[Quiet Earth]

Other than the Musical Fidelity V-Link series, are there any other non surface mount constructed USB to S/Pdif (AES) converters being made?

[Quiet Earth]

Ok, well I found another one. It's called the Audio-gd Digital Interface. No AES output though, but with optional power supply. Looks too good to be true at that price. http://www.audio-gd.com/Pro/dac/USBface/Digital1EN.htm

I guess most people don't care if their state of the art USB interface is made with surface mount technology. Or maybe it doesn't matter ?   

Still searching . . . . .

[srb]

I guess most people don't care if their state of the art USB interface is made with surface mount technology. Or maybe it doesn't matter ?   

According to an acquaintance who was the Demo Lab hardware specialist at Burr-Brown, good digital design keeps board trace length, component leads and component size to a minimum.  As long as close tolerance surface mount parts are used, I think it is a superior circuit topology, but just not as good for DIY rework without the proper equipment or skill.
 
Steve

[randytsuch]

I'm wondering why you care if it's surface mount or not?

These days, there are many parts that are only made as surface mount, the audio gd box you linked to contains surface mount parts.   I kind of doubt that you can find a USB audio device that is not surface mount.

Randy

[Quiet Earth]

Excellent points Gentlemen. Glad you replied.

First off, let me say that this is mostly something that I noticed as a general rule, and it sort of got me wondering. Second, I realize that this stuff is changing on almost a monthly basis, so it probably doesn't make sense to build a really nice converter by hand with premium parts when you are already designing version two and three.

- srb,

I totally understand what you are saying. I'm not sure that such a low frequency really needs a perfect 75 ohm (or 50 ohm) transmission line, but I suppose it wouldn't hurt. Even if it did require it, not very many of the boards that I have seen look like they adhere to the kind of rigid standards that you see in the microwave industry. Maybe some of them are built with that sort of thing in mind and I am just not seeing it.

- randytsuch,

Good question. And you are right. There is still the integrated circuit or the odd part that needs to be surface mounted anyway, so why not just carry it all the way through? I guess the main reason why it bothers me is because in general, SMT is mostly done as a cost saving strategy. It's a good way to cheat labor and get consistent quality out of lower cost manufacturing. I'm all for it, but it's not exactly what the advertising copy reads for the latest audiophile converter, is it?

I am also very leery of the sound of some of those parts. Not that there's anything wrong with them doing their job, but because tantalum capacitors, thin traces, and little tiny film resistors don't usually add up to natural sound.

I am totally open to the fact that I could be way off base here. That's actually why I brought it up in the first place. And also out of curiosity.

[srb]

I am also very leery of the sound of some of those parts. Not that there's anything wrong with them doing their job, but because tantalum capacitors, thin traces, and little tiny film resistors don't usually add up to natural sound.

Perhaps, but in this case you are referring to an all-digital USB to S/PDIF converter, not a complete DAC with an analog output stage that might have greater current requirements.
 
Steve
 
 

[randytsuch]

My 2 cents on the subject.

For a USB to SPDIF converter, as Steve said, everything is digital, and there are different issues than with a DAC, where you need an analog output stage.

For high speed digital circuits, SM caps can be better than leaded.  Leads add inductance, which is bad at high frequencies.

Also, SM is only cheaper and easier if you have access to special machines, that automatically place the parts.  I would guess that in some places in China, where labor is cheap, they can do through hole cheaper then SM. 

In the end, the design and implementation probably make more of a difference then SM versus leaded, IMHO.

Randy

[Tam Lin]

The upper photo appeared in a more technically oriented forum as an example of how not to do things. It’s a $10,000 CD player. Whatever the expected advantage provided by the fancy red caps is more than nullified by the extra inductance created by the long, skinny traces, vias, and component leads.

Every digital audio chip data sheet advises the bypass caps should be as close as possible to the IC pins. The lower photo is how it should be done: Short, wide traces, caps close to the pins, and no vias. (C329, C331, C333, C334, C336)

There is no cost saving using surface mount unless the PCB is machine stuffed and that is only cost effective for large production runs. That’s why limited production audio gear uses through-hole components. It also gives the manufacturer the ability to post pretty pictures with colorful components and brag about boutique brands because that’s what their audiophile customers demand.

[art]

WHY!!!!!!!! You are joking, right?

Let's see...........how many non-SMT USB controller chips are there?

"Uh..............zero?"

Yes!

And how many non-SMT I2S-SPDIF chips are there?

"Uh.......same answer?"

Correct-a-mundo.

BTW........has anyone noticed what happened to the price of through hole chips (like PCM1704) after they were made RoHS compliant? The cost went way up, as the costs of complying with that useless edict is now spread out over few chips.

The industry is going SMT. The availability of through hole is rapidly shrinking. Whether or not, as a manufacturer, I like it, is moot. There are few options.

Why you, as a consumer should care.............I have no idea. Your choices are less than ours. Welcome to the real world, bub.

[Quiet Earth]

not joking

just genuinely curious

of the correlation between SMT and generally poor sound

i have no idea why

you got all bent out of shape

but thanks for the warm welcome

anyway

[art]

For analog, you know - audio, yes, I would be very vary of SMT stuff. Especially the resistors. Add to that lots of folks use nasty ceramic caps, especially around the DAC chips and I/V and/or post-filter stages.

But, for the front-end, there are no through-hole alternatives. Since the subject is USB (to SPDIF) converters, it is all RF, and SMT has lots of advantages. If the subject was USB DACs, might be slightly different subject. Except all the DAC chips are all SMT.

Ok, in theory, you could use a PCM1702 (I mistakenly typed '1704, in previous post), but they are getting harder to get, and the price goes up weekly. Or so it seems. Not too long ago, I had to purchase some, to fix a customer's DIY DAC. Paid $48, which was up from $40, the last time I bought any. They are now over $50.

Compare that to $25, for a '1794, and under $10, for other parts in the family.

[Quiet Earth]

That makes more sense art. Thanks man. And I appreciate all of the replies so far. Food for thought.

Still having a little trouble regarding digital as data or RF, but I am coming around to it.