0
\$\begingroup\$

Further to my previous question, I also noticed that the crystal there was a 12MHz where this should have been 6MHz (due to an inventory error - the wrong part was fitted). I've removed the 12MHz crystal and fitted a 6Mhz crystal using a hot air tool but now it doesn't oscillate. It kind of looks like it's trying - the scope shows a fair bit of noise and occasional flashes of what looks like about 6MHz waves but it doesn't oscillate stably.

This is the datasheet for the 12MHz crystal that produced a stable sine wave and this is the datasheet for the 6MHz crystal that doesn't.

As far as I can see, the major difference is that the 12Mhz crystal has 40 ohm ESR while the 6MHz crystal has 80 ohm ESR. As far as I can tell, my rework hasn't damaged anything - the load caps and series resistance are still doing the right things - but it's a bit hard to tell if I've damaged the crystal or the IC.

What is my most likely problem here? Is the higher ESR of the lower-frequency crystal likely to be a problem? Or is it more likely that I have damaged a component with my rework?

The crystal has a 1.5k resistor in series with it (cribbed from the TUSB2046B reference design). I'm tempted to short this out to try to reduce the losses in the oscillator and see if that improves things - is this a useful thing to try?

\$\endgroup\$
  • 2
    \$\begingroup\$ Short the R78 resistor and see. ESR of 80 Ohms might ask for too much of negative resistance from the circuit, and the oscillator won't start. Also keep in mind that if the hub is not connected to active host and is not enumerated, it will fall into SUSPEND mode in 3 ms time frame, and the oscillator might stop. To get out of the SUSPEND, you just keep the upstream D+ shorted to ground, emulating USB_RESET bus condition. \$\endgroup\$ – Ale..chenski Jul 10 '18 at 16:35
  • \$\begingroup\$ Ah, I hadn't noticed that detail. As well as changing the crystal, I've added a 1.5k resistor between 3.3V and D+ - so this might explain it now going into suspend mode when the last version didn't. I'll try it. Thanks. \$\endgroup\$ – Tom Jul 11 '18 at 9:20
  • \$\begingroup\$ Yes, the device was going into suspend mode - tying D+ to ground brings out a nice 6MHz sine wave on the crystal. If you'd like to write it up as an answer, I'll mark it accepted. \$\endgroup\$ – Tom Jul 11 '18 at 9:58
1
\$\begingroup\$

ESR of 80 Ohms might ask for too much of negative resistance from the circuit, and the oscillator won't start. You might need to reduce or remove the R78 resistor.

Also keep in mind that if the hub is not connected to active host and is not enumerated, it will fall into SUSPEND mode in 3 ms time frame, and the oscillator might stop. The oscillator is usually shut down to meet low-power requirement in USB SUSPEND mode.

To get out of the SUSPEND, a simple method is just to keep the upstream D+ shorted to ground, emulating the USB_RESET bus condition. When seeing the USB_RESET on the bus, every USB device must wake up, and the oscillator should start.

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ The hub device was going into suspend mode - waking it up did the trick. \$\endgroup\$ – Tom Jul 12 '18 at 10:24

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.