USB-IDE puzzle
A friend of mine (spamsink) came to me with a puzzle. A USB-to-IDE puzzle. He had a little USB-to-IDE adapter just like this one:
The puzzle
What was puzzling is that when the adapter was connected to a USB port, it was detected as a full-speed (12 Mbps) device. In fact, Windows would complain that a high-speed device was connected to a full-speed port. Except that the device was connected to a real USB 2.0 high-speed port. The solution
First I carefully examined all device descriptors and found nothing wrong. I then concluded that the problem was at the hardware level and returned the adapter to spamsink. He was able to take the device apart and get to the PCB inside. With his permission to disassemble the device I dug in. Here is what I found inside:
The adapter is based on Genesys Logic GL811E chip. I took a look at the datasheet and started to figure out the schematic. I first determined that DPF and DMF (full speed 12 Mbps USB data pins) are connected directly to the USB D+ and D-. R4 and R5 are impedance matching resistors to DPH/DMH (high-speed 480 Mbps USB data pins). Both resistors looked fine and measured 33Ω. Next to them I saw R3 which is connected to pin 23 (RREF) of GL811E. From the datasheet it seems that this is some sort of a current reference resistor for the USB 2.0 transmit/receive cell. The datasheet mentions that:
(****) RREF must be connected with a 510 ohm resister [sic] to ground.
On the board I clearly saw the resistor marking: 514 (510 KΩ). I removed the part and replaced it with 470Ω - the only one I could find at the moment (later I replaced it with a proper 510Ω resistor). Bingo! Connecting the adapter to my Linux box I got the following display, showing that the device was properly detected as USB 2.0:
I ran some transfer rate tests and got 16.2 MB/s on writes and 19.2 MB/s on reads, clearly in the USB 2.0 territory.
To summarize, the problem was that the manufacturer screwed up and used resistor value 1000 times larger, than necessary. Here is a photo of the culprit:
The puzzle
What was puzzling is that when the adapter was connected to a USB port, it was detected as a full-speed (12 Mbps) device. In fact, Windows would complain that a high-speed device was connected to a full-speed port. Except that the device was connected to a real USB 2.0 high-speed port. The solution
First I carefully examined all device descriptors and found nothing wrong. I then concluded that the problem was at the hardware level and returned the adapter to spamsink. He was able to take the device apart and get to the PCB inside. With his permission to disassemble the device I dug in. Here is what I found inside:
The adapter is based on Genesys Logic GL811E chip. I took a look at the datasheet and started to figure out the schematic. I first determined that DPF and DMF (full speed 12 Mbps USB data pins) are connected directly to the USB D+ and D-. R4 and R5 are impedance matching resistors to DPH/DMH (high-speed 480 Mbps USB data pins). Both resistors looked fine and measured 33Ω. Next to them I saw R3 which is connected to pin 23 (RREF) of GL811E. From the datasheet it seems that this is some sort of a current reference resistor for the USB 2.0 transmit/receive cell. The datasheet mentions that:
(****) RREF must be connected with a 510 ohm resister [sic] to ground.
On the board I clearly saw the resistor marking: 514 (510 KΩ). I removed the part and replaced it with 470Ω - the only one I could find at the moment (later I replaced it with a proper 510Ω resistor). Bingo! Connecting the adapter to my Linux box I got the following display, showing that the device was properly detected as USB 2.0:
I ran some transfer rate tests and got 16.2 MB/s on writes and 19.2 MB/s on reads, clearly in the USB 2.0 territory.
To summarize, the problem was that the manufacturer screwed up and used resistor value 1000 times larger, than necessary. Here is a photo of the culprit: