Help me find the faulty component

[hyperknot]

I have the following components, connected with XLR cables.
- SONCOZ SGD1 DAC
- AUDIOPHONICS MPA-S250NC XLR Power Amplifier (based on Hypex Ncore NC252MP)

The problem is the following. After some hours of listening, there is no audio for about 2-3 minutes. Doesn't matter what I do, change volume, etc. the sound doesn't come back, until a random time, at which it returns.

To make it more interesting, when I connected the SONCOZ DAC to a different amp over RCA cables, it works perfectly. It was used like this for the last few months, many hours a day, everything is perfect.

To make things even more interesting, when I connected a different DAC to the Audiophonicsamp over RCA cables (+ cheap RCA-XLR adaper) it also works perfectly.

I don't have an other balanced source, and there is about a week left from the warranty of the amp. The Soncoz is over warranty period.

Can you tell me how can I diagnose if it's the amp or the DAC which is faulty? If it's the amp then I can just send it back and have it fixed. Audiophonics said that they haven't heard any case like this with Hypex amps.

 

[mhardy6647]

Two thoughts pop into my mind.
1) Does the amplifier have some kind of thermal cut-out? The '2 to 3 minutes before resuming' makes it sound like a safety 'circuit breaker' is activating and then resetting itself.

This being said, your other tests make it sound like XLR vs. unbalanced is a contributing factor, which leads me to

2) Are your XLR cables wired "properly"? There have been two standards, i.e., there was no standard.
Technically, the most likely configuration (which is at least sort of a standard today) is:

s
source: https://www.clarkwire.com/pinouts/xlr-audio-pinouts

note their droll comment(s) vis-a-vis "standards".

Note that I am no expert and I am completely "shooting from the hip" -- but if I were in your shoes, these are the two places I'd start.
Good luck!

 

[-Matt-]

Possibly some sort of thermal cutout in the amp?

Does it cut out if you play it very quietly?

Does it take longer to cut out if played quietly?


Edit: Beaten to it by @mhardy6647

 

[mhardy6647]

Oh... looking around online for a manual for the amp, I noticed this somewhat concerning notice on the link you posted, @hyperknot:

3) you might want to check your signal (input) source for a DC offset(?).

 

[Killingbeans]

The problem is the following. After some hours of listening, there is no audio for about 2-3 minutes. Doesn't matter what I do, change volume, etc. the sound doesn't come back, until a random time, at which it returns.

Both channels at the same time?

My first guess would be a bad solder joint in one of the XLR cables, but it's extremely improbable to happen in both cables at the same time.

 

[Sokel]

Oh... looking around online for a manual for the amp, I noticed this somewhat concerning notice on the link you posted, @hyperknot:

View attachment 374439

3) you might want to check your signal (input) source for a DC offset(?).

Yep,DC should be checked right out of the box ,then after some time on and not connected and then plug it to the rest of the gear.

 

[hyperknot]

Both channels at the same time?

My first guess would be a bad solder joint in one of the XLR cables, but it's extremely improbable to happen in both cables at the same time.

Yes, both cables, perfect silence, no pop or anything.

 

[hyperknot]

Thanks for the ideas. The XLR cable is this: https://www.cordial-cables.com/en/products/cpm-fm

 

[hyperknot]

Can you tell me about the DC offset problem? How does it manifest itself and why is it a problem for me?

 

[Speedskater]

Off Topic, but that Clark Wire XLR pin out is incorrect.
Correct:
Pin 2 = Signal Positive
Pin 3 = Signal Negative
Pin 1 = Shield (may not be connected at receive end)

 

[-Matt-]

Can you tell me about the DC offset problem? How does it manifest itself and why is it a problem for me?

Not an expert, so perhaps others can confirm...

If the output of the DAC has a DC offset then the amp will be constantly working (perhaps quite hard) to amplify the DC level (i.e. deliver a constant current). Even though this won't be generating any sound it could generate a lot of heat in the amp. If the DC current is passed through to the speakers it could also heat and potentially damage the driver coils.

The notice posted above warns you that the amp doesn't have a decoupling capacitor and that it will therefore pass DC current to the speakers. In which case you should be careful that the amp input doesn't have a DC offset.

The notice tells you that the protection circuit (potentially what causes your cut-out) doesn't kick in until 12Vdc. For 8Ohm speakers this would mean 1.5A of current, and 18W of heating in the drivers. For 4Ohm speakers it would be 3A and 36W of heating.

I definitely don't think it is a good idea to pass this additional current (which won't contribute to the sound) through your speakers. So you should try to eliminated DC offset at the DAC output if possible.

 

[mcdn]

Off Topic, but that Clark Wire XLR pin out is incorrect.
Correct:
Pin 2 = Signal Positive
Pin 3 = Signal Negative
Pin 1 = Shield (may not be connected at receive end)

Indeed, so strange! The entire point of balanced signaling is that there is no “signal ground”. The shield should usually be connected at both ends but it makes no difference to the signals on pins 2 and 3

@hyperknot to test for a DC offset issue you can connect a multimeter to pin 2 and pin 3 of the XLR output from the DAC. It should read zero volts.

 

[hyperknot]

Thanks. So how do I measure it? Speakers disconnected, DAC on, playing nothing, Amp on, and I measure DC on the speaker binding posts?

 

[mhardy6647]

Off Topic, but that Clark Wire XLR pin out is incorrect.
Correct:
Pin 2 = Signal Positive
Pin 3 = Signal Negative
Pin 1 = Shield (may not be connected at receive end)

As you can tell, I did not double-check and, if so (not that I doubt you) -- I think this hammers the point home quite nicely!
All XLR innies and outties are not created equal unless one knows that they are.
I am mostly unbalanced* so I don't think too much about XLR connections... but when I am in that mode, I have, and use, one of these.

CPTS Cricket Speaker Polarity & Continuity Tester - Galaxy Audio

CRICKET POLARITY TEST SET

galaxyaudio.com

I am a trust but verify kind of guy. A Cricket is an inexpensive and extremely handy tool to have for the testing/verification of wiring/polarity of both loudspeakers and cabling.

_______________
* yeah, yeah -- laugh all you want! Everybody funny -- now you funny, too!

 

[mhardy6647]

The internet claims, with its typical resolute-ness, that the Clark pinout is the AES14-1992 standard, also commonly referred to as “pin 2 hot”.
That's all I've got.
I do know that there are plenty of components wired the in either configuration. E.g., I had to wrangle appropriate cabling for my Otari MX5050 bii tape deck -- bless Otari's collective hearts.

 

[kemmler3D]

Oh... looking around online for a manual for the amp, I noticed this somewhat concerning notice on the link you posted, @hyperknot:

View attachment 374439

3) you might want to check your signal (input) source for a DC offset(?).

That was my first thought also. A little DC could be causing it to overheat after a while.

 

[hyperknot]

So how do I test DC on the DAC's outputs? Do I need to play a "silent.wav" with normal output levels and measure with a multimeter?

 

[hyperknot]

Great news, as I was testing it with higher volumes today and possibly that the outside weather is also warmer now, I was able to reproduce it with a different DAC as well. So it's definitely the Hypex amp. Sending it back in warranty.

 

[hyperknot]

I measured the DAC, playing "silence" on normal levels. DC offset is 0.2 mV.

 

[mcdn]

The internet claims, with its typical resolute-ness, that the Clark pinout is the AES14-1992 standard, also commonly referred to as “pin 2 hot”.
That's all I've got.
I do know that there are plenty of components wired the in either configuration. E.g., I had to wrangle appropriate cabling for my Otari MX5050 bii tape deck -- bless Otari's collective hearts.

The pin 2 and 3 bits are fine, it’s the confusion of shield and ”signal ground” that’s the issue. All XLR cables should have pin 1 (and hence the cable shield) connected to the chassis as closely as possible, unless the designer if the component has a very good reason to do it another way.