• The Vendor's Bazaar forum is for commercial offers and transactions. Only unmoderated members can post here.

    diyAudio provides this forum for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members. Use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Reference DAC Module - Discrete R-2R Sign Magnitude 24 bit 384 KHz

I didnt like NOS filter with dam1121, but I have preferred NOS (or at least not disliked it) with other true NOS DACs.
It may be that some or most of the NOS charm comes from the lower clock frequencies (and potentially lower jitter), which the soekris NOS filter will not have since the input is still oversampled to the same rate only with duplicated samples.

Was this with Paul's earlier NewNOS or gumisb's recent NOS filters?

If you haven't tried gumisb's filters, you should. Paul's NewNOS filters had a very audible raspiness in the midrange. Sounded broken to me, and I enjoyed C128dp for quite a while. Then gumisb came along with something I like to call "NewNOS++" and I think it sounds amazing.
 
No, the shift registers get their supply from the +/- 4V Vref
But there is 1 3.3V regulator feeding the mcu, oscillator, isolators, I/O-banks of the FPGA, etc... so allready very noisy to start with. I still use only that same single regulator, but removed the RC feeding the oscillator on the FPGA-side and made a new rcrc filter starting from the regulator tab (also 3.3V) straight to the oscillator. So I use the 'cleaner' side...

Did you make comparisons with the ceramic RC added vs only the polymer caps? As this is a much more invasive procedure than merely adding the polymer caps...
 
Here's an example of the incredible linearity the DAM1021 is capable of (measured with a Prism Dream AD-2).

It looks like this if and when you are lucky to catch a period of low drift.

The clock drift issue really is an insult to the quality of the converter and the Soekris customers. Soeren, are you ever going to fix this?
 

Attachments

  • DAM1021_linearity.jpg
    DAM1021_linearity.jpg
    613.4 KB · Views: 419
Did you make comparisons with the ceramic RC added vs only the polymer caps? As this is a much more invasive procedure than merely adding the polymer caps...

Sorry, no... but if in rev7, the oscillator got its dedicated regulator, some extra filtering won't do any harm 😀 There must be a good reason for Soeren to go this way...
With two soldering irons, you can easily remove the smd resistor heating it at both sides at the time and than feed the oscillator with two 1.5V AA batteries to give it a try.
 
Was this with Paul's earlier NewNOS or gumisb's recent NOS filters?

If you haven't tried gumisb's filters, you should. Paul's NewNOS filters had a very audible raspiness in the midrange. Sounded broken to me, and I enjoyed C128dp for quite a while. Then gumisb came along with something I like to call "NewNOS++" and I think it sounds amazing.

I dont know, but I think it was 'true' NOS (ZOH), so didn't think you could change/improve anything about it.

There were some 'pseudo-NOS' filters that I used, which I liked but the favourite was TNT's filter, are these the ones you're referring too?
 
Clock Drift, can you hear clock drift in domestic use? is it like Wow and Flutter!, how would you recognise it when you are listening to a song or audio in general? my TDA 1541a sounds different, the 2015 implementation that I have is probably more than enough for a chip made in the 80's. Philips didn't deceive anyone when they designed this chip..... or did they?..(dramatic music in the background).
 
Last edited:
Clock Drift, can you hear clock drift in domestic use? is it like Wow and Flutter!, how would you recognise it when you are listening to a song or audio in general? my TDA 1541a sounds different, the 2015 implementation that I have is probably more than enough for a chip made in the 80's. Philips didn't deceive anyone when they designed this chip..... or did they?..(dramatic music in the background).

If it were only wow and flutter - no. But we established with a 1/4 sample frequency test tone that something else is going on. It's hard to reproduce and measure at the same time, but I do think it may be audible as signal degradation, and I have an idea multiple passes through the DA AD sound different depending on drift. You can see the difference in waveforms carrying musical information, samples do not line up with the shape of the waveform as long as drift is high (corresponding to distortion in the 1/4 sample frequency).
 
If you didn't tell me about this would I notice it in the audio I am hearing?

I love the terms companies use like Bit Perfect, Miracle Dac, 1bit, MASH, New Class A, Reference, do these things mean anything or something completely different? they all have their weaknesses, valve amplifiers have a high THD some like it some don't, would you hear it though? personally I hate the sound of square waves but its the standard test.
 
Last edited:
...It looks like this if and when you are lucky to catch a period of low drift.

The clock drift issue really is an insult to the quality of the converter and the Soekris customers. Soeren, are you ever going to fix this?

Theoretically there is option to stop small clock constant change around the target frequency by putting on hold I2C communication with bringing SCL low.
Assume we have very good clock just before i2s input ,after initial adjustment of dac si514 to that clock only small updates overshooting target are send so if we temporary put on hold this communication we will have stable clock. Critical is how close to target dac clock was set.
 
The dam1021 adjust its oscillator with an accuracy of 1 Hz, it don't have anything to do with drift. The oscillator will adjust as needed, but can never hit exactly the frequency of the incomming sample rate. It have been explained multiple times since 2014, please read the whole thread 🙂

The adjustments happens in a interrupt routine, so If you want to test without updates, just enter uManager, updates then stop.... But I doubt that anybody can hear it.

And btw, I2C is a multimaster open drain bus, you can short the lines....
 
Last edited: