Honored Contributor
Honored ContributorI have created a USB-Blaster clone using an FT245R module (UM245R) and an FT232H module (UM232R) interfaced to a BeMicro-SDK. The FT245R works fine under Windows and Linux, but the FT232H does not ... jtagd SEGVs. If I put the FT232H clone behind a USB 1.0 hub, then it works fine. The only difference between lsusb with the UM232H behind the hub or directly connected is the wMaxPacketSize which is 64-bytes for USB1.0 and 512-bytes for USB 2.0. I believe this is the same issue the Cypress FX2 clones face, and that is why they report their interfaces with 64-byte packet sizes.
Both of these clones work fine under Quartus II Windows. The hardware works fine under Linux, as using libftdi I can put the JTAG TAP in bypass mode and send 16MB of data through no problem. The issue is with Altera's Linux implementation of their JTAG server ... if only they'd just publish the source, we could fix it for them ... Here's some additional lsusb outputs; Altera USB-Blaster, Terasic USB-Blaster/DE2/DE0-nano, and an Arrow BeMicro-SDK. I don't see anything particularly wrong in your USB-Blaster-ARM clone lsusb output. Can you provide a link to your clone? I've never heard of that one. Daixiwen - can you post the output of lsusb for your EEBlaster - if it uses the UM232H device, then perhaps its suffering the same problem as my clone - try putting it behind a USB hub. Cheers, Dave
