Or even cheaper from member Mravlca.
http://www.diyaudio.com/forums/vendors-bazaar/263618-tps7a4700-low-noise-ldo-regulator-pcb-8.html page 8.
I am very satisfied with this regulators. I installed it in wm8741 avcc and dvcc position, for 74HC... and SPDIF input receiver and USB module. So 2*5V and 2*3,3V
After first listening sound is more...fresh. More colourfull not so dark and dull. More fuller.
I don*t want to tune DAC into hiper resolutive sterille dac as ES9018 can be. After hearing TOTAL dac i know what i am looking for.
I tried a lot with caps and also one friend also member her, complaint about sssssterile sound when he put TPS7A4700 instead of 7805...but bypassing was not proper implemented. So you must try some others caps if you are hearing too much ssssss.
I used KZ for avcc and original low esr Nover 100uf plus 1uf MKT for dvcc... Os con SP and SEPC from 47uF to 470uF brings metallsounded to plastic sounded character with lot of ssssibilance. FM was better but not neutral enough. All music has a little bit of nasty high midrange pronounced character.
http://www.diyaudio.com/forums/vendors-bazaar/263618-tps7a4700-low-noise-ldo-regulator-pcb-8.html page 8.
I am very satisfied with this regulators. I installed it in wm8741 avcc and dvcc position, for 74HC... and SPDIF input receiver and USB module. So 2*5V and 2*3,3V
After first listening sound is more...fresh. More colourfull not so dark and dull. More fuller.
I don*t want to tune DAC into hiper resolutive sterille dac as ES9018 can be. After hearing TOTAL dac i know what i am looking for.
I tried a lot with caps and also one friend also member her, complaint about sssssterile sound when he put TPS7A4700 instead of 7805...but bypassing was not proper implemented. So you must try some others caps if you are hearing too much ssssss.
I used KZ for avcc and original low esr Nover 100uf plus 1uf MKT for dvcc... Os con SP and SEPC from 47uF to 470uF brings metallsounded to plastic sounded character with lot of ssssibilance. FM was better but not neutral enough. All music has a little bit of nasty high midrange pronounced character.
The price of this PCB is less than 1€, the price of the TPS7A4700 is ca. 4€, the rest < 1€. Summary it is less than 6€.
In case of 2 LDOs, the price 25$~25€ it is not a problem, but if you need more than 10 regulators?
I will not. If they don't know such a simple thing, it's their own fault. Remember, the age of engineers is over; marketing rules.You must inform TI about it.
So - you will buy from TI, and I will listen to music.
Seriously, I loved this little thing. One minor imperfection I'm trying to fight now is somewhat loose lower bass. It must be more collected and punchy. As of now I use bare ADA4627's, thinking of buffering them.
Last edited:
I think the problem is not the type of used op-amp, but the schematic of the output stage circuit. The summing of the differential outputs and the one stage low pass filter in one op-amp, it is not good. I will tray the output stage of the AK4399 DAC Lynx D60. This circuit contains 4 op-amps per channel (2 x LME49710, LT1368 and AD811). I am waiting for PCBs for this output stage (I've just replaced two LME49710 op-amps with one LME49720).
Attachments
Last edited:
I use two stage, single opamp DSD variant of the filter circuit from 4397 datasheet. Sounds just proper, except that loose bass.
Lynx is great, but LME49710 he chose works poorly in this place. It's bipolar input with significant bias current, FET opamps are much better here, imho. I'd rather try my fave 744+811 composite in the next iteration.
Lynx is great, but LME49710 he chose works poorly in this place. It's bipolar input with significant bias current, FET opamps are much better here, imho. I'd rather try my fave 744+811 composite in the next iteration.
Ja, right, BUT in your circuit, in your environment (power supply, output capacity) etc.
I was experiencing weird random scratchy pIayback issues. When I decide to investigate it strictly, I realized that analog supply regulators was drawing too much current, which was causing excessive voltage drop (goes way below 5V) at analog rails due to in-series resistor at CRC filter of my external supply. I checked pins but there wasn't any short-circuit failure so I replaced all tantalums around the dac chips..and that was it. I was using old axial type tantalums but it seems they were failing when voltage applied to them.
I replaced all of them with polymer types (sanyo sepc & oscon) except at analog supply bypass (which is an Elna silmic II). The result was very satisfying. Very noticeable low-freq reproduction which goes way down deeper as it could get. One noticeable thing is that gives a strange impression about the tightness of drum-like instruments (for example: http://www.youtube.com/watch?v=7yH6o7YGG4c ). I'm using transformer at output stage as I mentioned earlier.
I replaced all of them with polymer types (sanyo sepc & oscon) except at analog supply bypass (which is an Elna silmic II). The result was very satisfying. Very noticeable low-freq reproduction which goes way down deeper as it could get. One noticeable thing is that gives a strange impression about the tightness of drum-like instruments (for example: http://www.youtube.com/watch?v=7yH6o7YGG4c ). I'm using transformer at output stage as I mentioned earlier.
Attachments
Hi terranigma
Glad you eventually found a solution to the playback issues.
I never had long-term SQ satisfaction when using organic polymer caps (OSCONS in my case) for analog decoupling. In this 2xAK4399 DAC the VCOM and especially HREF positions are very influential, so I would recommend to try audio grade caps in these positions. Also the 1u filmcap for bypass isn't very wise when using OSCONs (oscillation due to very low ESR). I can't predict how sound will be, because you're using transformer output, just try it and decide for yourself what is best in your setup.
Nino
Glad you eventually found a solution to the playback issues.
I never had long-term SQ satisfaction when using organic polymer caps (OSCONS in my case) for analog decoupling. In this 2xAK4399 DAC the VCOM and especially HREF positions are very influential, so I would recommend to try audio grade caps in these positions. Also the 1u filmcap for bypass isn't very wise when using OSCONs (oscillation due to very low ESR). I can't predict how sound will be, because you're using transformer output, just try it and decide for yourself what is best in your setup.
Nino
yo cu6apum,
I had tried these IC's:
ad797
ad8065
LT1360CN8
LT 1028 ACN
LT1363
OPA134 BB
First, the LT 1028 ACN and LT1360CN8 do oscillate in the given output configuration, their sound being a disaster, whatever I tried
Second, the winners (for my rig) were LT1363 and AD8065.
LT1363 sounds incredible transparent with apparent more extended dynamics than the AD8065. Unfortunately, it is not so correct when it comes to and timbres accuracy and instrument own texture definition.
I finally choose, helped by a classical music/opera freak pawn, the AD8065, which fulfilled almost all the criteria for any kind of music in an elegant manner, without any other crazy mods concerning the output stage.
Still, for the jazz, electro, rock and pop music, I could live with LT1363, which is brilliant in his own way.
But the king for classical and opera music for my dac/rig is AD8065. By the way, the 8065 did not liked to me after the first brief tests. It's a video dedicated op-amp, with j-fet input stage, not available in DIP8 pinout; you will need a SOIC to DIP8 adapter. This chip does not support a supply value of more than +/- 14 volt, as I remember.
8065 was reccomended to me by an electronics engineer who works for laser projects at a tech university lab.
The BB OPA134 and AD797 are not in the same league. OPA134 has too much bass, and the sound is neither clean or dynamic as the others, while 797 sounds a bit sterile and the sound lacks the lower register, being on the other side of what we called a warm sound. All these results are accounted for the particular output diagram of the Weiliang Dual Ak4399/ Ak4118, and I don't think my conclusions could be extended if you will use the same chips in some other topology.
I tried to find a chip that keeps the project simple ( 3 operational IC per channel scares me !). If I could afford a proper designed output transformer (sowter, for example), I will gladly use it for my project.
Hope this will be useful for you project.
attitube dot com
Thank you Nino. Probably you're right. But I just can't help myself for giving a chance for odd combinations ..
Last edited:
Hi Nino, which capacitors have you soldered on the DAC pins. It is possible to buy the X7R caps wit capicity >100 nF or 2.2 nF COG/NOP caps. The 2.2nF is the greater NOP capicity for the size 0402. Because they are the decoupling capacitors and not the power tanks, I think the NOP caps should be prefered. What do you think about it?
Last edited:
I would try to use the Panasonic FC and Rubicon (Black, but not gate ) ca. 100 uF capacitors. I think that the FC is better than the Rubicon, but I will try both. Using low noise LDOs (the PCB has own output capacitor, tantalum or ceramic 40-100uF) I think it is possible completely remove the electrolytic caps from all pins, except HREF.
) ca. 100 uF capacitors. I think that the FC is better than the Rubicon, but I will try both. Using low noise LDOs (the PCB has own output capacitor, tantalum or ceramic 40-100uF) I think it is possible completely remove the electrolytic caps from all pins, except HREF.
Last edited:
Thank you Mr. Q!
Ad797 was never an option for me as it needs low-impedance source for proper operation, so it's merely not for this circuit. Ad8065 on other hand is a neat candidate. Still have to try LT1363, don't have them handy.
So - no surprise - this output stage needs a fast FET opamp.
In a nearest timeslot will try playing with composite ada4627/ad811, will tell when done. As of now, the fave is 4627.
Ad797 was never an option for me as it needs low-impedance source for proper operation, so it's merely not for this circuit. Ad8065 on other hand is a neat candidate. Still have to try LT1363, don't have them handy.
So - no surprise - this output stage needs a fast FET opamp.
In a nearest timeslot will try playing with composite ada4627/ad811, will tell when done. As of now, the fave is 4627.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.