Got mine today from diyhifishop.com.
Dull surface on the bottom covered with some sticky residue, like grease.
Smells like rubber. Maybe it was the foil they wrapped it in.
Nevertheless I took the effort to measure the board for those being interested.
The fixing holes are not placed symmetrical but there seems to be no reason for this by the layout , besides a fixing hole close to the AC input where a capacitor is in the way.
Maybe another hint that the board is kind of "universal" and that he fixing holes got their strange positions from an earlier design/application with different traces.
I have no transformers yet to provide power and check performance...
I would have loved to buy another ES9018 Dac from Twistedpear (own two of them) but for my current project the pcbs of the power supplies and output stage would take too much space.
Dull surface on the bottom covered with some sticky residue, like grease.
Smells like rubber. Maybe it was the foil they wrapped it in.
Nevertheless I took the effort to measure the board for those being interested.
The fixing holes are not placed symmetrical but there seems to be no reason for this by the layout , besides a fixing hole close to the AC input where a capacitor is in the way.
Maybe another hint that the board is kind of "universal" and that he fixing holes got their strange positions from an earlier design/application with different traces.
I have no transformers yet to provide power and check performance...
I would have loved to buy another ES9018 Dac from Twistedpear (own two of them) but for my current project the pcbs of the power supplies and output stage would take too much space.
Attachments
Last edited:
Checked the Weiliang today - WOW -but not in the way one might expect.
It sounds extremely DULL!
NEVER heard such big differnce to other DACS EVER!
Compared it to Sony CDP-X5000, Buffalo DAC II, even a 33 year old CDP-101.
Even checked cables.
The dullness is so strong that I thought that de-emphasis
was switched by some fault. So i took Sony YEDS-18 test disc that has emphasized tracks, but no difference when other Players have de-emphasis on. The Weiliang still sounds dull.
I can´t imagine that the different filter types of the ES9018 would have such an audible effect. Maybe bad implementation of opamps?
Opamps are NE5532 / NE5534, maybe generic, company JRC
Playing Silence and ramping Volume all way up, the Buffalo II and Sony X5000 are dead silent. Hum on the Weiliang. The Sony uses regular voltage regulators 78xx / 70xx as well as the Weiliang
This is what Weiliang / diyhifishop (or "minishow" on ebay) sent me as schematic
It sounds extremely DULL!
NEVER heard such big differnce to other DACS EVER!
Compared it to Sony CDP-X5000, Buffalo DAC II, even a 33 year old CDP-101.
Even checked cables.
The dullness is so strong that I thought that de-emphasis
was switched by some fault. So i took Sony YEDS-18 test disc that has emphasized tracks, but no difference when other Players have de-emphasis on. The Weiliang still sounds dull.
I can´t imagine that the different filter types of the ES9018 would have such an audible effect. Maybe bad implementation of opamps?
Opamps are NE5532 / NE5534, maybe generic, company JRC
Playing Silence and ramping Volume all way up, the Buffalo II and Sony X5000 are dead silent. Hum on the Weiliang. The Sony uses regular voltage regulators 78xx / 70xx as well as the Weiliang
This is what Weiliang / diyhifishop (or "minishow" on ebay) sent me as schematic
BuffaloII DAC has ES9018, so I know how good it sounds.
Does not sound like ground loop. Again, I simply played silence (Digital Zero) and raised the amps volume to max for that time.
Can opamps change the sound that dramatically? Never heard it that drastically.
But the feedback between the opamps output to the inverted input looks strange to me. I do not understand the two resistors and one capacitor between opamps output and inverting input...
Does not sound like ground loop. Again, I simply played silence (Digital Zero) and raised the amps volume to max for that time.
Can opamps change the sound that dramatically? Never heard it that drastically.
But the feedback between the opamps output to the inverted input looks strange to me. I do not understand the two resistors and one capacitor between opamps output and inverting input...
I had that board year or 2 ago....still have supply part. If you take a look my post i write down about that. First change R0 around lmv because spdif input termination is 75 ohm not 10kohm. About gnd. It is mess you must cut a few lines and connect it different. Something was also wrong with 7pin connector and supply...reference for 1 opamp must be better filtered . Unsolder that 100ohm 100p combination because i am not sure that 100p place did have cap on that place i think it was resistor. Solder at least 220p or more to do some filtering on 1 opamp. Second part of filter must have 100n to gnd and 10nf on feedback.... good luck
Last edited:
Hi Androa76,
thanks a lot for your reply. I read your post mentioned but did not understand it.But at least 75ohm termination is correct on my board..
I made a jpeg from the original schematic.
Maybe you can mark the areas
or even draw in the new connections...?
All the best and thanks for the effort,
Salar
thanks a lot for your reply. I read your post mentioned but did not understand it.But at least 75ohm termination is correct on my board..
I made a jpeg from the original schematic.
Maybe you can mark the areas
or even draw in the new connections...?
All the best and thanks for the effort,
Salar
Attachments
Interesting...i have found picture of mine...and is also v.1.1. but not the same as yours...
My has no 100n decoupling near ne5532...your termination is correct 75ohm.
on your schematic LC4,5 and RC4,5 are 102 ...it must be 104 on board. take out LC, RC10 and LC,RC2 and measure if the small one are capacitors. If this 2 are resistors than feedback is not right. Then we will go on.
Gnd under is hard to see...clean pcb with isopropyl. I think it goes, from rca out till 5pin vias into Kubota regulator...that is ok...mine is going opposite ...into es9018 3,3v gnd.
I had at least 7 different es9018 dac boards and commercial products ( i still have one 9018k2m dac lying in the table waiting for...) but i was never satisfied...sound is to unnatural sharp focused, full of details...a lot of people like that sound because on first note it has wow effect...but when you compared it to high class dacs you hear that real life world is not so sharp... there are good dacs on the market with ES9018 but i have not heard them. My last comparison was Total dac vs one high class es9018 dsd dac from this forum ... with discrete output not Lundahll... on acoustic arts high end equipment and Soulsonic speakers.
Dull sound with es9018 ...hard to belive.
My has no 100n decoupling near ne5532...your termination is correct 75ohm.
on your schematic LC4,5 and RC4,5 are 102 ...it must be 104 on board. take out LC, RC10 and LC,RC2 and measure if the small one are capacitors. If this 2 are resistors than feedback is not right. Then we will go on.
Gnd under is hard to see...clean pcb with isopropyl. I think it goes, from rca out till 5pin vias into Kubota regulator...that is ok...mine is going opposite ...into es9018 3,3v gnd.
I had at least 7 different es9018 dac boards and commercial products ( i still have one 9018k2m dac lying in the table waiting for...) but i was never satisfied...sound is to unnatural sharp focused, full of details...a lot of people like that sound because on first note it has wow effect...but when you compared it to high class dacs you hear that real life world is not so sharp... there are good dacs on the market with ES9018 but i have not heard them. My last comparison was Total dac vs one high class es9018 dsd dac from this forum ... with discrete output not Lundahll... on acoustic arts high end equipment and Soulsonic speakers.
Dull sound with es9018 ...hard to belive.
Here comes the proof:
https://www.youtube.com/watch?v=Hjb5wNw4EvE
The dullness is still audible despite the fact that
-the original recording played back through the DACs was made using a Zoom H4n, not High End studio equipment
-I matched loudness (the Weiliang is 1.4db quieter)
-used 20 year old equipment for recording the analogue outputs
(Audiomedia III card)
-recorded "only" in 16 bit / 44.1
-the sound was further compressed by youtube.
I am using a no name Gainclone, JBL Control 1, an Audio Alchemy
DAC in the Box hooked to my laptop and can still hear the difference.
Last edited:
O.K got it, it is 0.01 µf and a resistor of 300Ω.
On my board it is LC4 / RC4 and LC5 / RC5 have a value of 0,15µF
Resistors LR16 / RR16 and LR20 / RR20 have a value of 22Ω
This makes a low pass starting at 48kHz
According to the schematic the values should be 22Ω and "102" = 0.001µf.
This makes a low pass starting at 723kHz.
But the markings on the board state "103" = 0.1µF
Low pass starts at 72kHz
Will replace the 150nF with 0.01uF and see what happens.
BTW, sent the video to diyhifishop, they asked which kind of difference was meant. So they at least pretend to be deaf. No comment
On my board it is LC4 / RC4 and LC5 / RC5 have a value of 0,15µF
Resistors LR16 / RR16 and LR20 / RR20 have a value of 22Ω
This makes a low pass starting at 48kHz
According to the schematic the values should be 22Ω and "102" = 0.001µf.
This makes a low pass starting at 723kHz.
But the markings on the board state "103" = 0.1µF
Low pass starts at 72kHz
They are capacitors. But because the markings are unusual I replaced them with 100pF caps.
Will replace the 150nF with 0.01uF and see what happens.
BTW, sent the video to diyhifishop, they asked which kind of difference was meant. So they at least pretend to be deaf. No comment
It is multiple feedback not Rc. Filter with 100nf is 100khz...little less. Replace all opamps with lme49720he and you will get similar sound as buffalo. First you replace resistor on 100pf position with wima fkp 100-220pf and put out 22uf vreference electrolyt and put 3.3uf red wima mkt and mkp 100nf. Vref must be clean or noise will be on output and sound is not so clean.
For me buffalo is to sharp and thin weilllang to simple sounded no 3d stage and separation.
For me buffalo is to sharp and thin weilllang to simple sounded no 3d stage and separation.
I wrote 100-220pf something between...it can be til 1000-2000pf ... matter of taste and filter setting and scope mesurement... I discovered that less noise of dac modulation comes into opamp, better ...for me relaxed neutral big stage, deep bass, complexity of tones in music... for example if you have 2 different acoustic guitars playing, it must be immediattely hearing different kind of timbre of both instruments.
If you want do get rid of noise in 1 opamp you must use filter like Rasmussen suggested. If not than do more filtering in first opamp to go less noise into second opamp and its feedback. Maybe i am wrong but for my taste and scope mesurements convinced me to do filters in that way. AK4490 that i have at the moment gives smaller noise at output than es9018k2m.... and even on first RC filter stage, noise is almost zero...and don*t come into buffer filter opamp. And sound is very good, very neutral, excellent complexity.
https://www.youtube.com/watch?v=pS7KDE_DgFs&list=PLXv-reIETDOGY4N-GZIEspf9bDhUGkwAP
Listen to this 2 guitars...ok it is youtube but original cd is better. One guitar is magical romantical soft and fullsounded from floor and almost till ceiling and another one is more explosive ... take flow trought the song and more in the middle and front of vertical position. One friend told in joke that he could smell scent of a wood from guitar. Sorry am not good at describing.
If you like this 2 actors than you must hear also album Krushevo.
https://www.youtube.com/watch?v=P0UwmutM_ZI&list=PL8294FD8EB7F0E67C
My best combination in 22uf place was 10uf tantal and 1uf pps, both smd size pps under and tantal on top.
Something similar as here. But pps under, that was before upgrade. Second best non smd version is combination described few posts back.
For 3,3V reference a will use TPS7A4700 mini reg from Aleš Mravlcax
If you want do get rid of noise in 1 opamp you must use filter like Rasmussen suggested. If not than do more filtering in first opamp to go less noise into second opamp and its feedback. Maybe i am wrong but for my taste and scope mesurements convinced me to do filters in that way. AK4490 that i have at the moment gives smaller noise at output than es9018k2m.... and even on first RC filter stage, noise is almost zero...and don*t come into buffer filter opamp. And sound is very good, very neutral, excellent complexity.
https://www.youtube.com/watch?v=pS7KDE_DgFs&list=PLXv-reIETDOGY4N-GZIEspf9bDhUGkwAP
Listen to this 2 guitars...ok it is youtube but original cd is better. One guitar is magical romantical soft and fullsounded from floor and almost till ceiling and another one is more explosive ... take flow trought the song and more in the middle and front of vertical position. One friend told in joke that he could smell scent of a wood from guitar. Sorry am not good at describing.
If you like this 2 actors than you must hear also album Krushevo.
https://www.youtube.com/watch?v=P0UwmutM_ZI&list=PL8294FD8EB7F0E67C
My best combination in 22uf place was 10uf tantal and 1uf pps, both smd size pps under and tantal on top.
Something similar as here. But pps under, that was before upgrade. Second best non smd version is combination described few posts back.
For 3,3V reference a will use TPS7A4700 mini reg from Aleš Mravlcax
Attachments
Last edited:
In the attached photo is a roundup of what Weiliang made wrong.
As far as I remember, those wrong 150nF capacitors were mentioned in this thread before so this error might be in many boards.
I replaced the caps according to Weiliangs schematic and treble is back again.
Differences to BuffaloII are now almost negligible.
Still have to look at LM7805 feeding ADP4 and xtal, seems to make some
hum/noise in the analog output. (At least removing AC power from this
circuit results in lower hum.)
As far as I remember, those wrong 150nF capacitors were mentioned in this thread before so this error might be in many boards.
I replaced the caps according to Weiliangs schematic and treble is back again.
Differences to BuffaloII are now almost negligible.
Still have to look at LM7805 feeding ADP4 and xtal, seems to make some
hum/noise in the analog output. (At least removing AC power from this
circuit results in lower hum.)
Attachments
Last edited:
Still one basic question about the Weiliang board mentioned above (and elsewhere in this thread):
Input voltages are 2x9VAC and 15-0-15VAC.
But heatsinks get very hot! Especially the small one.
Doing some math, talking about the +/-15VDC supplies, feeded by 15VAC
15AC rectified is about 24V, correct?
Thus 9V will have to be dissipated as heat for the needed +/-15VDC.
Also as far as I understand schematics, 9V AC are rectified to 14VDC for 5V final voltage
Again 9V are dissipated as heat. Maybe my math is wrong assuming
that alternating current equals 1.6x in rectified / direct current...?
Only the 9VAC inputs have 5V voltage regulators attached, one even being low drop according to the schematics.
Dropout voltage is normally much lesser than 3V
Thus 5 to 6V AC (equaling about 8V/9.4VDC) should be sufficient?
The +/-15V lines are based on transistor design - don´t know whether this equals a "discrete" voltage regulation.
But again assuming 3V dropout voltage, 12VAC equaling 19.2VDC after rectification should be sufficient...?
Here is the schematic again.
http://www.diyaudio.com/forums/atta...i-end-usb-dac-es9018-schematicdiyhifishop.pdf
Please check CON1 / CON2 / CON3
Input voltages are 2x9VAC and 15-0-15VAC.
But heatsinks get very hot! Especially the small one.
Doing some math, talking about the +/-15VDC supplies, feeded by 15VAC
15AC rectified is about 24V, correct?
Thus 9V will have to be dissipated as heat for the needed +/-15VDC.
Also as far as I understand schematics, 9V AC are rectified to 14VDC for 5V final voltage
Again 9V are dissipated as heat. Maybe my math is wrong assuming
that alternating current equals 1.6x in rectified / direct current...?
Only the 9VAC inputs have 5V voltage regulators attached, one even being low drop according to the schematics.
Dropout voltage is normally much lesser than 3V
Thus 5 to 6V AC (equaling about 8V/9.4VDC) should be sufficient?
The +/-15V lines are based on transistor design - don´t know whether this equals a "discrete" voltage regulation.
But again assuming 3V dropout voltage, 12VAC equaling 19.2VDC after rectification should be sufficient...?
Here is the schematic again.
http://www.diyaudio.com/forums/atta...i-end-usb-dac-es9018-schematicdiyhifishop.pdf
Please check CON1 / CON2 / CON3
Last edited:
Hi I see you haven't used the right math before. It is a factor 1.41. Ripple voltage and diode Uf also need to be taken into account. You are right that voltages seem randomly chosen on the high side, as often with chinese designs. You could do with less heat indeed.
You could use 7V and if you replace the 7805 for an LDO version you could go as low as 5V AC. Preferably a bit higher to be safe. So 6V AC.
Never assume anything so please measure the dropout voltage of the 15V regulators before making statements.
* Dissipation is not just voltage, it is the product of voltage x current thus power. In Watts.
You could use 7V and if you replace the 7805 for an LDO version you could go as low as 5V AC. Preferably a bit higher to be safe. So 6V AC.
Never assume anything so please measure the dropout voltage of the 15V regulators before making statements.
* Dissipation is not just voltage, it is the product of voltage x current thus power. In Watts.
Last edited:
@androa76,
I know you did not touch your board for some time and already and posted some
photograps. Bur if you atill have your boar shelved, could you again make some
Close up photographs of your mods? (Top an Bottom side)
As one example you put in 3 LED to replace original diode, but how, if there is only one
Dolder Point - at leastwith my board. Same with the +/-15V power citcuitam not clear to me which transistors were finally used or Hof ferrite beads were implemented.
Phoros woul help more than explanations!
All the best, Salar
I know you did not touch your board for some time and already and posted some
photograps. Bur if you atill have your boar shelved, could you again make some
Close up photographs of your mods? (Top an Bottom side)
As one example you put in 3 LED to replace original diode, but how, if there is only one
Dolder Point - at leastwith my board. Same with the +/-15V power citcuitam not clear to me which transistors were finally used or Hof ferrite beads were implemented.
Phoros woul help more than explanations!
All the best, Salar
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.