Topping D50 and D50s coax input hit or miss

[Samuel T Cogley]

I have both a D50 and a D50s.  This isn't meant to be a review of the DACs, suffice to say they are adequate for the price and the DAC price to performance ratio in general continues to impress.

 

The issue is the coax S/PDIF input.  At first I thought my D50 was faulty when it wouldn't lock to a signal from a known good source.  Then I tried the D50s.  Same issue.  What?  Can't be.  So then I started experimenting with every coax S/PDIF source I could find.  A few worked, many did not.  And without exception, all of those sources work just fine with all my other DACs.  And to be clear, the TOSLINK input works fine on both units.

 

I searched around the internet and found some, but not an overwhelming number of similar experiences.  One poster said Topping's response was something to the effect of, "our DAC is S/PDIF compliant".  My suspicion is that the target demographic for the D50 mostly isn't buying it for S/PDIF, and that's why I'm not seeing more complaints.  I just wanted to post this as a heads up and maybe if someone else has this issue they can chime in.

 

I guess what I really need to do is put an oscilloscope on the S/PDIF sources that work (and don't) and see what the differences in levels might be.  But my experience is that at least with these particular DACs, coax S/PDIF compatibility is not universal.  And even if I knew what the voltage window was that worked, I seriously doubt most people know without testing what voltage levels are presented by their S/PDIF sources.

 

 

 

 

[mansr]

18 minutes ago, Samuel T Cogley said:

I guess what I really need to do is put an oscilloscope on the S/PDIF sources that work (and don't) and see what the differences in levels might be.  But my experience is that at least with these particular DACs, coax S/PDIF compatibility is not universal.  And even if I knew what the voltage window was that worked, I seriously doubt most people know without testing what voltage levels are presented by their S/PDIF sources.

A coax S/PDIF transmitter shall deliver 0.5 V ±20% peak-to-peak into a 75 Ω load. A receiver shall work with input voltages of 0.2–0.6 V.

 

Balanced (AES-3, XLR) sources shall have a peak-to-peak level of 2–7 V into 110 Ω. Receivers shall accept signals of 0.2–7 V. The transceivers are in fact the same as for RS-485.

 

In practice, an AES-3 receiver will accept an S/PDIF signal despite the impedance mismatch. An S/PDIF receiver, on the other hand, may be overwhelmed (or fried) by the higher level of an AES-3 signal.

[Miska]

What varies between receivers is how much input jitter they accept to stay locked. I wonder if they use the ESS Sabre built-in receiver (which almost nobody does).

 

32 minutes ago, mansr said:

The transceivers are in fact the same as for RS-485.

 

In the 90's for some of my first DACs, I used the standard RS-485 interface chips with good results. Because (IIRC) the YM3623 didn't have the electrical interfacing, but instead required TTL level input.

 

[mansr]

7 minutes ago, Miska said:

What varies between receivers is how much input jitter they accept to stay locked. I wonder if they use the ESS Sabre built-in receiver (which almost nobody does).

Just how bad do you think the sources are? The limit in the spec is rather generous.

[Miska]

55 minutes ago, mansr said:

Just how bad do you think the sources are? The limit in the spec is rather generous.

 

They can be very bad, something like Airport Express, or some Bluetooth interfaces (IIRC, both typically generate the clocks from the network receive packet interval). But it is more question how bad the receiver can be. I think ESS had some adjustable PLL lock window and people like to run it at tightest setting for some reason. A good but tricky receiver is the Wolfson one too, it has so many configuration parameters and not so easy to get running nicely.

 

But everybody I've talked to just say the S/PDIF receiver in ESS Sabre is so bad that they don't want to use it. Maybe someone made a mistake and actually tried to use it.

 

Anyway, there are huge differences between receivers in how much jitter they can deal with before falling off.

 

Another likely contributing factor may be ground currents adding noise to the S/PDIF reception, since the Topping D50 seems to be missing the isolation transformer it should be having on the coaxial S/PDIF input. Some S/PDIF transmitters have isolation transformer, some don't (they really should). If you combine setup where neither side has it, the likelihood for problems in this area is much bigger.

 

At least on this side of the PCB there's no needed magnetics for the coax S/PDIF:

 

 

[Ralf11]

How did Bruno Putzeys do his S/PDIF receiver?

[Lightstream]

Hi,

I have/had the same issue with my Beresford SEG DAC. Never had any problems with the predecessor and several other DACs, but the latest Beresford model remained quiet using the SPDIF input. Toslink and USB worked fine. The SPDIF only worked, when the SPDIF connector hardly touched the socket, so that it nearly fell out. Did you try this? As soon as I inserted the plug properly, no signal was transmitted anymore. The manufacturer told me that the DAC worked fine with SPDIF when he checked it and had no idea about the reason.

 

The reason was that I used an XLR-to-SPDIF adapter to connect the AES/EBU output between my Simaudio MOON 180 Streamer and the DAC with an SPDIF cable (Oyaide FTVS-510). Neither the Neutrik AES/EBU to SPDIF adapter I used nor a no-name adapter worked. Also an adapter on the input side of the DAC (AES/EBU to SPDIF) using a AES/EBU Cable did not work which is no problem at all with other DACs. When I used the SPDIF cable and the SPDIF in- and outputs everything worked fine. Maybe a highly bred racehorse reacts more sensitively than the average horse... 

 

Cheers

Stefan

[Samuel T Cogley]

Thanks everyone for your responses so far.  I will say the best results I've had so far is from an iFI iLink, so there may indeed be something to the jitter tolerance than @Miska notes above.

[Miska]

8 hours ago, Lightstream said:

The reason was that I used an XLR-to-SPDIF adapter to connect the AES/EBU output between my Simaudio MOON 180 Streamer and the DAC with an SPDIF cable (Oyaide FTVS-510). Neither the Neutrik AES/EBU to SPDIF adapter I used nor a no-name adapter worked. Also an adapter on the input side of the DAC (AES/EBU to SPDIF) using a AES/EBU Cable did not work which is no problem at all with other DACs. When I used the SPDIF cable and the SPDIF in- and outputs everything worked fine. Maybe a highly bred racehorse reacts more sensitively than the average horse... 

 

For AES/EBU <-> coax S/PDIF conversion, there are adapter boxes available, commonly based on pulse transformers. These do the needed level conversions too. There are also active converters that also support conversion to/from optical S/PDIF. You could take a look at some pro-audio shop for such (in Europe, Thomann is good place to start).

 

[mansr]

35 minutes ago, Samuel T Cogley said:

Thanks everyone for your responses so far.  I will say the best results I've had so far is from an iFI iLink, so there may indeed be something to the jitter tolerance than @Miska notes above.

Both voltage levels and jitter can be checked with a scope. If you have one, it would be interesting to see the results.

[Samuel T Cogley]

3 minutes ago, mansr said:

Both voltage levels and jitter can be checked with a scope. If you have one, it would be interesting to see the results.

 

I agree.  But appreciate that I would have to measure at least 3 sources that don't work and 3 sources that do to have enough data points to be useful.  I probably won't have the time and access to a scope for at least a couple of weeks.  Stay tuned.

 

EDIT: Don't I need a spectrum analyzer to measure jitter?

[mansr]

8 minutes ago, Samuel T Cogley said:

I agree.  But appreciate that I would have to measure at least 3 sources that don't work and 3 sources that do to have enough data points to be useful.  I probably won't have the time and access to a scope for at least a couple of weeks.  Stay tuned.

 

EDIT: Don't I need a spectrum analyzer to measure jitter?

There are many jitter metrics, not all of which require a spectrum analyser. Here's a useful article: https://www.sitime.com/api/gated/AN10007-Jitter-and-measurement.pdf

[Lightstream]

37 minutes ago, Miska said:

 

For AES/EBU <-> coax S/PDIF conversion, there are adapter boxes available, commonly based on pulse transformers. These do the needed level conversions too. There are also active converters that also support conversion to/from optical S/PDIF. You could take a look at some pro-audio shop for such (in Europe, Thomann is good place to start).

 

Thanks for the hint! I'll have a look at Thomann, he's practically around the corner from my place. 😀 I had used my Mutec MC-3+ reclocker as an active converter, but then I have to use a spdif cable from the Mutec to the DAC which I wanted to avoid.

[SQ5EBO]

Hi

I also experienced this problem and finally found a solution.  

Regards

Tomasz

[Lightstream]

19 minutes ago, SQ5EBO said:

Hi

I also experienced this problem and finally found a solution.  

Regards

Tomasz

Congrats! Will you tell us the solution?

[SQ5EBO]

https://sites.google.com/site/sq5ebo/topping-d50s-coaxial-mod?overridemobile=true

[sstenks]

I found this, there is solution.

Hope this will help.

Best regards

https://sites.google.com/site/sq5ebo/topping-d50s-coaxial-mod