Hope this helps:
http://h5.m.taobao.com/awp/core/detail.htm?id=527172582335&toSite=main
Bom that I linked to meant to be as reference, it is from the original design which I’m planning to build.
Good luck with asking the seller🤣
Exactly what I was looking for ! Thanks very much Koen, much appreciated.
I’m sufficiently brave/stupid enough to think I can do the reflow surface mount work after acquiring a cheap toaster oven and a thermocouple probe.
It’s a shame there are no templates for the surface mount pads enabling a stencil to be etched so as to apply solder paste like a workshop does, but I’m encouraged by the YouTube clips showing solder paste messily applied but retracting onto individual pads under surface tension when melted.
Next step is to upgrade the resistors and order the remaining parts from Mouser. When they arrive, surface mount work first then the through hole stuff and then think about power supplies.
You’re welcome @andybob
For a stencil: Make a photo of the pcb, edit it, print it on a translucent sheet and cut the parts out.
That would be the easiest way I could think of.
I can recall that jussi said that he used his iron instead of the oven. That would be mine modus operandi too.
At the risk of confirming that I am a complete newb at this Koen, which resistors in the SMT BOM (photo below) should be replaced with the above type ? There don’t appear to be any of an 8.06k resistance. The 64 pieces of 15k seem the most likely contenders.
This appears to be the item, using a case code of 1206.
Edit: Ah, I see this is referred to in the ad linked in my first post. It is indeed the 15k resistors that are replaced by 8.06k types. Excellent.
I get the same results as you, it all depends on which ppm you need.
Over at diyaudio where I get a lot of my info (I’m not much better at this than you😜) they mention you can use 0604 0,1 SMD too
Thanks for the high expectations but I’m probably even more a noob than you are lol
Ahh I see you already figured it out, which ppm did you choose and why?
After filtering by case code (1206 inches), Mouser only had 25 PPM/C in stock. Which sort of narrows the choice. I was thinking of going with that.
yeah couldn’t get any wiser than that it has to do with how many parts vary per million units, as 25, 10 an 5 are already low ppm I would go with 25 too.
I initially thought it was a quality control parameter, but this paragraph explains it as more like the cooking or chemistry ppm. I suppose that way of expressing it helps more than a coefficient when quick guessing the consequences in a circuit of temperature change for a particular resistance value. Given that 50 ppm seems a typical figure for metal film types we should be fine with 25 ppm.
It is heartening to see parameters like this for components. Too many audiophiles place far too much faith in designs as distinct from implementation. You can’t buy a brick wall filter in a shop; you have to build it out of parts and make compromises when doing so.
Have you manage to setup up and make it work together with Roon+HQPlayer? Roon 1.3 allows PCM to DSD conversion and this project looks interesting. I wonder how it sounds like through a discrete digital to analog conversion compared to off selves like Burr-Brown, Cirrus Logic, Asahi-Kasei Microdevices and ESS Sabre DAC chipsets.
I haven’t built it yet so: “No” !
Of those who have built it, I haven’t read a review of people sending DSD 512 to it.
My whole purpose for building it is to upsample to DSD 512 in HQP, using a new server with a CUDA card to do so, and see how it sounds. I’ve seen the reports from people who I trust about improvements with DSD 512 using the T+a DAC 8 DSD but am too cheap to spring for the $4,800 AUD to check it out !
My poor dual core i7 BRIX wouldn’t be able to handle upsampling to DSD 512 without CUDA assistance, even if Roon were to implement it. So a new server will be needed.
It will be interesting to see what people think about upsampling in Roon in 1.3 as compared to HQP. I can hear differences but have not carefully controlled for volume so am pretty leery about attributing those differences to the software. My uncontrolled and completely unscientific first impression is that fewer links in the audio chain is a good thing.
I didn’t want to ask!
To someone with my skills/time/knowledge, it seems like an interesting project, bordering on torture. 
How much do you think the entire thing will cost in the end, including any power supplies and other necessary sundries (but excluding your valuable time of course)?
Kit - $35.66 USD
Extra/upgraded parts about $100 AUD
Tools, solder paste etc - about $250-$300 AUD including a new variable temp soldering iron, magnifier with built in light, toaster oven and thermocouple
New kit and parts after I make a dogs breakfast of the first - $150
microRendu & LPS - ah already gotta one, it’s verra nice !
Amanero USB to l2s Board - $106 USD (including shipping)
Nice case, about $50 AUD
Power supplies, haven’t decided yet. Something cheap to see if I can get noise out of it first up.
OK, wow, so when you take away the tooling (that you can maybe use again - possibly when you start selling the made-up kits in bulk 
), its pretty cheap! Seriously cheap actually.
Dare I say it, without electronics training and diagnostics kit, what would one do in the event that it doesn’t work or doesn’t sound ‘right’? Of course in your case this won’t be necessary and it will work fine, but as a general question…
The likely culprits for error are:
-
misplaced parts. Checking twice and avoiding intoxicating beverages during construction is the plan here.
-
bad connections. Carefully examining the SMT work under magnification after reflow for solder bridges is the go here. Desoldering wick and thin thin solder wire might help correct any, but I hope this doesn’t happen. Similarly tombstoning (where an SMT part rears up on one end because it wasn’t wetted down in the paste). I can just see myself ruining ICs with heat if I have to manually correct SMT reflow errors.
-
heat damaged parts. Diagnosing these is the worst case scenario. I will have to check voltages with a multimeter at various points in the circuit.
Electronics training … ya sure. I got electronics training _
I can talk the talk, but can I walk the walk ?
Thanks Andrew!
That’s informative and clear!
Last night I have crosschecked the bom I planned to use with jussi’s bom that is included, which is correct.
Costs for the components of the dac only are +/- 80€ from mouser excl USB board.
Time to order the PCB’s from easyEDA.com. 5 pieces will be 40€ delivered.
From what I heard/read, the psu is a very important factor for a low noise floor. When using the Amanero the good results where achieved with the free isolator from diyinhk and a separate psu for the amanero.
Probably going to use the original dsc1 psu
For people that are interested I will have some boards extra
I’m likely to be interested in the PSU, but let’s see if you can find folks who’ll buy both before we split any pairs of boards.
I have ordered the dac boards 
PSU I will order later.
You may have one of the boards Andrew, but assuming that you live in another part of the world I guess it would be cheaper for you to just order 5 yourself.
DAC PCB is 25€ for 5 plus shipping makes 40€, PSU PCB will be around the same. Of course I chose the cheapest shipping so ETA will be around this time 2018😩
1 Like
This capacitor code table is helpful (necessary) to decode the BOM into a Mouser order. Where there is a P suffix I understand that it means picofarads so WIMA 100P means 100 picofarads. WIMA 152, however means 1500 picofarads as set out in the table.
Correct, my trick is (which in most cases works) is 3 number is amount of zeros.
But my knowledge is very rudimentary
1 Like