9
\$\begingroup\$

While routing my CY7C68013A based PCB I just realized that regular USB B receptacle and mini USB equivalent have D+ and D- effectively swapped.

It seems that the chip was prepared for regular size receptacle (DP is on the left if looking towards the chip), while for mini USB I have to route one of the tracks under the receptacle.

Is it really like this, or did I mess something up? How am I suppose to keep the tracks impedance and length constraints, while one of the tracks needs to go a different way, around the receptacle pads?

\$\endgroup\$
8
  • \$\begingroup\$ Most likely getting impedance and length matching perfect on those lines wont make that big of a difference, especially if the lengths aren't too different. I would like to see what a "professional" has to say though. \$\endgroup\$
    – Kellenjb
    Apr 20 '11 at 14:13
  • \$\begingroup\$ Which encapsulation (package) are you using? \$\endgroup\$ Apr 20 '11 at 14:24
  • \$\begingroup\$ For the connector? It's something like this: farnell.com/cad/452550.pdf The chip itself is TQFP-100 version. \$\endgroup\$ Apr 20 '11 at 14:33
  • \$\begingroup\$ Be careful when you're laying out this part. Do not run traces underneath, between the mounting lugs, because the metal is sharp and can scratch through the solder mask. \$\endgroup\$
    – markrages
    Apr 20 '11 at 15:23
  • \$\begingroup\$ The USB Spec is Extremely conservative. I've done some horrible things to USB cables, and still managed to run full USB 2.0 over them. \$\endgroup\$ May 7 '11 at 5:58
5
\$\begingroup\$

I've noticed the same problem when routing boards with the FTDI FT232 chips. Here's what the CY7C68013A datasheet says about routing those signals:

  • DPLUS and DMINUS trace lengths should be kept to within 2 mm of each other in length, with preferred length of 20 to 30 mm.
  • Maintain a solid ground plane under the DPLUS and DMINUS traces. Do not allow the plane to split under these traces.
  • Do not place vias on the DPLUS or DMINUS trace routing.
  • Isolate the DPLUS and DMINUS traces from all other signal traces by no less than 10 mm.

I would suggest the zig-zag method, but only if it that little detour around the pin added more than 2 mm to the trace length.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ FT232 chips are full-speed devices, running at a maximum of 12 Mbit/s. The Cypress chip referenced is a USB 2.0 High-speed device, which can do up to 480 Mbit/s. The routing requirements are similar (at least they're both using differential signals), but they won't be identical. \$\endgroup\$ May 5 '11 at 14:47
  • \$\begingroup\$ Sorry, I didn't make it clear that those recommendations are from the datasheet for the CY7C68013A that the OP is using. I don't think the FT232 datasheet has such strict requirements, as you suggested. \$\endgroup\$
    – ford
    May 6 '11 at 2:38
2
\$\begingroup\$

Are you sure that D- and D+ are swapped? I checked the USB2 specification. ECN 1 describes the miniB connector, and I see the same order for the regular and the mini: Vbus, D-, D+, Ground.

If the connections are swapped, can't you add a few zigzags in the shorter trace to make it as long as the other? (I don't think it is really necessary, but it might give you some rest :-))

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ Pins appear swapped at the PCB level, not at the plug level, due to different receptacle geometry - unless I've messed it up :) \$\endgroup\$ Apr 20 '11 at 14:58
  • \$\begingroup\$ @Czajnik, could it be that it's intended to be mounted on the bottom of the board and someone is just tricking it? \$\endgroup\$
    – kenny
    Apr 20 '11 at 15:45
  • \$\begingroup\$ I guess not, as PCBs with components on one side are obviously cheaper to manufacture. The receptacle I want to use is a standard, commonly used component. Perhaps the perfect impedance and length match is not as critical as I think? \$\endgroup\$ Apr 20 '11 at 19:15
1
\$\begingroup\$

This one docs states that the prefered mounting for microusb-b is on the underside. http://portal.fciconnect.com/Comergent//fci/drawing/10103593.pdf

I also had the problem when mounted on the topside that D+ D- would not line up. Had to put the chip below the connector instead and run the traces between shield pins.

\$\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.