What the heck happened? (P5 acting up after a fraction of a second power failure)

Went to power up my system this evening, and failed. The P5 was on, front panel screen on, but no power to components. I checked the display, and it showed output distortion in excess of 6%. I powered everything down, checked all the connections, turn off the unit at the back, turned it back on in the back, and pushed the reset button. Voila. The only thing I can think of that was out of the ordinary today was an interruption of power that lasted about half a second earlier this evening, well before I wanted to listen, when nothing was powered. The P5 was in its normal standby mode, rear switch on, but no display or blue light on front panel. Thoughts?

Just to be clear, everything is functioning normally now?

Listening right now. Perfect, like nothing happened.

Find a piece of wood and touch it. Maybe you just had an innocuous microprocessor hiccup.

Will do. My theory is that the half-Second power outage or surge may have tripped the P5’s protection. Anyway, all’s well that ends well.

Invoking a protection mode should not cause the THD out to go up like that.

I’m only guessing.

Invoking a protection mode should either shut down all the outputs, or by-pass the re-generator and directly connect the input to the output. A correctly functioning protection mode shouldn’t see something worse on the output than what’s on the input.

Speaking for myself: I’m not taking any chances. Not in South Africa you don’t! So I have a 3Kva pure sine UPS hardwired in my DB board driving my P20 in battery mode 24/7. Soon to be augmented by at least 2x12V 200aH deep cycle batteries for longer run time.

A reasonable guess.

It happened again tonight. A fraction of a second power outage took place, but this time the P5 didn’t go dark. It powered on, but outputted only 113v for an input of 121v, and 6.4% THD when the input level was only 1.7%. Fixed it the same way as before: powered off the unit with the remote, threw the power switch in the back, waited a bit, resumed power with the switch, hit the reset button, and turned on the remote. It’s back to normal now, with the output showing .1% THD.

Last night, yet again, same thing happened as in 9/18. Powered-up, showed over 6% THD at output for 1.6% or so THD input. Shut down both with remote and rear panel switch, waited a few seconds, restarted and normal service resumed: 1.7% THD in, .1% THD out. And again, there was a very short duration (1/2 second tops) power interruption earlier in the day, when nothing was playing but the P5 was in standby mode. Any PSA folks care to speculate?

The fraction of a second power outage, replete with whatever noise is on the line, apparently scrambles the brain of the unit - as evidenced by everything going back to normal upon reboot.

I would not give it any further thought.

Hmm, that may be, @Elk, but I think we may be able to agree it shouldn’t happen. Perhaps PS Audio can find a FW fix for this, hopefully.

I do not think we have enough information to agree, or disagree, as to whether it should happen. We do not know to what the equipment is being subjected.

Very short power interruptions, followed by spikes, noise, etc. will scramble all sorts of equipment - and reasonably so.

If the unit operates fine for 10 months, acting up only under these unique circumstances, I tend to believe something is odd about these rare outages/spikes/brown outs rather than assume something is wrong with the device.

I do not expect any piece of electronic equipment to always operate flawlessly, regardless to which it is subjected power-wise.

But perhaps PS Audio has some ideas.

1 Like

Well, I guess my view comes from a career in system design for highly reliable systems. Any interruption of that type could be dealt with by a properly designed recovery sequence that would reinitialize parts of the system that are needed to come back up in proper working order. We do things like this where necessary. But it would potentially add cost to the design.

As you suggest, I am certain additional features could be added.

I doubt the cost would be justified in light of the non-critical application of powering home audio, although some may argue no expense is to great when pursuing audio perfection. :slight_smile:

Just got with the team a little, and Elk seems to have hit the nail. The brains inside (the oscillator measuring the incoming waveform form) is getting a little confused. It’s likely not readjusting the phase properly. Flipping the switch in the back will reset said brains.

Thank you for the follow-up, including the specific cause.

Thank you @jamesh. Have you also determined if any failure modes that would cause this off-nominal behavior can in any way result in the regen outputs presenting any danger to the connected components?