We’re sorry, but in building a few more kits, we’ve noticed that there are occasional problems with a kit not working. At first, we figured that it was just some random errors in a handful of boards. However, with the help of the new oscilloscope, we’ve found the real culprit – on the D- line of the USB connection, there is a 10K pull-up resistor. It turns out that 10K is probably on the very limit of the effective range. Instead, it should be more along the lines of 1.5K Ohms, although anywhere between 1K and 5K should probably work.
The issue is that the 10K resistor is just doesn’t let quite enough current through to be able to pull the data line up to the “High” state and hold it there on it’s own. Depending on the hardware it’s connected to, the scope is showing that during long periods of being held “High”, with the 10K, the line drifts back to a semi-high state, where the host is presumably holding the line just too low below the 2.8V cutoff for the USB spec.’s High mark, or maybe just letting it float. At any rate, voltages that individual hosts see as “High” can vary surprisingly widely. For instance, some computers will see a 5v signal as “High” and some won’t – the spec. sets “High” at 2.8v-3.6v, but USB puts power out at 5 volts. Figure that one out. Unluckily, it turns out that the 10K resistors we’ve been using have been bringing the lines down to around 2.7-2.8v in a few cases, while being fine in others.
So we’re really sorry about all of this. If you’d like us to, we’re more than happy to send you some replacement resistors, or if you’d like, we’d be happy to pay the shipping and replace the resistors ourselves, test out your system, and update all the firmware for you. Just send us an e-mail, and we’ll work everything out.
If you’ve got some resistors lying around, here is the resistor in question that needs replacement:
