Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Design flaw causing power failure
#1
Hello,
I registered on this forum to share with you the conclusions resulting from the repair of the Phantom I device by Polish YouTuber Daniel Rakowiecki.
Yesterday, he published a video of repairing a device that had a power issue, and he identified a design flaw by Devialet engineers.

Youtube video title: "Najdroższy głośnik BT na świecie ma poważny BŁĄD elektroniki !!"
Link: https://www.youtube.com/watch?v=U9n59HaGavI

Video is in polish language but You can enable automatic closed captions (they are sometimes not 100% precise though)

He is identifying the problem in 18:20 (mm : ss)

In a nutshell, Texas Instruments TPS560200 Step-down converter used to generate 3,3V for the device has EN input directly connected with 15V line (VIN).
This is designed by Devialet against Texas Instruments specification (maximum voltage is 7V)
https://www.ti.com/lit/ds/symlink/tps560200.pdf

After replacing TPS560200 module to new one, device was fixed but Daniel is speculating that it will soon brake again.
That is why in 21:00 he is lifting EN input chip leg (it became unconnected)
Device is still working and now it meets texas instruments specification.

In 37:48 he is also explaining why Devialet engineers made this design error.
He is saing that 99% of step-down chips are activated using high state (ie. VIN) on EN input
This TPS560200 is unique.
To activate it You need to leave it unconnected. EN input here is used for soft-start with additional capacitor (but only to 7V maximum)

So, in short, the Devialet engineers did it out of habit without checking the manufacturer's chip specifications.
Reply
#2
(10-Jul-2023, 17:25)MiLAP Wrote: Hello,
I registered on this forum to share with you the conclusions resulting from the repair of the Phantom I device by Polish YouTuber Daniel Rakowiecki.
Yesterday, he published a video of repairing a device that had a power issue, and he identified a design flaw by Devialet engineers.

Youtube video title: "Najdroższy głośnik BT na świecie ma poważny BŁĄD elektroniki !!"
Link: https://www.youtube.com/watch?v=U9n59HaGavI

Video is in polish language but You can enable automatic closed captions (they are sometimes not 100% precise though)

He is identifying the problem in 18:20 (mm : ss)

In a nutshell, Texas Instruments TPS560200 Step-down converter used to generate 3,3V for the device has EN input directly connected with 15V line (VIN).
This is designed by Devialet against Texas Instruments specification (maximum voltage is 7V)
https://www.ti.com/lit/ds/symlink/tps560200.pdf

After replacing TPS560200 module to new one, device was fixed but Daniel is speculating that it will soon brake again.
That is why in 21:00 he is lifting EN input chip leg (it became unconnected)
Device is still working and now it meets texas instruments specification.

In 37:48 he is also explaining why Devialet engineers made this design error.
He is saing that 99% of step-down chips are activated using high state (ie. VIN) on EN input
This TPS560200 is unique.
To activate it You need to leave it unconnected. EN input here is used for soft-start with additional capacitor (but only to 7V maximum)

So, in short, the Devialet engineers did it out of habit without checking the manufacturer's chip specifications.

Correct me if I am wrong, but does it mean that every Phantom speaker will sooner or later break?
Reply
#3
Well, it depends...
If they are using same step-down converter in all Phantom speakers (and there are very few other (none?) for such high VIN voltage - according to Daniel Rakowiecki video) in the same way (EN input feeded with 15V) then they are making that against chip producer specification (doubles the maximum allowed voltage).
It could cause faster step down chip wear / higher risk of malfunction for sure.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)