Cambridge Audio CXA81 Review (Sample 2)

[amirm]

”After some investigation, they said that they could replicate my results at their China manufacturing facility but not their R&D lab.”

Can they now reproduce the problem with attaching a USB cable in their R&D lab? Or is there something that changed from prototype to production?

I have not heard back from them on this front. And no, I am not aware of any manufacturing changes.

 

[jasonhanjk]

The best result is always using the AP's input side ground to the DUT's star ground on the PCB; with a battery powered laptop.
If still did not see result as good as it's optical, I would classified it as a design issue.

 

[wwenze]

I'm surprised AP of all things actually has common-mode noise from USB since that thing is built so well. Then again it's below the -120dB mark. I'd bet that with a computer I'd see my harddisk activity at -90dB.

 

[beefkabob]

Kudos to Cambridge for dealing with you constructively. Looks like a significantly flawed but not actually broken product.

 

[amirm]

The best result is always using the AP's input side ground to the DUT's star ground on the PCB; with a battery powered laptop.
If still did not see result as good as it's optical, I would classified it as a design issue.

You can't battery power the AP and DUT. I also try many ways to reduce the impact of this with a dozen different approaches to grounding. The results you see are the best of those scenarios.

 

[jasonhanjk]

You can't battery power the AP and DUT. I also try many ways to reduce the impact of this with a dozen different approaches to grounding. The results you see are the best of those scenarios.

I see our setup are different.

I'm using a separate laptop as source.
Wav file created using Audacity with 24 bit 48kHz.
Player uses Groove at 97% volume (not Master volume).

 

[DMill]

Still not my cup of tea and I typically despise responses that echo others. But it is worth saying in this case nice job @amirm and Cambridge.

 

[PeteL]

You can't battery power the AP and DUT. I also try many ways to reduce the impact of this with a dozen different approaches to grounding. The results you see are the best of those scenarios.

Not a guarantee that it would change something, but the DUT is designed to be used grid powered. That’s what we want to see tested of course. Now laptops USB power suplies, and ports in general are far from being all equal. Some are very noisy. Powering the Laptop with the battery can to some extent help,but that said, filtering on the DUT should remediate that, and the product deserve to be criticised for it. it’s just that, as you mentioned, results may vary on the Laptop ports. If we want to minimise the laptop contribution on overall noise mr Jasonhankz suggestion is relevant.

 

[eriksson]

Many thanks Amir!
Since I have similar amp (Denon PMA1600) with integrated DAC I concur - USB is what I use to connect my various devices and it's a big deal if it's causing problems.

Now the question is, does the "bad" unit represent the quality of tiny fraction of CXA81 already sold or considerable ratio?

 

[aktiondan]

You could use something like this:

Hifime Hi-Speed USB Isolator - 480Mbps speed, USB 2.0, 3.0, 3.1

The Hifime High-Speed USB Isolator is designed specifically for audio and supports USB Hi-speed mode that is required by many Audio devices. It will prevent ground loops, prevent interference and protect your equipment against power surges.

hifimediy.com

I have the cheaper low-speed version and can attest to its performance in my setup. It absolutely does remove audible noise, likely from some kind of ground loop, since all this does in effect is isolate the grounds between the in and out. It may work here as well, even though that options appears built-in. Also that's just one more thing you have to buy.

 

[Chester]

Are we talking about the same situation as Rob Watts described with the ASR Mojo 2 tests?

“The issue here is the APx 555, in that the analogue oscillator is permanently on internally even when it's turned off - at least that's my supposition, as the APx 555 permanently creates analogue ground noise at 1kHz intervals and with careful measurements this signal will become visible. The issue is the APx 555 creates a 1kHz signal current, which flows out of the USB port, into the device under test (DUT) then back into the APx 555. Due to finite ground impedances, this creates an error voltage, which is then picked up by the APx 555 input ADC. But this error is nothing to do with Mojo 2, but a problem with the APx 555.”

 

[pseudoid]

It would be the best invention in the world. I use my tiller to attack them but they come right back up when you plantings are in there. And then becomes a tedious hand pulling.

Soon, our next generation is going to wonder what on earth you are talking about gardens.
Isn't it LA County that has just enacted one-day/week lawn watering; maximum of 8 minutes???

 

[laudio]

I think the real question is if the measuring equipment introduced the ground loop over USB or is it something one would expect using a real world USB device driving it that isn't an AP analyzer

 

[Doodski]

Soon, our next generation is going to wonder what on earth you are talking about gardens.
Isn't it LA County that has just enacted one-day/week lawn watering; maximum of 8 minutes???

LA is going to need a pipeline from the Columbia River if it keeps going where it and the Colorado River basin are heading.

 

[Snoopy]

Thanks for the second review. Having a usb cable connected shouldn't cause such problems. That's to bad really.


I've always used the CXN V2 (XLR) as DAC streamer.

 

[Bryan T]

I think the real question is if the measuring equipment introduced the ground loop over USB or is it something one would expect using a real world USB device driving it that isn't an AP analyzer

Presumably, I could test for this by listening for noise on a non-USB input with and without a USB cable plugged in, no?

On Analog 2 with the volume at 100% I can hear noise up to about 16 inches away from the tweeter. I unplugged the USB input and the max distance didn’t change.

 

[respice finem]

Fine? Not really yet IMHO, even if it will satisfy many customers.
Meanwhile not really terrible too, but the vote can't be changed.

 

[AndreaT]

Excellent analysis. The capability to output 2 v for a pre is indicative of flexibility of a device to adapt to different configurations. The mess created by a USB cable connection is a serious problem, as it is often how I use the USB input: the cable reaches the front of the unit for intermittent connection with a digital source. Thank you Amir!

 

[amirm]

Are we talking about the same situation as Rob Watts described with the ASR Mojo 2 tests?

“The issue here is the APx 555, in that the analogue oscillator is permanently on internally even when it's turned off - at least that's my supposition, as the APx 555 permanently creates analogue ground noise at 1kHz intervals and with careful measurements this signal will become visible. The issue is the APx 555 creates a 1kHz signal current, which flows out of the USB port, into the device under test (DUT) then back into the APx 555. Due to finite ground impedances, this creates an error voltage, which is then picked up by the APx 555 input ADC. But this error is nothing to do with Mojo 2, but a problem with the APx 555.”

His statement makes no sense as USB port that talks to the DAC is on the computer, NOT APx555. The analyzer has no direct support for USB. The measurement software uses computer's USB port for this purpose. Any indication of 1 kHz spikes indicates digital logic in the DAC bleeding into the analog section. USB micropacket length is 1 Kbytes and hence that cadence. He needs to buy the myriad of good DACs I have tested which show no indication of such spikes despite being measured identically.

 

[F1308]

Another CONCLUSION:
I am glad I went PA5. Terrific.