Denafrips 'Pontus' R2R ladder DAC - close up view

[Nayanovmax]

Question:

Has anyone encountered a similar issue, or does anyone have suggestions for resolving this without introducing new problems?

I had exactly the same thing on the new Pontus 2 "out of the box" and I completely got rid of these "stutters". It took installing the original 1.0 firmware and the problem disappeared immediately. The new pontus was running version 1.5 and it behaved as you described above.

 

[Honeypot]

I had exactly the same thing on the new Pontus 2 "out of the box" and I completely got rid of these "stutters". It took installing the original 1.0 firmware and the problem disappeared immediately. The new pontus was running version 1.5 and it behaved as you described above.

That's great news! One more thing to try. I will report this to Denafrips too. If it's something related with the firmware they should address it somehow.
Thanks!
Meanwhile:

 

[medon]

...
But yes, I'm about to try it right now. Just with a couple of dummy cables, just to try and check if it helps it.

Thanks for trying it out! ... already curious what your results will be!

On the other hand, as I stated, the ground also helps the devices to get rid of the EFI captured from their metallic encapsulations (body), and EMI from power supply, etc, via the power cable.

I added a commercial grounding solution (Puritan Groundmaster City + distributor) to my chain last year, and it made a noticeable improvement.
So maybe the regular grounding links on AC power can also introduce RFI noise into your chain?

 

[Honeypot]

I added a commercial grounding solution (Puritan Groundmaster City + distributor) to my chain last year, and it made a noticeable improvement.
So maybe the regular grounding links on AC power can also introduce RFI noise into your chain?

I don't know that specific solution, will check it out. Generally speaking ground shouldn't introduce RFI as long as your outlet ground connection works correctly.
This means Ground to Live (Hot): should measure approximately 230V (or 120 mains voltage). In my case 232 stable.
Ground to Neutral: You should measure approximately 0V or a very small voltage. In my case 1-2v. Which is fine.

Edit: ground should work as a sucker for any exciding current or pulses. This includes the chasis if it's connected to ground (which should be) for safety first and to act as a faraday cage).

What introduce RFI are cables (not shielded) picking random electromagnetic waves all around the place, as well as enclosures or any conductable surface on or in the equipment.

I had exactly the same thing on the new Pontus 2 "out of the box" and I completely got rid of these "stutters". It took installing the original 1.0 firmware and the problem disappeared immediately. The new pontus was running version 1.5 and it behaved as you described above.

Damn since Denafrips has being stopped being part of vinshine the links for the old or new firmware are gone.
I wish I had downloaded it while I could. I may ask them.

 

[sajunky]

The issue manifests as a brief glitch—it lasts for just a split second but is enough to disrupt the listening experience. The best way I can describe it is that it sounds like a fast-forward effect or perhaps a jitter/clock-related anomaly

Denafrips tend to use long FIFO buffers as an alternative to PLL which lock frequency to the source clock.

Synchronisation with FIFO results the lowest jitter possible (as it use a local fixed frequency oscilator), but due to a minimal difference of clock frequencies buffer will eventually run empty or overflows after certain period of time.

Fast forward effect comes when buffer becomes full and firmware has to drop series of extraordinary samples. Firmware seems triggering panic procedure on the extreme condition instead of continuously monitoring buffer level and dropping a single sample or inserting zero sample during quiet moments.

Different firmware revisions can make adjustments to this algorithm.

As for removing ground wire, I agree, it was not a good advice, especially because you use filtered sockets. Switch to unfiltered sockets for this test

 

[medon]

ground should work as a sucker for any exciding current or pulses. This includes the chasis if it's connected to ground (which should be) for safety first and to act as a faraday cage).

That's what I deem to be true as well.

What introduce RFI are cables (not shielded) picking random electromagnetic waves all around the place, as well as enclosures or any conductable surface on or in the equipment.

I believe RFI does not only come "from the outside" (e.g. WiFi, LED lights, ...) but also from the inside.
As my brick and mortar dealer put it, regarding clean AC power and grounding: "The HiFi chain is its own biggest enemy". -- regarding noise, that is.

In my chain for instance I found out that
- my REL subwoofers induce HUGE amounts of AC noise into the mains, probably due to their SMPS. Hearable negative impact on SQ, even when the devices are only switched ON but not putting out any sound (by setting them to zero gain). Remedy: put the RELs on a different wall socket with a filtering AC distributor.
- Ethernet noise: I put a cheap & pretty noise resistant Netgear (GS108E) switch between my ISP's router (FritzBox) and my streamer. --> sound improvement.

My philosophy by now is: Try different wall sockets, try different AC distributors, maybe AC filters, grounding (we had this already), good cables. I have experienced too many "oh, really!?" moments regarding these HiFi "hygiene" factors ...

 

[Nayanovmax]

Damn since Denafrips has being stopped being part of vinshine the links for the old or new firmware are gone.
I wish I had downloaded it while I could. I may ask them.

Write to me in a personal message, I will post you different versions of the Pontus 2 software, they remained on my PC.

 

[sygnus21]

When I got my 12th-1 from Denafrips (Feb. 2024), I never updated the firmware as it had already come with the latest, and I wasn't going to chance creating issues buy putting an older firmware update on a newly released device that already had the latest update. Connected to my Denafrips Iris, and later Hermes, I had zero issues.

Fast forward, I now run a Pontus 15, and it too came with the latest firmware update and so like with the Pontus 12th-1, I'll not be updating the firmware.

This is the firmware info from the Pontus 12, as read without being connected to the DDC, but from PC USB to DAC....

Looks like the firmware is now "N2"? (P15.N2)

 

[Thomsen132]

How is the SQ of Pontus 12th compared to 15th?

 

[Honeypot]

Denafrips tend to use long FIFO buffers as an alternative to PLL which lock frequency to the source clock.

Synchronisation with FIFO results the lowest jitter possible (as it use a local fixed frequency oscilator), but due to a minimal difference of clock frequencies buffer will eventually run empty or overflows after certain period of time.

Fast forward effect comes when buffer becomes full and firmware has to drop series of extraordinary samples. Firmware seems triggering panic procedure on the extreme condition instead of continuously monitoring buffer level and dropping a single sample or inserting zero sample during quiet moments.

Different firmware revisions can make adjustments to this algorithm.

As for removing ground wire, I agree, it was not a good advice, especially because you use filtered sockets. Switch to unfiltered sockets for this test

Thanks for the detailed description.
For the record, with ungrounded power cables, it still does it. So it may well be what you say. The firmware itself.
I will try with unfiltered sockets, but from your description I'm afraid not much will be achieved.

When I got my 12th-1 from Denafrips (Feb. 2024), I never updated the firmware as it had already come with the latest, and I wasn't going to chance creating issues buy putting an older firmware update on a newly released device that already had the latest update. Connected to my Denafrips Iris, and later Hermes, I had zero issues.

Fast forward, I now run a Pontus 15, and it too came with the latest firmware update and so like with the Pontus 12th-1, I'll not be updating the firmware.

This is the firmware info from the Pontus 12, as read without being connected to the DDC, but from PC USB to DAC....



Looks like the firmware is now "N2"? (P15.N2)

My Denafrips Pontus 2 non 12th anniversary is updated to 1.4 but the controller panel is reading differently.

That's what I deem to be true as well.

I believe RFI does not only come "from the outside" (e.g. WiFi, LED lights, ...) but also from the inside.
As my brick and mortar dealer put it, regarding clean AC power and grounding: "The HiFi chain is its own biggest enemy". -- regarding noise, that is.

In my chain for instance I found out that
- my REL subwoofers induce HUGE amounts of AC noise into the mains, probably due to their SMPS. Hearable negative impact on SQ, even when the devices are only switched ON but not putting out any sound (by setting them to zero gain). Remedy: put the RELs on a different wall socket with a filtering AC distributor.
- Ethernet noise: I put a cheap & pretty noise resistant Netgear (GS108E) switch between my ISP's router (FritzBox) and my streamer. --> sound improvement.

My philosophy by now is: Try different wall sockets, try different AC distributors, maybe AC filters, grounding (we had this already), good cables. I have experienced too many "oh, really!?" moments regarding these HiFi "hygiene" factors ...

Believe me I take a lot of effort into these factors and it's not a matter of ground or AC. Issues.

Write to me in a personal message, I will post you different versions of the Pontus 2 software, they remained on my PC.

It will all depends on the version of your device.

 

[sygnus21]

I just got it so not even 30 hours yet to make a fair judgement. I do expect an improvement as it's now using Venus parts now, so...

Anyway, the Pontus 15th was more a replacement as I accidentally fried something in the Pontus 12th-1 killing both I2S and Optical inputs.

 

[CalvinandHobbes2020]

Thanks for trying it out! ... already curious what your results will be!


I added a commercial grounding solution (Puritan Groundmaster City + distributor) to my chain last year, and it made a noticeable improvement.
So maybe the regular grounding links on AC power can also introduce RFI noise into your chain?

It could be due to better grounding, more effectively draining EFI/RFI from your system.

 

[Honeypot]

Denafrips tend to use long FIFO buffers as an alternative to PLL which lock frequency to the source clock.

Synchronisation with FIFO results the lowest jitter possible (as it use a local fixed frequency oscilator), but due to a minimal difference of clock frequencies buffer will eventually run empty or overflows after certain period of time.

Fast forward effect comes when buffer becomes full and firmware has to drop series of extraordinary samples. Firmware seems triggering panic procedure on the extreme condition instead of continuously monitoring buffer level and dropping a single sample or inserting zero sample during quiet moments.

Different firmware revisions can make adjustments to this algorithm.

As for removing ground wire, I agree, it was not a good advice, especially because you use filtered sockets. Switch to unfiltered sockets for this test

After testing Denafrips support proposal solution and yours (ungrounded wire to filtered and unfiltered sockets) the problem remained.
Wrote Denafrips support about it, and also about the findings I made: your FIFO explanation and the other cases found in the forums.
I also mentioned about the old firmware, and they provided me the 1.1 version along with the 12th anniversary (1.4) version of my FPGA chip. Flashed it this morning without issues and the DAC seems to be working normally.
I didn't have much time to test (about 1h of listening) for the fast forward glitch, but so far so good. I will continue checking later and try to see if it's really gone.
By the way, now that I have gone back to the older firmware version I can say I prefer the sound of the 1.1 over the 1.4. hard to explain but it seems better layering, depth, texture and tonality. Or maybe I'm biased. Who knows....
Anyway, thanks everyone for the inputs and hope you a good 2025 entrance!

 

[Smoothstereo]

After testing Denafrips support proposal solution and yours (ungrounded wire to filtered and unfiltered sockets) the problem remained.
Wrote Denafrips support about it, and also about the findings I made: your FIFO explanation and the other cases found in the forums.
I also mentioned about the old firmware, and they provided me the 1.1 version along with the 12th anniversary (1.4) version of my FPGA chip. Flashed it this morning without issues and the DAC seems to be working normally.
I didn't have much time to test (about 1h of listening) for the fast forward glitch, but so far so good. I will continue checking later and try to see if it's really gone.
By the way, now that I have gone back to the older firmware version I can say I prefer the sound of the 1.1 over the 1.4. hard to explain but it seems better layering, depth, texture and tonality. Or maybe I'm biased. Who knows....
Anyway, thanks everyone for the inputs and hope you a good 2025 entrance!

You flashed Pontus to version 1.1 and so far all is good? I didn't know there was a version 1.1, I thought the original firmware on Pontus2 was 1.0.

 

[Honeypot]

You flashed Pontus to version 1.1 and so far all is good? I didn't know there was a version 1.1, I thought the original firmware on Pontus2 was 1.0.

Well, that's what the firmware and the device panel says. Now I don't know of there is a previous version 1.0 or what I've been sent by Denafrips it's a 1.1 or 1.0 named 1.1. I asked if it was my FPGA chip code, and they confirmed. So I flashed. So far I just used 1 hour. But sometimes the glitch didn't happen that fast. As some stated it can happen every 3 hours or so. This afternoon I will be listening to music for hours.. I will report back in a couple days with final results.