Position of the TCXO
Is it ok if I glue the TCXO upside down to the chip Av100 (rebranded CMI8788 or whatever)? It will also act as a heatsink and ensure shortest path for ground and clock out. Does the chip heat much during use? I'm asking cause I have it at the bottom of the mainboard and can't put my finger in to check the chip's temperature during working conditions... when I remove the card and check it after a minute it's been on it seems that it does not heat that much.
Is it ok if I glue the TCXO upside down to the chip Av100 (rebranded CMI8788 or whatever)? It will also act as a heatsink and ensure shortest path for ground and clock out. Does the chip heat much during use? I'm asking cause I have it at the bottom of the mainboard and can't put my finger in to check the chip's temperature during working conditions... when I remove the card and check it after a minute it's been on it seems that it does not heat that much.
An externally hosted image should be here but it was not working when we last tested it.
While the TCXO arrives - the CMI8788 chips has 0.15W dissipation power -low- so it's safe to glue the XO there using something else than super glue*, I added two .33uf polyester mkt caps to these positions with great results (clarity and dynamics) vs direct bypassing of the caps near the opamps. These caps are followed by ferrites marked L on the board so the whole deal doesn't take out of the clarity but takes care of the noise on the +12/-12V rails. The -12V rail hasn't got any stab on it so it's noisy.
* superglue destroys IC's
Last edited:
There is not so important the position of the oscillator.
If you turn your device upside down, then I suppose the distance between its pins and the soldering points it will increase... You may care to have the shortest connexion for the clock signal, and GND. The power pin it can be connected with a longer wire. A 0,1µ cap should be soldered directly and in between the oscillator V+ and GND pins.
If you turn your device upside down, then I suppose the distance between its pins and the soldering points it will increase... You may care to have the shortest connexion for the clock signal, and GND. The power pin it can be connected with a longer wire. A 0,1µ cap should be soldered directly and in between the oscillator V+ and GND pins.
I have two choices for this: an .26uf (strange value) WIMA MKP 10 160V or an ordinary 1uf polyester 100V bypassed with an 3300pf MKP also WIMA. Which one would you choose? I will also add a series coil (L) to filter out analog 5V garbage from the 7805.
I'll get back with pictures when I get the TCXO which is economy int'l shipping from China (Hong Kong)... was shipped about 5 days ago, does anybody know how much does it take to Europe? Meanwhile chears to everybody, this is a great thread
Last edited:
There are 22mm between GND and V+ on the TCXO. Ceramic SMD is 2-4mm or so... how do I fill the gap? I think I'll use the film cap first then change it later
Vanguard 0.1ppm 24.576 TCXO
Here's the pinout for the TCXO and also the distances on the package... where to fit the ceramic cap? Need as little distance to the pins as possible, can't fit it on the 5V reg as it will be too far and wrong ground plane. I'll go with the WIMA MKP 10.
Here's the pinout for the TCXO and also the distances on the package... where to fit the ceramic cap? Need as little distance to the pins as possible, can't fit it on the 5V reg as it will be too far and wrong ground plane. I'll go with the WIMA MKP 10.
Last edited:
missing 7912 is dpak a direct replacement?
This is from a post on head-fi.org: MAKING A DECENT HI-FI POWER SUPPLY FOR XONAR ESSENCE - Page 19
The guy put in an ordinary 7912. Pinout is Gr Inp Out. I thought of using one but then the protective shield won't fit. Would a proper dpak 7912 fit there? Is it dpak 2 or 3?
I'll insert the photos found on headfi and the most important things which is changing a resistor (marked R2 in the pic) to 120k to get -15V a proper tension for regulation and removing L1 which is on the -12V line. Anybody tried it? Is the resistor 0805 or 1206 SMD as I have to order it...
Would this fit directly? d2pak 7912 after changing the resistor and removing the ferrite on the negative supply http://www.farnell.com/datasheets/1631331.pdf
This is from a post on head-fi.org: MAKING A DECENT HI-FI POWER SUPPLY FOR XONAR ESSENCE - Page 19
The guy put in an ordinary 7912. Pinout is Gr Inp Out. I thought of using one but then the protective shield won't fit. Would a proper dpak 7912 fit there? Is it dpak 2 or 3?
I'll insert the photos found on headfi and the most important things which is changing a resistor (marked R2 in the pic) to 120k to get -15V a proper tension for regulation and removing L1 which is on the -12V line. Anybody tried it? Is the resistor 0805 or 1206 SMD as I have to order it...
Would this fit directly? d2pak 7912 after changing the resistor and removing the ferrite on the negative supply http://www.farnell.com/datasheets/1631331.pdf
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
My card is a different revision from what's on head-fi. I marked with red what I think is the resistor to be changed... I also measured on the card and d2pak 7912 fits with the measurements (I measured 0V, -12V on the pad (input) and -12V on what should be out which is now passed through the L). All the necessary caps are in place, I'll do the mod if I can get confirmation that the resistor near the testing point is the one to change.
An externally hosted image should be here but it was not working when we last tested it.
Discard the last picture, here's the right resistor, hope 120k is the right value, just tested with a 20k on top of the existing resistor and the test point dropped to -3.6V
An externally hosted image should be here but it was not working when we last tested it.
Why not buy an STX II? Modding IS FUN
Last edited:
I orded D2pak 7912 couldn't find original dpak3 on farnell but with some leg bending should fit... 100W soldering gun awaiting and 110k, 120k resistors 0805 case. I think 0603 is the original but 0805 should solder in place as well. 110k should get to 14.3V vs 15.xxV for the 120k and is 14V is more than enough for the reg. Be back with the results.
and 110k, 120k resistors 0805 case. I think 0603 is the original but 0805 should solder in place as well. 110k should get to 14.3V vs 15.xxV for the 120k and is 14V is more than enough for the reg. Be back with the results.-12V mod: D2pak won't fit directly, dpak-3 or dpak is needed for soldering on the pad. I'll use a 7912 in an upwards position. Here's the pic before soldering, legs bent to fit the card. The top has been cut to fit under the card's shield. Should I be concerned with heat dissipation? I'll use 120k for the divider to give some -15.xxV regulated down to -12V or whatever tolerance the piece has. Someone please delete post #671
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Soldered the 7912 in, removed the L as per the instructions on previous page, sorry for the quality of pic, out of focus but you get the picture... I got approx 1-2mm clearance from the shield, going tomorrow to a GSM service to have the 120k cleanly soldered and get the -15V needed for the regulator to work. Now it's working with -12V in and approx -11.98 out. At least I soldered it properly
Oh, and the big sanyo SEP capacitor on digital dac decoupling - smooth as silk, all harshness from the treble is gone and bass it well rounded using OPA2107 IV which used to sound bright. Maybe some would find the sound dark but I'm happy with it.
Oh, and the big sanyo SEP capacitor on digital dac decoupling - smooth as silk, all harshness from the treble is gone and bass it well rounded using OPA2107 IV which used to sound bright. Maybe some would find the sound dark but I'm happy with it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I used 110k for the resistor marked in the green pic. It yields at least on my card -14.6V which is more than enough for the 7912 - -12.08 out. So you may want to try 110-120kohm resistor as the 7912 isn't heatsinked and you don't want say -16V or -17V. So far happy with the result, the regulator seems to give better sound across the audio range (more spacing and starts to remind me of tape players) than the coil that was in there (see previous page). Measurements are about the same. Waiting for the TCXO, it's stuck in Beijing for now
Last edited:
I need confirmation for these 5V analog DAC caps:
*tip for changing the caps... don't try to clean the ground hole, just only the +/- hole (+ is marked on the board, pay attention to polarity for -V caps), insert the new cap as you heat the ground hole. You need at least 60W of heat power for everything to go nice.
An externally hosted image should be here but it was not working when we last tested it.
*tip for changing the caps... don't try to clean the ground hole, just only the +/- hole (+ is marked on the board, pay attention to polarity for -V caps), insert the new cap as you heat the ground hole. You need at least 60W of heat power for everything to go nice.
Last edited:
As I advised previously, removing the caps on this board by using the soldering iron is not a very good idea. The very thin/small traces and vias on the board it may be damaged very easy, by the heat of a standard soldering iron. Special induction tools are more appropriate for the operation, but these are quite expensive.
The best solution in this case is to broke the legs of the caps and let it the rest of these legs soldered in place on board. Then the new component may be soldered over in a more convenient way, with much less thermal stress for traces/vias on PCB.
To broke the caps legs, one should move these components back and forth, as the available place it permit. After a while with such movements, the terminals are broken. The cap it can not be used again (too short terminals) but this is not a problem. Just throw it.
The tagged in picture caps are not the 5v L/R decoupling caps, but some used for the internal filtering on the DAC chip. The decoupling caps for power rails. are the two another ones, just beside these green tagged.
For these filtering caps it is better to use SMD ceramic ones, of larger capacity than the original. For the power decoupling caps is better to use also a larger capacity. Working tension for these caps it may be 6v as never here the power it exceed 5v. Oscon as tantalum or even ceramic types can be used.
The metal shield placed very near these caps, should be removed to have more available working place for this mod...
The best solution in this case is to broke the legs of the caps and let it the rest of these legs soldered in place on board. Then the new component may be soldered over in a more convenient way, with much less thermal stress for traces/vias on PCB.
To broke the caps legs, one should move these components back and forth, as the available place it permit. After a while with such movements, the terminals are broken. The cap it can not be used again (too short terminals) but this is not a problem. Just throw it.
The tagged in picture caps are not the 5v L/R decoupling caps, but some used for the internal filtering on the DAC chip. The decoupling caps for power rails. are the two another ones, just beside these green tagged.
For these filtering caps it is better to use SMD ceramic ones, of larger capacity than the original. For the power decoupling caps is better to use also a larger capacity. Working tension for these caps it may be 6v as never here the power it exceed 5v. Oscon as tantalum or even ceramic types can be used.
The metal shield placed very near these caps, should be removed to have more available working place for this mod...
Last edited:
You are right, I looked in the datasheet of the DAC and for the caps marked above picture it says 'L/R internal bias decoupling'. I measured 5V on them so I thought they were the analog decoupling caps. I changed them with low impedance rubycon caps and the sound changed for the better. Wish somebody tells ASUS to make the vias larger so you can change caps easily.
Last edited:
Silvered mica
I just added a 4700pf silvered mica (the cap has 10mm between the leads so you have to bend them for 5mm on the board) bypass to the one of the caps on the 5V 7805 (noisy!) line that powers the analog portion of the DAC and jaw drop... details galore (do you like hiss on AAD recordings?) but almost too bright for my Beyerdynamic DT880. One ought to try that.
I just added a 4700pf silvered mica (the cap has 10mm between the leads so you have to bend them for 5mm on the board) bypass to the one of the caps on the 5V 7805 (noisy!) line that powers the analog portion of the DAC and jaw drop...
details galore (do you like hiss on AAD recordings?) but almost too bright for my Beyerdynamic DT880. One ought to try that.
Last edited: