USB malfunction ?
in [Repair]
|
Prev: instructor solution manual for Reinforced Concrete: Mechanics and Design (5th Ed., James G. MacGregor & James K. Wight)
Next: Rob@welzijnpernis.speedlinq.nl
From: N_Cook on 3 May 2010 09:38 I know nothing of USB function other than what I've just read on usb-in-a-nutshell.pdf and the datasheet for Cypress AN2131QC USB chip . I downloaded legacy installer.exe and firmware stuff from www.event1.com/Support/EZbus/Updates/firmware.htm for an Ezbus digital mixer. Unzips and apparently functions on the pc. Cannot connect PC to the Ezbus, to reload firmware, as not recognised hotplugging USB and if plugged in before booting PC then it hangs the PC. On mixer on its own, USB D+ line is low during POST and then goes to and stays at 3.3V when presumed corrupted system code starts up. The main controller hangs up at this point also (2 of the 3 clocks present but third comes from a stuck high pin) and no function of any sort on the mixer . Could corrupted sysyem code make the USB hang like that or could corrupt USB chip hang the main control? What to look at next?
From: N_Cook on 4 May 2010 03:05 Both in the datasheet and the mixer 1K5 between G+ pin and DISCONnect pin, pins 79 and 1 ,of 80. Desoldering it and DISCON goes high but not D+ now, so will have to retry connecting pc to mixer
From: N_Cook on 4 May 2010 03:37 G+ -------> D+
From: N_Cook on 5 May 2010 03:56 that 1K5 will have to go back in , It is the first point of recognition for USB linking. Next to find why DISCON is permanently H.
From: Franc Zabkar on 8 May 2010 04:06
On Wed, 5 May 2010 08:56:58 +0100, "N_Cook" <diverse(a)tcp.co.uk> put finger to keyboard and composed: >that 1K5 will have to go back in , It is the first point of recognition for >USB linking. Next to find why DISCON is permanently H. My understanding is that the pullup resistor on D+ signals to the host (your PC) that a USB device has been plugged in. The host then attempts to enumerate (ie identify) the device. If the device takes too long to go through its POST and initialisation routines, then the host may time out. In such cases the device may implement an active pullup which is disabled during its POST. When the device is ready, the pullup resistor is enabled, allowing enumeration to begin. AIUI, the normal state of DISCON# is high. See this EZ-USB Development Board (circuit on pages 5 & 6): http://www.minford.ca/MF3001EZUSBManual.pdf I found this detailed document: The EZ-USB USB Integrated Circuit Technical Reference: http://www.digchip.com/data/115/115-00027-0-AN2121S.pdf It states that EZ-USB devices can power up in RAM-only mode, in which case the host downloads runtime code into the 8051 CPU's internal 8K RAM. During this time the device identifies itself as a "Default USB Device". After the code has been downloaded, the device disconnects and then reconnects, this time renumerating as the device defined within the downloaded code. Page 41 describes the EA (external access) pin. I think this may be the key to how the device behaves. The tech ref also talks about mask ROMed versions. Perhaps there is a way to force the EZ-USB chip into RAM-only mode??? If so, then this will remove any corrupt firmware from the equation. You could then use UVCView to see how the device enumerates under these circumstances. You could also disconnect the serial EEPROM, if any. The EEPROM would contain Product and Vendor IDs for the manufacturer of the mixer. If the EEPROM is absent, then the EZ-USB chip will ID with the PID and VID of 0x0547 (Cypress Semiconductor) and 0x2131 (EZ-USB), respectively. - Franc Zabkar -- Please remove one 'i' from my address when replying by email. |