-low-noise regulators, and low jitter clock--produces a completely new USB signal. Think of it as an outboard SOtM or Paul Pang USB card, but at the DACs input jack instead of inside the noisy computer. Schiit Wyrd is much more similar to Regen in terms of function (John feels the Regen has several advantages, but I have yet to compare them).

 

Sonic performance of AudioQuest Jitterbug versus UpTone USB Regen? Who knows--nether are shipping yet!

[Superdad]

Its unbranded, chinese made and model numberHR-1220 9v 1.5amp

 

Well that's the same case style, and in modest quantities the above can be sourced from China for $4 each (plug freight). But there is nothing to verify that its insides are the same as what AMR/iFi specs for their private label model and which I recall them claiming was a lower noise unit.

 

While I don't look forward to it, I may have to order in a bunch of different wall-wart and little tabletop types to both measure and listen to. And I just realized that if I supply a wall plugin "wart," I'll have to stock them in the various world plug styles. Or get the kind that offers snap-on plug plates. Hmm… the little tabletop kind with detachable mains cord is starting to sound like a better plan. We'll see.

[Superdad]

I think the main point of the DIY Swenson is the input filter inductor which prevents noise from traveling back into the AC ground.

 

Actually that is really the secondary purpose of the choke in John's PS design--though it, and the RC snubber he puts on the secondaries both do an excellent job of that.

 

The primary advantage over a single large C filter is that the remaining ripple is pure sine wave, which makes the regulators job MUCH easier. A traditional cap only filter (transformer, diode bridge, big cap) produces raw DC with a sawtooth riding on top. That sawtooth produces lots of high frequency components that the regulator has to deal with.

 

Traditional regulators do very well at low frequencies, but have lousy characteristics at high frequencies which means a fair amount of those high frequency components from the cap only filter get through the regulator. Those discrete regulators do well at blocking the high frequency components, but add cost and complexity to the PS.

 

As long as we are talking about power supplies, allow me to pass along a few more of John's opinions. Most cheap linear supplies are frequently worse at these goals than good switchers, but some switchers are REALLY bad. Since there is so much variability between PS implementations it's hard to make generalizations.

 

There is a group that has latched onto "lab supplies". There is no guarantee that a lab supply will be any better either. John has measured some lab supplies that inject just as much noise back into the mains as other supplies. And some are very good. But just going to eBay and getting the cheapest "lab supply" you can find is not going to guarantee good results.

 

Anyway, I am ordering up a variety of small Chinese tabletop models (you know, the ones with detachable AC cords--might make life easier for folks overseas, plus wall warts are always a pain to plug in). I'm going to try both regulated linears and SMPS. I just need to find out if any of those will sound different or if it is just that my JS-2 is so much better than any possible cheap supply. I am sure that is the case--and really for the Regen it is not huge--but I might as well try a few before I commit for a pile of them.

 

Oddly, one of my concerns is to find a decent one that is not more than 1.5 inches thick--so it will fit in a US Postal Priority Mail Small Flat Rate Box. I can send those anywhere in the world with partial tracking and full insurance for $23.50 ($19.50 to Canada, and $5.25 in the US). Wall warts with plugs don't fit and that jumps up overseas shipping by a lot.

 

Have a great week everyone,

 

--Alex C.

[Superdad]

Semi related:

Im experimenting with adding a ferrite bead at the dac end of my USB cable. There is a definite change (for the better - needs more testing…).

 

 

My dac is a Ayre QB-9 DSD that has its own 5V power with optical isolation (but needs the 5V for initial “handshake” with the computer).

 

 

What does the ferrite do here? What part of the USB signal is affected?

 

 

Funny you should mention this Fredrik. Just last week I dug out of a drawer some clamp-around ferrites (no autocorrect, not ferrets--those stink!). Tried them years ago on speaker, power, and interconnect cables and hated the results. But Miska was mentioning some positive things he measured with them on USB cables, so I thought I'd give it a try. I too found a minor benefit (nowhere near the scale of the Regen--which I was testing with Geek Pulse, power supplies, etc. in the same session--of course changing only one thing at a time). Miska says the ferrites knocks down some very high frequency hash. John was concern about what band that might be in (I'll ask him about it again), but I did not find that the clamp had a bad effect. Just was not that significant on my USB cables.

 

Here is a link to TDK brand clamp-ons that Mouser sells, sorted by I.D.:

TDK Ferrite Clamp On Cores | Mouser

[Superdad]

How does the effectiveness of your regen device compare with solutions like the JCAT or SOtM cards?

Would it make sense to use the regen with a Berkeley Alpha USB, or would it be redundant?

 

Hi Rick:

 

Thanks for the good questions.

The short and honest answer with regards to actual sonic benefit of the USB Regen in comparison to a JCAT or SOtM USB card or for use with a Berkeley USB>S/PDIF interface is:

"We don't know." Although every DAC I have tried the preproduction piece with has benefited greatly, I have neither of those USB cards nor a Berkeley unit.

 

The longer answer includes the following points:

 

a) Think of the USB Regen as an external JCAT or SOtM USB card--but out of the electrically noisy computer and PCI bus, close to the DAC, after the USB cable, and with carefully impedance matched output.

 

b) The Regen will likely still benefit computer>DAC connections that use a JCAT or SOtM in part because of factors mentioned in 'a)' and also because we use a very low jitter clock and ultra low-noise regulators to completely regenerate a new USB signal--and for those DACs that use 5V VBUS power for anything, the Regen provides really clean (and ground isolated) power as well.

 

c) While the Berkeley Alpha USB is one of the very best USB>S/PDIF converters out there, its USB input still consists of a USB PHY and an XMOS processor. They do a nice job running the I2S through an isolator and keep a separate ground plane, but non of that is much different than the USB>I2S boards we have played with--which the USB Regen still helps.

 

The primary purpose of the Regen is to increase "signal integrity" to make the PHY's job of decoding the data from the noise easier (read some of Swenson's many posts on this), and to potentially reduce packet noise as well (that is going to happen inside the DAC--typically around the 8Khz USB packet rate, but broadband too--no matter what, but the Regen may help).

 

Of course we all have to wait until the Regen begins shipping. Then lots of people can try it in lots of different applications and hear for themselves what benefits it might bring to their music systems. The price should (I am still waiting on a couple of quotes) be under $200, so it is not as if there will be a nick risk for folks to try it.

 

Best,

--Alex C.

 

P.S. Here is a pic of the USB section of the Berkeley Alpha USB:

(USB stuff all on the left; the isolator--type not known--is hidden under the U-shaped iron lamination stack separating the USB side from the S/PDIF side.)

 

[Superdad]

Time for an update!

 

Indeed it is!

 

Got the final quote yesterday and the order for PCB fabrication/SMD assembly is being place today! I'm crossing my fingers and going ahead with 100 pieces for the first run. Between these and another 100 pieces of the MMK boards, it is a really BIG bank wire transfer--gulp.

 

Unfortunately, China (yes, that's where everyone gets their boards stuffed, but the actual circuit parts are exact spec and shipped over there from Digi-Key) is going on New Year's holiday from February 14-27, so I won't have the boards until early March. That's probably okay since we still need to get the final chassis drawings (end cap openings and silk screen printing) and customization quote. So I hope everything will come together for production units to ship out to all you good folks by mid-March.

 

Should be able to announce the price in the next week or two. Will be under $200 for sure--just not sure by how much.

 

I can't wait to get these out as the positive effect is really easy to hear.

 

Have a great week gang!

 

--Alex C.

[Superdad]

OK, not everyone gets their stuff done in China btw.

 

For instance, Schitt Audio uses Jaxx Manufacturing, Inc. - Home I have about 4 that I've used in my state alone.

 

I know. It is just that every time we have gotten domestic quotes for PCB assembly, they either want runs of 500 units or they are several times the price of our Canada/China vendor--or both. But since you mentioned JAXX (I had bookmarked them months ago after seeing mention in Stoddard's very enjoyable blog), I went ahead today and submitted the Regen to them for bid. Would love nothing more than to have our boards made in California. But we'll see. I can't order on the scale of Schiit.

[Superdad]

^^^ My thought exactly. If you factor in the shipping costs and delays (be it in shipping time or time required to redesign and implement production changes), I'm not sure China is worth the trouble.

 

It was very well described in Schiit story (whch is an excellent reading BTW !):

 

Hi Adam:

 

Yes, I have enjoyed Jason's blog/book from the beginning. I got a lot of laughs from it since we lived through a lot of that sort of thing in my Hovland days (a room with $20K of scrapped custom machined chassis parts; not being able to ship due to just one stupid screw being back-ordered--and on and on).

Plus I have learned a ton from him. His recent "expose" of the costs and perils of working with Amazon? Priceless.

 

China can be tricky for chassis and custom stuff, or for complex boards where one may be changing parts and design. But for average stuff with a bunch of surface mount and a few through-hole parts, they are quite good, especially if you work with a reputable and reliable vendor. Such houses take care of all the parts procurement--buying and using only the exact items you specify.

Our turnkey firm, in Canada, analyzes everything (they will even catch mistakes in the PCB design), buys all the parts from Digi-Key or Mouser (whichever is cheaper and has stock), ships them to China where they make and photograph a first article. They test every board, pack them beautifully, and ship them to me via FedEx or DHL. During the process they even send me a link to a detailed tracking system so I know where they are in the manufacturing process.

 

How long does all that take? From the time I send payment to the time I have boards--just 14-16 days! The quality is really great and the prices are hard to beat.

 

I'd love to source more locally and will continue to try. For the R-core transformer we use in the JS-2, the one USA firm still making R-cores wanted 3 times the price of what I pay from China--and that was after shipping from China but without including freight for the USA vendor. And the lead time was 12 weeks as opposed to 3.

 

On the JS-2, the very high quality (finish, accuracy, machining, and printing) chassis I source from Japan costs me close to $160 landed, but I think it is worth it. Yet with a total parts cost of nearly $400 for the JS-2 (including cables and nice packaging, not including $50 royalty to John or anything for my labor), one can see that my costs are higher than the RETAIL price of the various typical transformer/diode/cap LPS units coming out of China/Taiwan. So I stay direct sale and concentrate on providing a flexible, high quality, high performance unit that people will be proud to own for a long time.

 

Still, I am somewhat baffled as to how Schiit Audio manages to domestically produce and sell some of their products for as little as they do. Big capital and economies of scale work wonders. Maybe I'll get there someday.

[Superdad]

Alex,

 

How do you make your drawings ?

 

With the good graces and talents of my dear friend, architect, and talented industrial designer Jeffry Tonkin. He was president of Hovland Company and the sole source and inspiration of everything beautiful that ever came out with the Hovland name. He is definitely stooping pretty low to do the CAD work for my current products, and for that I am grateful. Here is a sample of his capabilities:

 

[ATTACH=CONFIG]16855[/ATTACH]

 

[ATTACH=CONFIG]16856[/ATTACH]

[Superdad]

Alex, I get "Invalid Attachment" when I click. Please do attachments that can walk!

 

Yes, that is odd. They work for me, but I was puzzled that they did not show up as pic, just Attachment link names. Let me try again here:

 

 

[Superdad]

Hi Alex,

 

I'm hoping to receive my JS-2 and MMK in the next few days! Unfortunately it now looks like I won't be able to pick them up from my HK office till the month... Bugger!

 

In any case, please put me down for one of these too! Maybe it'll arrive in HK before the end of the month and pick up all the goodies at the same time!!!

 

Art

 

Sorry to hear that Art. As you know, it shipped on Monday and FedEx still has delivery set for Thursday.

 

As for the Regen, late March is when I am expecting to begin shipments to clients.

 

Best,

ALEX

[Superdad]

So when do you want real orders, and would it help that cash flow to have payment with orders (asssuming we know what it costs)?

 

Hi Bob. Since I keep an e-mail list--not used in months since my last 'newsletter', and now it is a very large list--I will send everyone an announcement when we are ready to take orders and payments.

 

Thanks for the offer, but I have a major ethical aversion to taking peoples' money prior to product production.

Even with the JS-2 LPS I avoid collecting funds more than a week or so in advance--and that I do only so I can reserve spots, assign serial numbers, configure mains voltage, and quote shipping costs.

Call me an anomaly in high-end audio.

[Superdad]

You mentioned that the Regen would be in the black box right ? Not that it is of importance to me, but...

 

Black body for sure. Have not yet decided on the color for the end caps. I am leaning towards the cobalt blue.