GB interest check on HP amp and USB DAC

[GearSlave]

OK guys, this project has received some revived interest of late (especially from skinny, Mr. Slave Driver ).

To paint a little picture of what this is going to end up as:
It'll be a USB DAC based on the PCM2702 with a TPA6120 driving the cans. This amp will have a little switch to select either USB or line in if you have another source. I've done the proto part and tested it with very good results (verified with RMAA), so now comes the fun part, having professional PCB's made. Initially it was only meant as a quick little project for skinny and myself, but to get the best possible performance you have to use a double sided SMD board. Also, the pitch on the SSOP28 DAC chip is very fine so you can't really home brew the boards. It stands to reason that such a board is not cheap in small quantities, so I'm doing a little interest check to see who would be interested in such a little project. Any thoughts/ideas/suggestions? I'm hoping to get a price break where the boards can be done for under R150 each (including tooling charges).

I'm still capturing the schematics and hope to be done this evening. I'll post it here for any possible input / criticism (I'll ignore BAD criticism ) a little later for you guys.

Let's try and see if we can have our first AVForums Project?!

Thanks

Gert

[Viagara]

After the good results I get from the Soundblaster Live! USB sound device, I would be keen to get my hands on your product, so count me in.

[skinnyfat]

ummmmmmmm, let me see, can I get back to you ? ?



Seriously guys, I would suggest you take advantage of this little development. I know Gert has been putting in some late hours on this one (thanks to some encouragement from me and a bottle of 12 yr old single malt <we wont elaborate on what happened there > ) and I would like to see this as the first project offered to AVForum members

[GearSlave]

Please bear in mind that the price given is for just the PCB's and that there is some SMD work to be done. If it's a small amount, we can make a plan so that I can mount the SMD parts. Also, all the expensive parts are available as samples from TI, so that also helps. I envisage the total project to not be more than around R600.

Also, any suggestions or improvements are welcome

[GearSlave]

ummmmmmmm, let me see, can I get back to you ? ?

funny bunny

[skinnyfat]

funny bunny

LMFAO LOOOOL, heh heh heh

[GearSlave]

OK, this is what I quickly put together.

I shamelessly copied a lot of what was done before and just brought it all together. Because of the low impedance requirements of the TPA6120, I included a buffer stage so that it can be used with line sources if you don't want to use the USB dac. It can be excluded (saving an buffer) if you only intend using it with the DAC.

Like I said, I've done measurements with a crude prototype and the results was very encouraging. I'm itching to see the performance once it's properly constructed on a nice PCB with SMD parts.

Again, comments are welcome

First image is the USB DAC portion, the second is the HP Amp and the third is the PSU part. The only funny parts are the PCM2702, TPA6120 and the TPS73033 3.3V regulator, but like I said, all these fparts are available as samples from TI.

[Family_Dog]

I don't see any valves? 


-F_D

[Hi-Phibian]

I am in, how many people do we need to get to the PCB price break?

[GearSlave]

I don't see any valves? 
-F_D

It's interesting that you mention that I'm going to make the DAC and HP board separate, so you can conceivably use a valve stage instead of the TPA6120 if you wanted to. I'm not too hot (pardon the pun) when it comes to valves, but you welcome to propose something and we can work something out. It'll be interesting, to say the least.

@croak
I got a quote last night from a fab house in China. It'll come to around $190 for 10 boards and that includes tooling and shipping via DHL. So that comes to R130 each and add 20% worst case for duties locally so you in for about R160 for the PCB. So, I'll need about 10 guys interested.

So far:
Me
Skinny
Abe
Croak

That's 4 boards (duh). I still need to complete the layout of the boards, so I'll let this run for a week or two. If I get around 7 guys interested I'll order the boards; I know once these are completed and we have a few happy campers out there the last ones will be snatched up quickly.

I'll be going to Communica today to go get some parts that I need to confirm the final layout, so I should be able to post a proper BOM soon for you guys. My aim is to not use exotic parts so that you can get all the stuff locally. The TI parts you'll have to order as samples - I've ordered my quota already and they won't ship me anymore - but it's very easy and they are very fast A lot of it will be SMD though, this is unavoidable for best results unfortunately.

Cheers

Gert

[Hennie]

For a current output DAC:

- a resistor, followed by a passive filter stage and a valve amplifier, like this:

http://members.chello.nl/m.heijligers/DAChtml/dig_r2b.pdf

http://members.chello.nl/m.heijligers/DAChtml/suphv2b.pdf

It is not as mad as it looks. A virtual ground opamp based I/V does not provide a linear, low impedance load to the DAC at the higher end of the spectrum. It won't be quite a virtual ground anymore. There will be HF intermodulation but how much of it folds back into the audio band I am not sure. Probably not too much but it would be interesting to see a comparison with a valve stage. The design referenced above limits the input spectrum to the valve stage with a passive filter.

[GearSlave]

Nice comments, thanks Hennie.

The above DAC is voltage output though, not current output, thus no I/V stage. The low pass filter stage after the DAC is simply to remove some residual HF components that still exists in the output of the DAC. A valve stage is still a possibility though

[Hennie]

Sorry Gert, I was looking at the MFR's application circuit for the headphone driver which had an I/V DAC connected. I should have looked at your circuit.

With the voltage output and the 2134 opamp design you have, I am not sure that a valve would have any benefit other than a warm feeling that there's a valve somewhere.

Another comment:

 - the great German headphones all require an output impedance of 150R or 120R (not dead sure which but I think it is 150R). With too low an output impedance they tend to sound slightly bright and slightly bass shy.

[skinnyfat]

Come on guys, LOOOOL who else is taking one, I got no more patience

[Viagara]

The projected price is close to the price of a SB Live USB and I would expect that the quality would be better than the SB, so yes I am still in.

Just a question though. You mention USB and Line-level input. Am I confused or would SPDIF make more sense than Line-level? Or would the Line-level input bypass the DAC?