1
\$\begingroup\$

I'm trying to understand a schematic document about a USB 3.0 controller (Renesas UPD720201.)

They are using two converters (buck converter from 12V to 5V and boost converter from 3.3V to 5 V) to obtain the same voltage output. I don't understand why.

enter image description here

P.D: Both voltages come from PCIE Connector, maybe 12V goes down in any time and only is used 3.3V supply?

EDIT: I have added more images from schematics.

enter image description here

enter image description here

\$\endgroup\$
10
  • \$\begingroup\$ Do you believe it to be an error? Are you privy to the original design requirements? \$\endgroup\$
    – Andy aka
    Aug 2 at 11:40
  • \$\begingroup\$ This schematic is an original design from Renesas. \$\endgroup\$
    – Juanma
    Aug 2 at 11:46
  • 7
    \$\begingroup\$ One creates 5V from a 12V supply, the other creates 5V from a 3V aux supply. With what looks like to be circuitry to turn off the 3V supply when the 12V supply is present (? schematic partially cuts off at that point) \$\endgroup\$ Aug 2 at 11:48
  • \$\begingroup\$ Is this on a dev board? Perhaps there's two circuits for flexibility? \$\endgroup\$
    – mike65535
    Aug 2 at 11:56
  • \$\begingroup\$ Just an educated guess… it can do both since there are two different use cases and it has to work with whatever it comes in. And by experience buck-boosts are quite more tricky do do efficiently. Or maybe one it's a bootstrap supply and the other converter takes over later. Without the full device plan it's difficult to say \$\endgroup\$ Aug 2 at 12:05
3
\$\begingroup\$

It it likely that the 3.3v line here is connected to the 3.3v auxiliary (standby) supply from the PCI-E connector. The USB3.0 controller you mention is designed to work with PCI-E.

I would suspect this is related to some kind of 'wake-up' functionality for whatever PC it is connected to. The PCI-E bus offers a WAKE# signal.

Alternatively, it might be there to offer a trickle charge for connected USB devices, similar to the feature seen on modern Gigabyte motherboards.

\$\endgroup\$
2
  • \$\begingroup\$ Yes, it is designed to work with PCI-E. But as i understand this, might 3.3 or 12V supply go down in any time? Because they are using other converter to obtain 3.3V (3.3V_DUAL) as Controller Supply but they are using 3.3V from PCI-E port to obtain 1.05V to supply controller too (it doesn't make sense for me). \$\endgroup\$
    – Juanma
    Aug 4 at 7:04
  • 1
    \$\begingroup\$ Based on the Wikipedia article linked, it seems that the PCI-E slot provides, 12v, 3.3v and 3.3v AUX, the last of which is on during standby. Without a background in actually designing these devices, my answer is purely speculative. \$\endgroup\$ Aug 5 at 1:12
1
\$\begingroup\$

Since the two converter outputs are connected together, this can produce +5V from whichever supply is present. In other words, if the 12V supply is absent, then +5V is derived from the 3.3V supply, and vice versa.

D9 and D10 form a "diode-OR" configuration which "steers" the greater of the two converter outputs to the ultimate 5V output, and simultaneously prevents the output of either converter from injecting current into the other.

You might think that this would cause the output to drop under 5V (by 0.7V), but cleverly the designers have taken feedback from after the diode, to compensate for this.

Edit: I suspect that the designers also configured one of the converters to produce a slightly greater output voltage than the other (like 5.0V and 5.05V), so that when in operation, the other converter senses an "overvoltage", and ceases its switching, becoming effectively idle. This function is perhaps implemented in the large transistor network preceding both converters, though.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.