Fix your AC leak

[One and a half]

Curious as to why SMPS wall warts are no good for audio, for a recent trip preparation, I thought to measure in a typical CA setup and report on the findings. Suspicion arose when connecting a portable drive to the computer, I could see a small set of sparks when connecting the USB cable. It wasn't static discharge and the sparks would continue.

Out with a DMM (digital mutlimeter) to measure the voltage between the USB connector's case and the computer, a MacBookPro 2012 model 6,5. Wow. With a Fluke 87V, the measurement showed 99.7Vac to earth from a 230V/12VDC supply. Removed the cable from the portable drive and measured just the pin connector from the SMPS wall wart. Yes, that's where it came from 99.7V to earth from the SMPS. Now that's just the HDD power supply, what about the rest, Logitech 5V powered hub, ifiUSB and the DAC, m920.

 

 

Connected altogether in one system as shown in the drawing, the best results were obtained when the charger for the computer was disconnected. This allowed everything to float and be referenced by the DAC, however it wasn't quite zero as expected. Leaks occur in the SMPS and superimpose the AC crud on the DC power supply inside the components. There's no true galvanic isolation, so the 0V of each component is then tied to the other forming a chain, propagated by the shield voltage of the USB cables to make things worse. Achieving 0V on the shields of USB cables or the 0V of the PSU is difficult, as each one couples ot the other. The ifi's original power supply was nothing special, it leaks as well.

 

 

 

On its own, the HDD connected to the computer with the computer's charger unplugged, and using the internal DAC of the computer created a bad hum when using the headphones. The hum dissapeared (well, really went inaudible) when the charger was re-connected to the computer. In this case the coupled 99V was referenced to earth, more or less shorting the leaks to earth.

 

 

All goes to prove a linear (transformer, rectifier, caps, regulation) is the about the best way to supply small devices and avoiding hum and crud leaks from wall warts and their SMPS supplies.

[mayhem13]

Non audiophile grade........gotta love threads like this. There's been a very sound and sufficient reference for electronic performance beyond consumer grade......it's called lab grade and the accuracy of instrumentation connected has proven itself outside the boundaries of our solar system and inside the tiniest of particles. No ground breaking going on here at CA. Best to stick to subjectivity where no proof of concept is required.

[One and a half]

Non audiophile grade........gotta love threads like this. There's been a very sound and sufficient reference for electronic performance beyond consumer grade......it's called lab grade and the accuracy of instrumentation connected has proven itself outside the boundaries of our solar system and inside the tiniest of particles. No ground breaking going on here at CA. Best to stick to subjectivity where no proof of concept is required.

 

It's good to see your cynical self has not changed.

[mayhem13]

Not cynicism, just plain ol' reality really. The consumer electronics industry has been sticking it to people for years....I'm agreeing with you. The problem I have is depending on 'audiophile' manufacturing and design as an answer. It's market driven and marketing steers the ship and profit is the navigator. It's why we have crappy USB audio in the first place.....avoid the license fee of the FAR superior HDMI protocol but instead develop and market the noisiest interface in the computer industry?...really?.... If you examine what HDMI is capable of, you might find me just a tad bit less cynical. Our past experiences with HDMI are a prime example where early adaptations of the protocol were less than acceptable as they were cost driven. Even the latest protocol doesn't scratch the surface of what is possible with HDMI.

[One and a half]

OK, I'm on the road for the moment, and will perform the same conditions, but use HDMI. Have a few receivers about the place and BD players with DSD audio capability. The audio consumer only has USB, S/PDIF/HDMI/RCA/Ethernet/XLR/TRS if you're lucky ST Fiber as interfaces. You'd think after so many variants we would have got it right by now. I really like XLR, a true balanced system, big chunky connectors, polarised, noise immune, all it needs is a high bandwidth with error correction for DSD. If it can do DSD not DoP, it can do anything.

 

Anyway, USB lends itself to produce many fine USB cables which can clean up the mess, well try to. The reason for the mentioning the non audiophile distributor for the AC was it was I had laying about, if I would have used a pukka design, the results should be much the same, maybe a cleaner 98V instead of a fuzzy one.

A conditioner could make a difference, however, what I measured against a Monster Power Center earth point and a real earth point to a shell of a computer's USB out were very different. There's something in the earth line which shouldn't be there, will report on what I find.