Abraxalito mentioned that you have extensive knowledge relating to the TDA1543, and so I found a post of yours from some time ago. My experience with this chip is like yours, that it was relaxed and soft (with limited detail). This was somewhat expected for the reasons you describe.
My question relates to the reason for subjective limitations. Do you have any experience with I/V converters being attached to the TDA1543 with a more or less fixed input voltage and an input impedance perhaps below 10 Ohm's? I am wondering if subjective limitations exist in part for the reasons that many implementations fall far outside the +/- 25mV AC compliance limitations of the TDA1543.
It is considered that die sizes were much larger back then, whereupon dielectric constants of capacitances generating the bits could have greater influences in softness and masking. By terminating the motion with a low value load impedance of an I/V the influence of dielectrics would diminish, simply because they are no longer moving. This seems as the only chance the TDA1543 has in todays marketplace of alternative DAC's.
The answer to the first part is 'yes' - see my recent update in post #1 : lingDAC - cost effective RBCD multibit DAC design
As for the last part - I hope so but if you're new to SMT you might want to practice soldering a bit first as most of the parts (resistors, caps) are 0805 footprint which is 2mm long by 1.2mm. ICs are 1.27mm pitch so nothing very fine but compared to TH (through hole), half the pitch.
Nice!
That'll give me at least some time to practice smd-soldering, which is something that I dearly need... Sounds like a superinteresting project! Is there an estimate what the pcb's will cost and maybe a complete build as well?
Thanks for the answer!
/Hilding
Very glad to hear you find it so interesting.
PCBs - you can order those yourself when I publish the gerber file set. That's generally the cheapest way to get them - direct from the factory PCBs are almost unbelievably cheap nowadays and factories get good deals on shipping rates. Of course you can also get PCBs from me as part of a kit.
As for the complete cost I've not yet calculated it as there's a substantial labour element (my time) involved. I hope it will be under $100 (USD) for a kit. The kit will include the LC filter as a ready-built item and that's where most of the labour time is spent. The customer can choose between the two filters I've shown characteristics of in recent posts in this thread. The single inductor filter is quicker to build hence will be considerably cheaper.
PCBs - you can order those yourself when I publish the gerber file set. That's generally the cheapest way to get them - direct from the factory PCBs are almost unbelievably cheap nowadays and factories get good deals on shipping rates. Of course you can also get PCBs from me as part of a kit.
As for the complete cost I've not yet calculated it as there's a substantial labour element (my time) involved. I hope it will be under $100 (USD) for a kit. The kit will include the LC filter as a ready-built item and that's where most of the labour time is spent. The customer can choose between the two filters I've shown characteristics of in recent posts in this thread. The single inductor filter is quicker to build hence will be considerably cheaper.
Here's a shot of the first completed prototype PhiDAC hex r1 which played music first time on switch onI changed the colour of the nylon bolts through the inductor cores to closer match the sleeves of the 'lytics. What do you reckon - is black better than off-white?
The whole thing looks great. Black is better against the inductor core, creating an integral solid block look about the inductors themselves, as thereupon the whole of this board is more substantial looking. Otherwise, the dark inductors against the off-white nylon bolts creates a confused busy look generating an off-putting tension in the subconscious. Though one could ask "What does better mean?"
Here's a shot of the first completed prototype PhiDAC hex r1 which played music first time on switch on
White is ok, but black does match the inductors better
I have the board mostly ready. Just waiting on the new resistors and capacitors from Mouser. Should hopefully arrive beginning of next week.
I've turned my 2nd prototype PhiDAC hex into a 'DodecaDAC' - couldn't resist soldering on an extra layer of chips to get 12. The DAC chips need a reduction in supply voltage to achieve this result as there's otherwise more current than the I/V stage can reasonably handle. I've gone down to 2.35V supply for the TDA1387s and haven't noticed any SQ degradation nor measured any significant increase in distortion as a result of running the chips below their minimum.
Of course this is all just a distraction from getting the BOM finished....
Of course this is all just a distraction from getting the BOM finished....
There are always upsides to any downside - for example the people in India who can see the Himalayas for the first time due to the huge reduction in pollution.
If not for the virus I'd be still focussing on my transformer-based DAC design and getting kits out for that so this particular project wouldn't be in existence.
If not for the virus I'd be still focussing on my transformer-based DAC design and getting kits out for that so this particular project wouldn't be in existence.
Hey Abraxalito, quick question.
The resistors and capacitors for the quad version will arrive in a few days. I have a single dac chip soldered onto the board right now. When the other resistors and capacitors get placed on the board, I wanted to power it up and check that it all works before adding the other dac chips. Is there a way to measure each chip as I go for function? Should I just solder it ontop, leaving pin 6 floating and measure that one for voltage? What is the best approach?
The resistors and capacitors for the quad version will arrive in a few days. I have a single dac chip soldered onto the board right now. When the other resistors and capacitors get placed on the board, I wanted to power it up and check that it all works before adding the other dac chips. Is there a way to measure each chip as I go for function? Should I just solder it ontop, leaving pin 6 floating and measure that one for voltage? What is the best approach?