Pontus turn on “thump”

[gjpmho]

They have mixed a boatload of PSUs in there, both linear and SMPS looking to appear as a low noise design, right?
And low noise means low noise, at any state.
Does it sound sane to you that the opposite would occur, more so by design?

They, themselves admit that it's not normal. Something is asking for trouble in there, and it won't stay put at the current state, it will only get worst.
If you have some good gear upstream any potential failure can harm them too.

The 1VDC spice you measure with your DMM can very well be 15VDC, DMMs are not fast enough to catch such spikes. It did its job but your gonna need an analyzer to see the true extent, maybe some friend near you?

I think the 1VDC spike I measured with my DMM does not paint an accurate picture of the behavior. I just double checked and can hear a light thump through my speakers at my regular listening volume. It’s not blasting, it sounds similar to the light thump I heard when changing inputs on my amplifier, which uses a mute relay when changing inputs. So whatever it is is manifesting as a light thump in my speakers, I think me fixating on that 1VDc spike may have distorted the overall issue.

Update: the voltage spike at the speaker terminals is 20mV. I don’t understand why it would be 1VDC at DACs RCA output and 20mV at the speaker terminals. Ugh, maybe there’s something wrong with how I’m measuring.

 

[Wolf11Man]

Maybe one more thing. The main parts of his DAC are build outside the manufacurers factory by subcontractors. In China everything is about copying and keeping things secret. A cultural habit.
With these programable integrated cirquits in delivery state, the whole thing is without any function. The software is loaded up in house, as the last part of assembly. So no one can make a copy and sell it out of the back door. A common practice from contractors to multiply profits. Just look at Aliex. Any kind of PCB is sold there, many without the official makers service. The vendor usually knows nothing about the product. Ask why.

For a software update the manufacturer must give the software to the service partner. They don't want that. So if you erase or damage the software, the thing is done. Maybe if you send it to China they can repair it. If that happens better just keep the enclosure and plant some nice DAC PCB from Ali inside.

 

[gjpmho]

I think from your standpoint it is a wise decission to leave it alone. Maybe think about checking the regulated voltages on the DAC board, for peace of mind, but that may be all you can do. They should be a volt or so lower than the unregulated numbers on the board.

From what I see, there are programable IC's on the board, so the functions will be done in the software. I could not see any "normal" DAC parts. At least all functions after initial power up, like switching from standby to on, are software commanded. I'm pretty sure an update could solve your problem, if the manufacturer cared. There are all kinds of connectors on the board which may be there for this task, but only be used once, during assembly.
Anyway, all you could do would be some USB connected update and that will not be possible. You need some programmer directly connected to the board.
I think it is not wrong to suppose the manufacturer doesn't care too much for after sale service. So best is to leave it allone and use it as it is. If you break anythink a repair will be complicated or impossible.

PS from a technical view, there is a lot of useless material used in the build. The power supply is lowest electrical engineering, made impressive by using countless parts.
That doesn't say it can not sound fine, like most DAC's do.

where do I check the regulated voltages? At LT1763?

Ha your analysis doesn’t surprise me. Lots of eye candy and snake oil indeed. But I still enjoy the Pontus more than the Ares II, I’ve tried to let go of preconceived notions. The reviews seems to resonate some truth.

Perhaps the manufacturer updated the newer models. I know Vinshine Audio was providing FPGA firmware updates until 2023. I don’t have consensus from other Pontus I and II owners on whether or not they’re hearing this thump. Some just turn on their amp afterwards. I know of another Pontus I owner who says they also hear a thump, then I heard from a Pontus II and 15th owners that claim they don’t hear a thump or anything (what about the relay that clicks when you presss standby?). But we’re not using the same setup or following the scientific method.

 

[Wolf11Man]

The LT1763 is a regulator. You will find the voltage for the DAC section at its output. It only supplies 500mA, you should find more of them.
A few voltages for the digital part and at least 2 for the analog section would be my guess.
As I said, from the pictures I can not identify much.
I can't identify the output stage as well. There is a small chance that one of the anlog voltages is missing, while the outputstill works. That is all you should check IMO.
Again, be carefull. It is very common to short the legs of these integrated cirquits while trying to measure! There are no fuses you can blow as a safety!

 

[Wolf11Man]

The LT1763:
Fixed Output Voltages: 1.5V, 1.8V, 2.5V, 3V, 3.3V, 5V
Adjustable Output from 1.22V to 20V

The output voltage is always a little lower than the input.
I expected to find dual 12 Volt or more for the analog part, but they can also drive some OP-amps from something much lower.
I think of "usual" high end stuff. For sonic reasons, most line out stages have a quite high output voltage.
So maybe they use only +-5Volt. Lower voltage = cheaper electrolytics.

 

[gjpmho]

The LT1763:
Fixed Output Voltages: 1.5V, 1.8V, 2.5V, 3V, 3.3V, 5V
Adjustable Output from 1.22V to 20V

The output voltage is always a little lower than the input.
I expected to find dual 12 Volt or more for the analog part, but they can also drive some OP-amps from something much lower.
I think of "usual" high end stuff. For sonic reasons, most line out stages have a quite high output voltage.
So maybe they use only +-5Volt. Lower voltage = cheaper electrolytics.

Here are measurements. I found 5 LT1763 LDOs on the DAC board and 5 on the DSP board. I can’t access the power board. These all seem to be in the digital section.

DAC board LDOs

IC1
Vout=5.24V
Vin=7.09V

IC2
Vout=3.90V
Vin=7.09V

IC3
Vout=5.28V
Vin=7.09V

IC4
Vout=3.47V
Vin=3.83V

IC7
Vout=3.46V
Vin=3.83V

DSP board LDOs

IC22
Vout=3.25V
Vin=3.83V

IC27
Vout=1.22V
Vin=3.83V

IC28
Vout=2.43V
Vin=3.83V

IC29
Vout=0.14V
Vin=3.83V

IC4
Vout=3.22V
Vin=3.83V

Am I correct that +3.8V and +7V are the unregulated voltages for the digital section, and +/-7.2V are the unregulated voltages for the analog section? +3.8V and +7V remain on in Standby.


I found these opamps in the analog section.
LM732


That’s funny all this could fit in a Walkman, at least thermal management doesn’t appear to be an issue. Maybe one day you’ll have the option to purchase different sizes and looks lol. I’m pretty sure the toroidal transformers are more for aesthetics too.

 

[Wolf11Man]

IC 29 is strange. No one needs 0.14V. This may be interesting, the rest should be OK, even as the voltages are a bit off from the usuall values, like 5.0V or 3.3V.
Do these regulators have values on them? There are fixed and variable versions.

Your measurements may depend on the ground you use.

Can you find out what voltages feed the LM732 OP-amps? Look into the data sheet to identify the pins. If you measure the + and - feed, you get double the voltage, this will eliminate any ground error.

 

[gjpmho]

I used the same digital ground as the reference for all the above voltage readings. I didn’t want to risk shorting pins with the probes. I think all 10 lm1763s are variable, and they would be marked with fixed voltage (-x) if they were fixed. It’s hard to tell from the datasheet.

I managed to probe only -V on the 8 lm7321a opamps.
This time using analog ground
From left to right starting at IC44.

-7.4V, 0V, -7.4V, 0V, -7.4V, 0V, -7.4V, 0V
Does it make sense +0V on alternating opamps? This seems strange the 4 opamps are off. Is this for the single ended output?

I couldn’t safely probe the +V pin(7) (I might try again after powering off and using a safer contact to probe.

 

[Wolf11Man]

If you can verify the positive voltage om the OP-amps is missing, you got the problem I wrote about many posts before.

Does the positive leg of the OP-amp conect to the regulator showing only 0.14 Volt? Then the regulator is defective and the thump could / should be the result because of loading the output cap.
If so, maybe you should order such a variable regulator from a honest source. Mouser, Digikey etc. Consider buying a cheap hot air soldering station from Ali, ebay or Amaz.
No need to go expensive for a single IC swap. You also need some soldering paste, cleaning copper braid and some tweezers. This is still a very large SMD part. Easy to change.
All together this should be about 50-60 Euros.

Single ended output has noting to do with the voltage feeding the OP-amps! Like - for the negative and + for the positive side of a symetrical output. No!

If the voltage is already so low as inside your DAC, you need both, negative and positive feed to have some headroom in the output signal. Limit for some versions is +-8 Volt, so the dual 7.xy Volt would be the usual dual voltage for the OP-amp.

 

[gjpmho]

Thanks for all your help! I turned off power and verified the the analog power supply (+\-7.2) connects to +/-V pins on all the opamps. No direct connection between the voltage regulated output voltage and the opamp V+ pin. I have a soldering iron, and knowledge and skills to replace the voltage regulator from mouser. I’ll verify the voltage you asked for. I’ll check if the opamp and regulator are on the same circuit. I’ll also recheck my measurements, it seems odd all opamps are connected to the same voltages, and half measure 0 volts.

 

[gjpmho]

The IC29 (LT1763) is in SHDN shutdown Off mode (since pin 5 is 0 Volts), so 0.14V for Vout is expected. I think it’s ok. I verified all of the other regulators have Vin on pin 5, so they’re in On mode. I’ve seen this scenario before on the Ares II, one of the voltage regulator turned on when USB mode was selected.

I managed to measure V+ (pin 7) and V- (pin 4) supplies on the 8 op-amps.

IC 44(LM7321a on the far left)
V+ = GND
V- = -7.4V

IC 43
V+= 7.4V
V-= GND


And so on…

The rest of the pairs repeat this pattern.

I noticed the capacitors for the negative analog power supply are holding charge much longer than ones for the positive analog power supply. After several hours without power. V+ was 0.004, and V- was -0.4V.Could this discrepancy imply an issue?

 

[Wolf11Man]

Look up the data sheet for any OP-amp you find. Then measure the voltage according to the schematics pinout.

I still don't belief that there are OP-amps feed by a positive and others by a negative voltage. This is near to impossible. Maybe you did get the usual (+) signal input and the inverting (-) input wrong, these are for signals not voltage feeds
Look for the V+ an V- voltage feed and measure across these according to the pinout.

The geniuses that build this DAC don't even have technical data on the web page, so one could verify the line output voltage of the DAC, which depends on the voltage the OP-amps get.

 

[Wolf11Man]

The LM7321 you mention are no Audio OP-amps as far as I understand, so not part of a thump problem, if one exists.
I struggle with making sense of this build. Usually HIFI gear has some clearly identifiable sections, but I miss the Audio part in this one. So I fear I can not help with fault finding. I don't want to instruct you to do things that may mess this thing up.
If you find the voltages at he OP-amps to be OK, I'm out of ideas.
As far as I understand this is mainly a software, not a hardware DAC operating in the usual way.
Such stuff is nothing im comfortable with. I can only guess where things might be located. I got some ideas, but without having the object for measuring stuff I'm out.
So maybe just close the lid and enjoy it as is. I'm sorry I could not do more.

 

[gjpmho]

The op-amp is LM7321a

https://www.ti.com/lit/ds/symlink/lm7322.pdf?ts=1768820582252

I know the pin configuration: Pin 1 is indicated by the dimple, then count counter-clockwise to 8.

I measured V- (pin 4) and V+ (pin 7) to analog GND. The output of each op-amp ranges from 0 to the +/- voltage supply. This makes sense to me in my limited understanding, the positive and negative components construct the waveform.

I just measured 7.4V across V+ and V- on IC44.

 

[gjpmho]

The LM7321 you mention are no Audio OP-amps as far as I understand, so not part of a thump problem, if one exists.
I struggle with making sense of this build. Usually HIFI gear has some clearly identifiable sections, but I miss the Audio part in this one. So I fear I can not help with fault finding. I don't want to instruct you to do things that may mess this thing up.
If you find the voltages at he OP-amps to be OK, I'm out of ideas.
As far as I understand this is mainly a software, not a hardware DAC operating in the usual way.
Such stuff is nothing im comfortable with. I can only guess where things might be located. I got some ideas, but without having the object for measuring stuff I'm out.
So maybe just close the lid and enjoy it as is. I'm sorry I could not do more.

No worries I appreciate your help and I’ve learned. I’ll likely not open the lid until there’s an issue or time for replacing the capacitors.

 

[Wolf11Man]

I struggle to find any dedicated audio OP-amp. Chinese usually want something with Burr&Brown writen on it, at least OPAxyz or the like. A famous name...
I could not find any ot such kind on the pictures. They may be at he output of the ladder DAC, but I struggle to identify anything in that region. As far as I have understood, this DAC can use different filters, which have to be switched as well. Honestly, I don't understand this construction, at least from here.
So if you can adapt to the thump and the sound is nice and undistorted, it should be OK.

 

[babadono]

Update: the voltage spike at the speaker terminals is 20mV. I don’t understand why it would be 1VDC at DACs RCA output and 20mV at the speaker terminals. Ugh, maybe there’s something wrong with how I’m measuring.

don't forget about your volume control
but boy o boy that looks like a lot of power supply in your pictures to only put out +/- 7.4 volts...there must be other voltages(?)

 

[gjpmho]

I struggle to find any dedicated audio OP-amp. Chinese usually want something with Burr&Brown writen on it, at least OPAxyz or the like. A famous name...
I could not find any ot such kind on the pictures. They may be at he output of the ladder DAC, but I struggle to identify anything in that region. As far as I have understood, this DAC can use different filters, which have to be switched as well. Honestly, I don't understand this construction, at least from here.
So if you can adapt to the thump and the sound is nice and undistorted, it should be OK.

the thump is not loud through my setup, and it doesn’t seem like it’s damaging the DACs components, that was my main concern. Yes the sound is nice and free of distortion with abundant detail. I am very happy with the sound signature, but I have to admit the size and cost feel like pressure to make believe lol. I hope that will dispel after some time and I’ll forget about the dac and enjoy the music.

 

[gjpmho]

don't forget about your volume control
but boy o boy that looks like a lot of power supply in your pictures to only put out +/- 7.4 volts...there must be other voltages(?)

Volume was set to regular listening which is around 2 or 3 out of 10.
3.8 and 7 for the digital section. These are on in Standby mode.
+/- 7.4 for the analog section.

 

[gjpmho]

I replaced IC29 with an extra LT1763 I forgot I had from a few years ago, sourced from mouser.com. IC29 is in the Standby circuit. Vout is now much closer to 0 Volts, still thumps, oh well.

I did some searching and although I don’t have EE fundamental knowledge or really much solid knowledge to anchor my understanding… this could explain the op-amp configurations above:

Google AI Overview produced this result. I can’t vet the overview through the linked source material… yet it does make sense at least, and it doesn’t seem like a coincidence there are inverting and non-inverting op-amp circuits in the Pontus.

“A truly balanced R-2R DAC uses
two R-2R ladders, one for inverting (LSB-driven V_ref/2/4...) and one for non-inverting (MSB-driven V_ref/2/4...) with switches connecting to +Vref and -Vref, feeding a differential op-amp buffer(like a Fully Differential Amplifier) or two separate single-ended op-amps(inverting for one ladder, non-inverting for the other), summing their outputs for true bipolar/balanced output, achieving better linearity and common-mode rejection than single-ended designs”