2016-03-10 05:42 PM
Hi All,
I have two STM32F407 discovery boards. A version B2 board and a new version D1 board. The B1 starts on external power just fine with the same code but the D1 board does not. It only starts up after I connect it to a USB port while connected to the external power supply. The external power comes from at computer power supply giving it 5.01V while the USB power comes from my Windows 7 Desktop. I think it gives it 4.97V. Any idea what causes this weird start-up failure? #stm32f407g-discovery2016-03-10 06:46 PM
I'm pretty sure it isn't the difference in voltages.
If you are not using the USB/ST-LINK, disconnect it and the NRST from the target processor. I'm apt to think the issue is with the mbed firmware, rather than a board level issue. I have no RevD boards, perhaps ST can send me some?2016-03-11 01:16 AM
You can buy some at the usual distributors, they appear to have plenty - it appears the old ones (with STLINK/V2) are already sold out and were restocked with these new ones (with the ''mbed-compatible'' STLINK/V2-1).
Unfortunately, ST chose to confuse us (and presumably themselves too) with a poorly chosen naming system. The ''original'' F4 DISCO was labelled STM32F4DISCOVERY, the ''new one'' is STM32F407G-DISC1. Also, the F4DISCO's materials (product page, UM) refers to STLINK/V2-A, which is presumably the same as STLINK/V2-1 mentioned elsewhere. Another unfortunate fact is, that according to UM the serial inputs on STLINK/V2-1 are NOT connected to the target processor in any way (there are solder bridges on these lines on other STLINK/V2-1 - equipped DISCOs and NUCLEOs) and one is required to solder wires directly on the STLINK mcu's pins to use this - undoubtedly useful - facility. It would be nice if somebody from ST could comment on all of above. JW2016-03-12 10:10 AM
The problem was solved by upgrading the ST-Link Firmware from V2J25M14 to V2J26M15.
This same bug exists in V2J25M13 also.