DDDAC1543 mark II, asking the forum for Inputs

[dddac]

As many people have built the dddac1543, I consequently received a lot of feedback on ways how to improve the design. Specially on the PCB lay out I had several.

Currently I am planning a version "mark II" where I will incorporate a number of ideas. As I do not want to rush into the new version, I like to ask the Forum inputs on 2 possible versions of the PCB design and values of the components find in the circuits below...

Next steps is to build a prototype with exactly the same components as my "reference" DAC at home, so a fair A-B comparison is possible between current and enhanced versions.

I am specially interested in feedback on the 2 versions of pcb design: One layer Starground or a 2 layer pcb with Groundplane.

Any other feedback is welcome, but please refrain from off-topic discussions about DAC chips for example for that we can open other topics

Looking forward to your feedback !

doede

[dddac]

And here the second option:

[2A3SET]

To get easier placement of heat sink on top of each 1543.

Regards

[pburke]

now you do this - and I JUST started soldering the thing together I have been collecting parts for since December

anyway - I'll send ample feedback

first of all, I think the PCB needs to be larger around the DAC tower to allow room for heat sinks that extend from the tower all around it.

Another thing - drop that battery checker section - who needs it? I read you can hook up a smart charger through a diode and never have to turn the thing off (or is that wrong? ever test it?)

I may just build the mk II if the mk I one turns out as sweet as everyone who has built it describes it.

more in a few days as I find a few resistors I fogot to order d'oh!

Peter

[dddac]

Originally posted by 2A3SET
To get easier placement of heat sink on top of each 1543.

Regards

I still like the tower concept for its ease and short connections, but it is true that you need to make sure it cools well...

8 chips next to each other on a double layer board is ofcourse les of an issue...

Thanks for the feedback, I will keep this one in mind. Hope to get some more inputs on this subject (tower or loose chips on pcb)

doede

[dddac]

Originally posted by pburke
now you do this - and I JUST started soldering the thing together I have been collecting parts for since December

first of all, I think the PCB needs to be larger around the DAC tower to allow room for heat sinks that extend from the tower all around it.

Another thing - drop that battery checker section - who needs it? I read you can hook up a smart charger through a diode and never have to turn the thing off (or is that wrong? ever test it?)

I may just build the mk II if the mk I one turns out as sweet as everyone who has built it describes it.

more in a few days as I find a few resistors I fogot to order d'oh!

Peter

Hi Peter, no issue, this will not be ready for a longer time, I am sure you will enjoy the current version.... and by then you might want to do some experimenting again
If the results are not clearly better than the current version, I will not move on to a mk2 version any way. But lets wait and see ! I see this as an experiment to look for bounderies and what can be done....

Larger PCB, yes, some extra space could be good,

The battery checker is bought by more then half of dac-builders, so why not?

[ThorstenL]

Konnichiwa,

Originally posted by dddac
Any other feedback is welcome, but please refrain from off-topic discussions about DAC chips for example for that we can open other topics

Some thoughts. Make the DAC differential and modular.

Use small vertical daugther boards, 4 TDA1543 each with local PSU Reg and decoupling, re-clocking and buffering with the 4 Chips on one side and all the other logic on the other side, make the whole thing differential to start, so the basic version has per channel 2 Chips in Parallel/Differential.

You should then be able to use the modern nice copper RRAM cooler thingies to keep the chips running way cool and you can make the DAC highly scalable, all the way up to (say) 32 Chips (8 modules in parallel).

Also, the three dimensional layout with horizontal base PCB and vertical DAC modules can be used to keep all current loops really tight and short. I would design this also such that you can use a pair of batteries, one for the Receiver et al and one for the massive parallel differential DAC, which will with 32 Chips draw the better part of 2.5A!!!!

Lastly, a modest RAM buffer is not that hard to implement, why not go for it?

Sayonara

[till]

yea, balanced!

[Urban]

Hi

Your DAC looks nice!

One problem for me though. I don't see how it can work...

You are using a clock in the DAC that runs async from the clock in the transport. I.e. the DAC-clock will either be running faster or slower than the transport-clock.

How can this work ???
Is the single reason for using the TDA1543 that it doesn't require a master clock ?

Confused

/Urban

[dddac]

Originally posted by Urban
Hi

Your DAC looks nice!

One problem for me though. I don't see how it can work...

You are using a clock in the DAC that runs async from the clock in the transport. I.e. the DAC-clock will either be running faster or slower than the transport-clock.

How can this work ???
Is the single reason for using the TDA1543 that it doesn't require a master clock ?

Confused

/Urban

Hi Urban,

it works fine, no problem, this is not a new design, I am just looking for some improvement on pcb layout, grounding and power supply etc

The difference between the clocks is not audible. If you use 12 or 18Mhz, you get some problems, but as they are very close, it is no problem. Interesting result however, is the more open/wide/deep sound stage and better pinpoint....

It works with all DAC's. I had one version with PCM63 as well... worked perfectly



bc
doede