SMSL SU9-n issue

[horias2000]

About one year ago I bought the SU9-n based on Amir's review. I wanted a good quality DAC that had impeccable measurements and the SU9-n ticked all the boxes. The lack of MQA is not an issue. I I do have a TIDAL subscription but I'm ok with CD quality only. After enjoying the DAC for about one year, it stared having interruptions while playing over USB. I use Roon and the DAC is directly connected to a RPI3 and the Roon Core is installed on a Ubuntu virtual machine. I initially thought that it was a network problem. After fiddling around with the network I came to the conclusion that it's not network related. The interruptions were pretty random. I had days of continuous listening without issues and than, out of nowhere, it stated to interrupt and some times even stop the audio output to the amp. I noticed that the music was still playing in Roon. So this meant that the USB interface didn't disconnect. I then took the DAC out or the setup and connected it directly to my laptop and to a A50s headphone amp. I used local files and Foobar to listen to music. It behaved exactly the same. It was clear at this moment that the DAC had an internal issue. I wrote to the support email and they replied and said that if I want, I can send the DAC back (I bought it from Shenzenaudio) so that they can repair it. I did not want to do this as that meant that I will be without my DAC for a long time (at least 8-9 weeks). So I asked if this was a known issue and if I could repair it myself. They said that I can go ahead and try but they did not say weather this was a known issue or not. So I proceeded to take it apart and started measuring. First measurement was the output of the AC-DC converter that should output +/- 12Vdc. The -12Vdc rail was 0V and it was shorted. I took the converter out to see if the short was inside of it or on the DAC itself. It was the converter. I then used a lab power supply to supply the DAC with the correct voltages. It worked but the interruptions were still there. I continued measuring around with the scope to see what was causing the issue. When I was measuring around one IC (I do not know the part number as it was erased) the DAC would start working. I noticed that if I gently push on the IC, the DAC started working properly. I thought that there might be some bad soldering and I re-soldered the IC. It worked perfectly and I thought that the issue was fixed. I reinstalled the DAC in the system and I listened to it for the entire weekend. It started interrupting again after a couple of days but it is not as worse. It works but from time to time I get these interruptions. I asked the support from shenzenaudio if they could ship the two parts (the IC and the AC-DC converter) and they said that they could but only after the Chinese bank holidays that end in March . So I will ask for the parts again in March. The AC-DC converter is not such a big issue as I can either find a similar one on Mouser of Digikey or I can even make a linear one with a nice snake oil filled toroidal but the IC I can't buy as I do not know what part number it is. I think that the IC might have been damaged when the part number was removed from its package. I had a similar issue in a product I designed and the damage was ESD and mechanical as well. Some of the bonding wires were damaged when the part number was milled out. If anybody has any suggestions what this IC can be, please do let me know. One side of it is connected to the clrystal footprints near it. So it looks like something that would switch between different crystals or a buffer of some sort. I will keep the thread updated so that other people that might have the same issue can fix their devices. Overall, I have to say that I really like the DAC and the sound and I would not want to change it but if I can't fix this issue I will have to look for another DAC.

 

[777]

AC/DC +/-12V in fact has only +12V and -12V DC is 12ac !!! There is one rectifier diode for -12v.

 

[777]

This is at "-12v" in reality.

 

[horias2000]

This is at "-12v" in reality.

OK. That is similar to what I have. The text on the module itself says +/- 12Vdc @ 1A . I noticed that the analog side of the DAC is supplied at +/-11Vdc taken from the +12Vdc rail only. I'm not sure what the -12Vdc rail is used for but it does not work without it. So the converter is not faulty then. Still what you have there in the scope capture is not pretty at all

 

[777]

OK. That is similar to what I have. The text on the module itself says +/- 12Vdc @ 1A . I noticed that the analog side of the DAC is supplied at +/-11Vdc taken from the +12Vdc rail only. I'm not sure what the -12Vdc rail is used for but it does not work without it. So the converter is not faulty then. Still what you have there in the scope capture is not pretty at all

No, no, no. The convertor use -12V after rectification. For both, +12v and -12V there is an CLC filter and then a MOSFET relay and after it, there are two capacitance multipliers, one for each rail. After that we have +/-11vDC for op-amps.

 

[horias2000]

No, no, no. The convertor use -12V after rectification. For both, +12v and -12V there is an CLC filter and then a MOSFET relay and after it, there are two capacitance multipliers, one for each rail. After that we have +/-11vDC for op-amps.

OK, that makes sense. Then it's clear that the converter is not faulty. Then it's just that IC.

 

[777]

OK, that makes sense. Then it's clear that the converter is not faulty. Then it's just that IC.

Or maybe PCB

 

[horias2000]

Or maybe PCB

Yes, indeed. The PCB might have an issue as well. I can only say after I replace the IC. I will try to de-solder the IC completely, clean the PCB and re-solder the IC back on the PCB.

 

[Gradius]

Which 10-pin IC is that one? Lots of dirty flux around him (impossible to read from the photo).

No mqa isn't an issue... It's blessing!

 

[horias2000]

Which 10-pin IC is that one? Lots of dirty flux around him (impossible to read from the photo).

No mqa isn't an issue... It's blessing!

That's the thing. The writing on the IC was intentionally removed for some reason. So I can't really tell. I'm hoping that somebody from smsl might help with the part number or at leas to sell me one.

 

[horias2000]

I have a short update on this. I tried contacting SMSL support but so far no reply. I went ahead and removed the IC from the PCB, I cleaned the PCB and the IC and I re-soldered it back on. For now it works without any interruptions but I'll have to wait and see if the problem is solved or not. From what I can tell by looking at the PCB, the IC switches between several clock sources (TCXO). There is only one TCXO on the PCB so I think that this IC is not really used in this DAC. Therefore, next I will look around this IC with the scope and confirm that this is the function. If this is indeed what it does, I can route the signal from the input of the IC to the output and then take the IC out completely.

 

[horias2000]

I just discovered something new. If the IC heats up, the interruptions start appearing. When I use the DAC with the lid on, it starts malfunctioning after a few minutes and it starts working after I remove the top cover. I went ahead and used some hot air directly on the IC and almost instantly when the hot air was applied, the interruptions started again. When I touch the IC and I cool it down, it starts working. So I think I have a clear indication that the IC is faulty. By looking at the PCB traces I can say with some confidence that this is something similar to an analog switch like the ADG884. I might have some ADG884 somewhere and I could try and replace it. But first I will look around with the scope to confirm this suspicion. I should see the signal coming from the TCXO going to one pin and then the same signal on another pin (when the IC is working). Then I can remove the IC and add a strap wire from the TCXO signal to the output of the IC.

 

[restorer-john]

Ask SMSL directly what the actual IC is. You have every right as a purchaser and repair person to obtain the appropriate service information, schematic etc. Anything else is unacceptable and will result in much reputation damage for SMSL if they don't help you.

Honestly, I'm so sick of companies obscuring or removing component details. It pretty much ended in the 1970s and now it's sadly back due to paranoid Chinese manufacturers.

 

[horias2000]

Ask SMSL directly what the actual IC is. You have every right as a purchaser and repair person to obtain the appropriate service information, schematic etc. Anything else is unacceptable and will result in much reputation damage for SMSL if they don't help you.

Honestly, I'm so sick of companies obscuring or removing component details. It pretty much ended in the 1970s and now it's sadly back due to paranoid Chinese manufacturers.

I've written to the only SMSL email I found ([email protected]) and so far no reply. To be fair it's been only one day and I think they have some bank holydays these days.

I do not think they will give me the schematic as this is intellectual property. I'm not even a repair person. I work in R&D and I have access to expensive equipment and I just want to listen to my DAC and I said it will be faster to try and repair it myself.

I hope they will at least sell me another part although I think I might get to the bottom of it before they reply (if they reply). That is why I started the thread, so that other people having this issue can fix their own equipment if they choose to do so.

I personally hate when component markings are removed. In this DAC there are three components with the markings removed. This analog switch IC and the two LDO's supplying the analog side of the circuit (+/-11Vdc).

Overall the DAC itself is really good. It measures very well and I like the sound. The enclosure is nicely done and the PCB is pretty well made. I hate it that they used very small SMD components (0201) although there is plenty of room to use 0603 or even 0805. Other that this, this seems like a well designed and built product. I will get to the bottom of the issue and I will keep using it. But I agree that this will have an impact on the reputation of the company. At least for me. I was actually looking at the D400EX or ES as an upgrade but now I just want to fix this one and keep using it.

 

[horias2000]

The saga continues. I started measuring around the IC and as it turns out, it's not a simple analog switch that switches between several clock sources. I think it is some king of clock synthesizer as it has the 24MHz on pin 2 and I have 24MHz on pin 9 (with the edges rounded). What's interesting is that this IC has 49MHz on pin 6 and pin 10. When it heats up, the clock on pin 10 gets very distorted and it even stops completely. Once I cool the IC down, it starts working properly. So I think that this IC generates the 49MHz from the 24MHz crystal. And the 49MHz from pin 10 is used by the DAC as MCK. The only solution I have at the moment is to try and cool down the IC with some sort of heatsink. I'll keep digging around to see if I find some similar IC from any manufacturer. This IC is actually the "Ultra low phase noise and optimized clock processing circuit to achieve ultra-low clock jitter" that they advertise on the website.

This is the signal on pin 9 of the IC. This has the edges rounded so I'm assuming that this is the output of the 24MHz clock that comes from the TCXO and into pin 2 of the IC.

This is the signal from pin 10 when the DAC works properly.

And this is the signal from pin 10 when it doesn't work properly:

It's clear that when the IC heats up, it starts misbehaving and the 49MHz clock that is used by the DAC is completely messed up. The temperature I measured with the FLIR camera is around 51C. So it's not really that big but it's enough to make it misbehave. At the moment the solution I have is to use it without the top cover and I will try to install some sort of heatsink on this IC.

 

[AnalogSteph]

This IC is actually the "Ultra low phase noise and optimized clock processing circuit to achieve ultra-low clock jitter" that they advertise on the website.

So a PLL clock generator of some description (presumably fractional-N or DDS). I would check at distributors which of the usual suspects (AD, SiLabs, ...) have chips that match the functionality, package and what you know about the pinout. I kind of don't think it's something totally off-brand and exotic.

 

[horias2000]

So a PLL clock generator of some description (presumably fractional-N or DDS). I would check at distributors which of the usual suspects (AD, SiLabs, ...) have chips that match the functionality, package and what you know about the pinout. I kind of don't think it's something totally off-brand and exotic.

Yep. Already did this. I think it is a Si5351A with i2c interface. I just checked and it clearly gets some data on pin 5 which it should be SDA. The package and pinout checks out. I will most probably buy a couple from digikey and check. It's actually a nice idea to use a clock generator like this but I have no idea why they would scratch the PN off. Anyway, I will come back with update once I test the IC.

 

[zenon]

Yep. Already did this. I think it is a Si5351A with i2c interface. I just checked and it clearly gets some data on pin 5 which it should be SDA. The package and pinout checks out. I will most probably buy a couple from digikey and check. It's actually a nice idea to use a clock generator like this but I have no idea why they would scratch the PN off. Anyway, I will come back with update once I test the IC.

I think the Si5351A is not the right one.
Please check the pinout of Si5351A and PCB layout if they match.

Good luck!

 

[horias2000]

I think the Si5351A is not the right one.
Please check the pinout of Si5351A and PCB layout if they match.

Good luck!

I looked several times and I think it is the Si5351A in the MSOP-10 package. It's the only one with this pinout and in this package that I could find.
In any case, it's just 2.4$ a piece so I will order 2 or 3 and I'll see if it works or not.

 

[horias2000]

Just a short update. I ordered the IC and I will replace it once it arrives. I did another test and indeed the optical and coax input work just fine when the USB doesn't. SO this IC is only used for USB playback. I guess that the clock is generated differently for COAX and optical. I just want to add that I also measured the Vcc voltage of this IC and it is 3.28V. So this is definitely ok. Not an issue of over-stressing the IC. It does seem though that the problem is getting worse as time goes by. Now even if I cool it down, it won't work.