Shit! Another X-DAC V3 Problem

[XTRProf]

Hi guys,

I'm back with another X-Dac V3 problem. This time after changing to 60V 10 A Schottky diodes, the optical digital input initially had been able to log on but now cannot. The coaxial digital input still can log on but with high breakup and distortion in the sound. Initially, the optical input is already having this intermittent breakup of sound but now completely cannot log on. Previously, after the dac warmed up all intermittent breakup of sound will be gone.

Measured all the Vregs and all outputs are within specs of 5.08 and 3.3V. So what's the problem, gurus? The 2 digital input SOICs problem? That's what I suspected. Please confirm.

Danke.........

[Deleted]

XTRProf
Would appear to be what you said, but whenever something goes wrong after changing something, I first revert to the way it was, before delving too deeply.
Alex

[XTRProf]

sandyk said:

XTRProf
Would appear to be what you said, but whenever something goes wrong after changing something, I first revert to the way it was, before delving too deeply.
Alex

Hi Alex,

I will do that but what have the schottky diodes to do with the problem. Aren't they just diodes but only fast but soft recovery only?

[Deleted]

XTRProf
I presume that you mean the 2 diodes at the front that feed the 5V and 3.3V regulators ?
The only thing they should do is increase the input voltage to the 5V and 3.3V regulators by 300-400mV (at a guess) and reduce switching noise,but I have found from experience that when something goes wrong, put back things the way they were, and see if the fault persists. You need to know if you inadvertently damaged something, or not.
Alex

[XTRProf]

sandyk said:

XTRProf
I presume that you mean the 2 diodes at the front that feed the 5V and 3.3V regulators ?
The only thing they should do is increase the input voltage to the 5V and 3.3V regulators by 300-400mV (at a guess) and reduce switching noise,but I have found from experience that when something goes wrong, put back things the way they were, and see if the fault persists. You need to know if you inadvertently damaged something, or not.
Alex

Good morning Alex,

Ya, the diodes upstream of the 2 L7805CVs and LF33CV. On paper, they should not have done any damage n also from your confirmation. Also, I had carefully checked all the soldering for short b4 switching on.

I will do the reverse engineering as suggested by you after I had changed one of the 7805 to test as that particular one is not running as hot as the next 7805 although output voltage is 5.08 and 5.04 respectively. Just a noob change and nothing technical about it. A calculated lucky dip.

[Deleted]

XTRProf
The Optical MAY be run off the 3.3V supply.I think they may be using a superseded 3.3V opto module ?
Alex

[XTRProf]

sandyk said:

XTRProf
The Optical MAY be run off the 3.3V supply.I think they may be using a superseded 3.3V opto module ?
Alex

Oh, I didn't know that. I also suspected the optical receiver previously as that gave intermittent sound when cold. But then I also tried the coaxial after this problem and the same problem with distorted sound. Thanks.

[Deleted]

XTRProf
If you are not careful when working on the X-DAC V3 , the front panel can fall down and damage the opto module's terminations. I did recently when the front panel accidentally slipped. (I should have unscrewed the module) Luckily, I don't use Opto, as I prefer the potentially faster SPDIF.
Check for dry joints/damage to the opto module's terminations.
Alex

The previous mention of running off 3.3V may be incorrect.
I think it may actually be a 5V module, but the commonly available types appear to be 3.3V

[XTRProf]

sandyk said:

XTRProf
If you are not careful when working on the X-DAC V3 , the front panel can fall down and damage the opto module's terminations. I did recently when the front panel accidentally slipped. (I should have unscrewed the module) Luckily, I don't use Opto, as I prefer the potentially faster SPDIF.
Check for dry joints/damage to the opto module's terminations.
Alex

The previous mention of running off 3.3V may be incorrect.
I think it may actually be a 5V module, but the commonly available types appear to be 3.3V e.g. TORX176

Actually, I didn't have much of a choice when choosing between opto and coaxial as I just only have a glass toslink lying around. I also prefer coaxial as technically less jitter unless interference is very great. But this problem gave me the impetus to rip out the component occ cable to try the coaxial to better understand the problem.

Usually, Ilm the meticulous type who will rip out everything b4 work commenced as there will be less interference from those in the way. So, at least from me, there is not damage. I'm not sure of the previous owner but judging that he got an Audionote dac to replace the x-dac, I'm sure he's not the diy and surgery type. Yes, I will give all those opto joints the hot solder treatment to erase that possible dry joint problem and try again before taking the "colder" 7805.

[XTRProf]

Nope, no luck after resoldering the opto and changing the "colder" 7805. Still the same problem of intermittent and distorted sound when cold and the opto input did not lock in. After warm up become lesser of intermittent and distorted sound. I suspect there is multiple problems now as when I switched on, the opto receiver did not give a red laser light when I try to see it. Is the opto receiver supposed to give red laser light when on? If is, then this is gone.

Next stop to put back the original diodes and bridge.

[XTRProf]

sandyk said:

XTRProf

The previous mention of running off 3.3V may be incorrect.
I think it may actually be a 5V module, but the commonly available types appear to be 3.3V e.g. TORX176

Nope TORX176 and even TORX179 are not 3.3V but 5V types when I checked up spec sheet now. Yes, the x-dac v3 uses TORX176.

[Deleted]

XTRProf
The ones that I can get locally from Jaycar are 3.3V types and not pin compatible either.The one I wanted to replace was the TORX176, which is of course a 5V type.
Have a look at the attached link
Alex

www.head-fi.org/forums/f6/where-get-torx176-144301/

[XTRProf]

sandyk said:

XTRProf
The ones that I can get locally from Jaycar are 3.3V types and not pin compatible either.The one I wanted to replace was the TORX176, which is of course a 5V type.
Have a look at the attached link
Alex

www.head-fi.org/forums/f6/where-get-torx176-144301/

Hi Alex,

Thanks for the link. Actually, can also replace with the higher bit rate TORX179. Twice the rate of TORX176. But then pin for pin have to modify the board to put one in. So I still settle for a TORX176 if I can find one. If not, then have to modify board and 179 receiver to fit in.

Today, I had started replacing back the diodes for reg2 first and the sound went back to normal as before. But I can still hear the breakup, like a failing transistor sound, faintly. Perhaps, this is already there in the first place and I can't hear that as clearly as now. This is because I had changed the internal audio wire to three runs of twisted Teflon OCC copper wire for each channel. Will be shielding the wires later on when I have the time again. Then I try the TORX176 opto input and still did not lock-in. But there was lock-in for coaxial input.

Having this success, I went on to change the bridge diodes for reg3 back to the supplied one and expecting the faint breakup to completely disappear but it didn't. It was still there. So conclusion, putting Schottky diodes at reg3 is ok and will be a disaster for reg2. Really, this baffle me. Any technical reasons for the technically superior Schottky failing to perform at reg2? Also, there is other problems giving the faint breakup sound. The reference CD used was Amanda McBroom - Greatest Hits from highend Gecko label. Used Track 1 to 3 which can easily hear the breakup if you listen carefully and loudly.

I tried the TORX176 input again and again no lock-in whereas there is lock-in for coaxial, proving the TORX176 is most likely gone with the wind. Will look out for 176 and comes to the worst 179 which is available at Digikey wrt to your link.

OVERALL FAULT DIAGNOSIS: Schottky fail to perform at reg2 but was ok at reg3. The TORX176 had most likely failed by failing to lock-in after all diodes had been replaced. There was still faint breakup in sound and, hence, there are multiple problems. Will work to get rid of that faint breakup. Reg2 will try the less superior Hexfred diodes of similar rating.



Gurus, what are your likely comments?

[Deleted]

XTRProf
Have you tried SPDIF from another source ? How good is the cable used ? C45 ?
SandyK

[XTRProf]

sandyk said:

XTRProf
Have you tried SPDIF from another source ? How good is the cable used ? C45
SandyK

Have but not critical listening using an old NAD 502 to see the likely X-dac problems. I will try that later for critical listening to the faint breakup. Using the internal dac of the universal Pioneer 686 under critical listening there was not even the slightest faint breakup. This is the same player I used for the x-dac when it gives that faint breakup. The 75ohm cable used is an OCC Daiyo component coxial cable. Material wise not question able it but only the design. The cable as thick as normal antenna 75ohms cable. Fyi.

[Deleted]

XTRProf
Try another player's SPDIF. Try replacing C45, or try another cap in parallel.
Alex

[leo]

You have to be careful with Schottkies, I've known these cause problems before in digital equipment.
you sure you checked all the tracks for any hair line breaks?
What digital input IC is used in these units?
Does the break up happen on both channels?

[XTRProf]

sandyk said:

XTRProf
Try another player's SPDIF. Try replacing C45, or try another cap in parallel.
Alex

Tried the nad 502 spdif and the same faint breakup on both channels. I would say the pioneer sounds much better with spdif than the nad spdif. The nad is more grainy and distorted. I will try replacing the c45. What cap value u suggest for paralleling? Thanks.

[XTRProf]

leo said:

You have to be careful with Schottkies, I've known these cause problems before in digital equipment.
you sure you checked all the tracks for any hair line breaks?
What digital input IC is used in these units?
Does the break up happen on both channels?

Superduber! Another guru to the rescue. Nope, I do not used all tracks to check for hairline breakup on the amanda cd. Only tracks 1, 2 and 3. When her voice growl deeply, there will be breakup and distortion like transistor failing or too low a voltage. But faint here I emphasise.

The ic number I do not have on hand as I'm using a public pc now. Will check and give the number when back home. Roughly, one of them is a philips ad898 and there are 2 soics. The other begins with PIC letters. Sorry, I check up and post again and this is not definite.

Yesternight, b4 u mentioned using all tracks on the amanda cd to test, I did try all sorts of music like depeche mode, jazz and a chinese cd teresa teng. Although the faint breakup is there especially like heavily modulated vocal tracks, it's not so apparent. I'm sure if I use paravatti tracks, the problem will be very highlighted. Yes, the faint breakup does occur for both channels.

I can't wait any more, I'm going to farnell singapore to pick up a torx177 at 15mb/s superior to 176 and 179 now to try as it's not too expensive at sgd 5.40.

[Deleted]

XTRProf.
Value shouldn't matter. Try a 100nF or something. What were those wires your Zig and Zag (or was it Zen ?) told you to play with ? I hope they weren't digital ? I just read in DiyAudio that ill informed fiddling with digital wires like that can degrade performance. Was the problem there before that ? I will be very interested to see your report on the new optical receiver IC.
SandyK

[fanboi]

XTR,
I think Leo was referring to checking the PCB tracks for hairline cracks NOT using all the CD audio tracks.

[XTRProf]

leo said:

You have to be careful with Schottkies, I've known these cause problems before in digital equipment.
you sure you checked all the tracks for any hair line breaks?
What digital input IC is used in these units?
Does the break up happen on both channels?

That answered the distortion. The reg1 (3.3V) and 2 (5V) are for most probably for the digital side whereas the reg3 (5V) are for the analog side.

I'm quite meticulous and will check for any problems visually before putting on tbe power. I guess I'm that as I had formerly worked in a PCBA plant before, not at a technical level but assembly level.

The 2 digital input SOICS are:

Philips AD898
TI OBAK6YM


Picture here:





Hopes this gives your a better picture for advice.

[XTRProf]

fanboi said:

XTR,
I think Leo was referring to checking the PCB tracks for hairline cracks NOT using all the CD audio tracks.

Oops......., distorted signal as in this thread. Sorry, all. Have fault corrective action.

[XTRProf]

sandyk said:

XTRProf.
Value shouldn't matter. Try a 100nF or something. What were those wires your Zig and Zag (or was it Zen ?) told you to play with ? I hope they weren't digital ? I just read in DiyAudio that ill informed fiddling with digital wires like that can degrade performance. Was the problem there before that ? I will be very interested to see your report on the new optical receiver IC.
SandyK

Noted!

Ha, had zig zag to true zen using 3 runs of twisted Teflon OCC copper gauge 20 for both AUDIO channels. Now, I can hear more details until distortion!

My zen will tell me not to zig zag to any cable unless true 75ohms and OCC copper or silver for digital. Anyway, the internal "digital" cables are not true 75ohms type but just normal lossy cable.

As to whether there is any problem before this, I couldn't tell as I'm using the glass toslink and not coaxial. I'm now using coaxial as the opto had failed. I didn't notice this clearly until I changed to Schottkies and had zen to OCC audio internal cables.

[Deleted]

XTRProf
Are you able to try a return to the original cables that were shielded ?
Just to make sure not picking up radiated digital artifacts.I agree, that it does sound unlikely, but you haven't been able to verify that fault wasn't there previously.After all , you thought that Schottkies were a safe bet too !
BTW, I know that Leo will agree that only ONE modification should be attempted at a time, then fully evaluated, before performing the next modification. BTW, short lengths of shielded cable aren't the impedance you think they are ! The true characteristic impedance of a cable is only reached after a certain minimum distance. This will depend on type of cable and it's construction. You will need to research any further info on this subject for yourself. I have forgotten !
Alex
Alex