No business, I'll post the BOM and who likes/trust it buy things for himself...
Regarding mod levels, I'll post also those.
The 7812 for sure needs an heatsink...
It sounds good but a bit unbalanced, until LPF is corrected.
All caps did burn-in, between 50-100 hours.
Unfortunately I don't have a scope but if you worry about rectification noise I've, at least partially, managed it with a10nF X2 cap between AC rails of rectifier.
Probably better diodes (MUR820 or SBYV27) would give another step-up in performance.
hi guys, i just soldered everything but psu caps, just to test it so far... anyway i've tried it off 12v battery in place of 7812 and directly to 7809.
lm1117s on 7812 side measure correctly each at 3.3V and 5V, but on the 7809 side i only got 1.1V? and even heatsinked 7809 and it got awfully hot in a matter of seconds?
could it be current rush as i put 3 oscons 680uF after 1117 regulator and 220uF panasonic after 09 :/ cs uses about 30mA, whole dac uses 200mA(mostly analog section) and none of the leds light up
i've measured all caps prior to soldering and double checked polarities
okay i've measured again 7809 and seems like its putting only 1.2volts out? i've wired it correctly..oh i guess i'll have to try it with lm317 then
could it be internal thermal/current protection? but then i guess it wouldn't output anything at all..
thanks a lot guys!
lm1117s on 7812 side measure correctly each at 3.3V and 5V, but on the 7809 side i only got 1.1V? and even heatsinked 7809 and it got awfully hot in a matter of seconds?
could it be current rush as i put 3 oscons 680uF after 1117 regulator and 220uF panasonic after 09 :/ cs uses about 30mA, whole dac uses 200mA(mostly analog section) and none of the leds light up
i've measured all caps prior to soldering and double checked polarities
okay i've measured again 7809 and seems like its putting only 1.2volts out? i've wired it correctly..oh i guess i'll have to try it with lm317 then
could it be internal thermal/current protection? but then i guess it wouldn't output anything at all..
thanks a lot guys!
Last edited by a moderator:
Clave - That's a minefield of testing you've carried out - Well done !
Can I ask one thing tho - The differences you noted during the useage of the caps mentioned (in the paragraph above) - is this down to the quality/material of the cap or the value, or even a combination of the two? Would anyone care to comment further?
Thanks 🙂
All of them are Solid Electrolitics, so pretty similar.
SEPC are the only ones that have copper leads (IMHO from this cames their sweeter/warmer sound).
Probably value also has a role but I've selected them with this rule: the bigger value with 6.3mm diameter which I can find.
BTW it seems that over 100uF there's no much difference...
I came home today an switched on the system (dac left 'on' for burn-in with ipod random track playback) only to find distortion on music peaks. I had the dac & pwr supply housed away safely in a make-shift housing box, and checking everything else found the voltage regulator (the one that gets kind of hot) REALLY hot. Disconnecting the pwr supply for a while and then powering back up again (with box lid removed!) restored normal operation.
After hunting around for a heatsink to fit, I managed to knock-up this little make-shift piece while a better solution arrives in the post ! 😀
..an before you ask...yes, there is heat transfer compound in there as well 😉
Also, I've completed some more changes as per Clave's BOM. It's a very detailed sound but more burn-in time is required as it's now sounding a little edgy on the top end. Bass is somewhat improved tho !
-A-
Last edited:
Mmh, 3300uF Rubycon ZLs after regulators.... it's a no-no!
They're ultra-low ESR caps! And with that value poor regulators pass all the time charging them... no surprise they became hot!
Also edgyness could be due to them.
You should use a regular elco between 47uF and 220uF (Jean-Paul suggests the low side), BCs are good, Silmics, Cerafines, KWs or KZs even better.
In this post you'll find a link to a suitable heatsink... your it's a real 'heat-pipe'... 😀
Last edited:
oh i just found that dac pins 11 and 12 are shorted :/ uncareful soldering job whoever has done it...
11 is grounded - and 12 is on digital 3.3V ..i just hope chips survived this...
11 is grounded - and 12 is on digital 3.3V ..i just hope chips survived this...
Interesting....... I have left my DAC powered on for 70 + hours straight, and no problems with the 7812 and I don't use a heatsink.
Yes its gets quite hot, but no problems at all.
Yes its gets quite hot, but no problems at all.
Post regulation Advice..
Ok - they're gone. Replaced with Nichicon 100uf/50v (same type as was on pcb originally, but 100 instead of the 220uf)
These were already in circuit prior to the last mods, and no edginess apparent. It's getting better tho as time & burn-in takes place
A piece of microbore pipe would normally be utilised in heating systems - Heat pipe for sure ! 😉
Cheers for the heads-up Clave !
Mmh, 3300uF Rubycon ZLs after regulators.... it's a no-no!
They're ultra-low ESR caps! And with that value poor regulators pass all the time charging them... no surprise they became hot!
Ok - they're gone. Replaced with Nichicon 100uf/50v (same type as was on pcb originally, but 100 instead of the 220uf)
These were already in circuit prior to the last mods, and no edginess apparent. It's getting better tho as time & burn-in takes place
A piece of microbore pipe would normally be utilised in heating systems - Heat pipe for sure ! 😉
Cheers for the heads-up Clave !
'No Bypass' on B.O.M
Clave -
On your BOM of parts upgrade you have listed 'No bypass' (i.e. C11, C12, C15, C18, C19,C20). 😕
-'Position left void'- Meaning....you've removed the original part and left the location 'empty' of any component ?
Cheers,
-Andy-
Clave -
On your BOM of parts upgrade you have listed 'No bypass' (i.e. C11, C12, C15, C18, C19,C20). 😕
-'Position left void'- Meaning....you've removed the original part and left the location 'empty' of any component ?
Cheers,
-Andy-
Long term thinking should be the way to go when designing/building equipment. Maybe not in your case but small equipment sometimes stays powered on and in case of failure the results can be unpleasant. Way more unpleasant than adding a 1 Euro heatsink.
Electrolytic caps don't like heat. The less heat they get the longer they will live while functioning correctly during their life span. The reg probably will survive but still: heat is not recommended for longevity. The buffer caps are very close to the reg. For very little money you won't ever experience a problem due to heat and no need to replace too early dried out caps. The DAC will be built in a closed case in 99% of cases so no ventilation contrary to an "open" construction for testing.
Easy mathematics IMO. Just a small piece of aluminium with a 90 degree angle is enough.
Last edited:
Exactly 😉
I've attached a real BOM with RS Components, Distrelec and Mouser codes.
Please note that BC038s are not available from Mouser and have been substituted with Nichicon FWs which should be equivalent or even better, some codes are still missing.
Some caps are easier to find on eBay (like SEPC and BC036) or from PartsConnexion (Silmics, Cerafines, FWs).
Well spotted...
Good work. It's a nitemare trying to trace a fault.
Well, you can now at least start to hear what the dac has to offer eh ?
Lets us know what you think...
-Andy-
Good work. It's a nitemare trying to trace a fault.
Well, you can now at least start to hear what the dac has to offer eh ?
Lets us know what you think...
-Andy-
hi, and thanx every one on this thread. i wonder about the output voltage of this dac-board?
i got a pair of Sowter 9545. somebody mentioned it before and i want to ask the masters here. is it a good marriage withe this board? anything to concern about? (i'm not that technical)
output voltage of the chip is +/-2.8v! what will happen to the 1:5 ratio?
i got a pair of Sowter 9545. somebody mentioned it before and i want to ask the masters here. is it a good marriage withe this board? anything to concern about? (i'm not that technical)
output voltage of the chip is +/-2.8v! what will happen to the 1:5 ratio?
BOM Update
I slightly revised the BOM, attached.
Some additional info about some caps choice:
BC 038 (high voltage rating versions) are really good caps and sounds great in a lot of positions and have a very low DF (0.6 to 0.9).
I encourage American friends to buy them from eBay if they can't find a local source.
I've tried Nichicon VRs, also good caps, but BC038 are WAY better.
Regarding solid elcos SEPC, LS and LF are all great sounding caps, choice is a matter of taste.
SEPCs are warm and round, LSs open and snappy with LFs a good mix between the other two.
Some may think that is strange that I've preferred BC038s in C3 position but after all this is an ANALOG supply, isn't it? 😉
As always, though advanced, this is a work in progress...
In fact LPF still needs attention.
I've tried 3.3nF Amtrans caps from parts connection.
Not cheap at all but those are GREAT caps
I slightly revised the BOM, attached.
Some additional info about some caps choice:
BC 038 (high voltage rating versions) are really good caps and sounds great in a lot of positions and have a very low DF (0.6 to 0.9).
I encourage American friends to buy them from eBay if they can't find a local source.
I've tried Nichicon VRs, also good caps, but BC038 are WAY better.
Regarding solid elcos SEPC, LS and LF are all great sounding caps, choice is a matter of taste.
SEPCs are warm and round, LSs open and snappy with LFs a good mix between the other two.
Some may think that is strange that I've preferred BC038s in C3 position but after all this is an ANALOG supply, isn't it? 😉
As always, though advanced, this is a work in progress...
In fact LPF still needs attention.
I've tried 3.3nF Amtrans caps from parts connection.
Not cheap at all but those are GREAT caps
Last edited:
Alas, I've not managed to get as far as to connect up the sowters -concentrated on getting the basics of the on-board mods completed first to appreciate the gains therein.
But, the loading resistor mentioned for the output can be negated by the fact that the transformer will itself present a load. But, I've been advised that a loading resistor needs to be placed across the secondary output of the Sowters. The beauty of this approach is that the sowters will still be utilising the 'balanced' output operation of the AKM chipset. Some, have had mileage here with a single dc blocking resistor straight off the either of the '+' or '-' output pins, but this again is side-stepping the Low pass filter which surrounds the currently resident opamp. What gains/qualities are brought about will be down to your listening preference..
I found some KA7812 and KA7912 regulators in my spares boxes. Possibly made by Fairchild.
Are these likely to be any better than the ones fitted to this board?
Are these likely to be any better than the ones fitted to this board?