FDD-UDD U144K

[nretro]

By accident (resp. stupidity) I bought a FDD-UDD U144K.

Looks quit similar, but it's not obvious how the jumpers have to be connected to flash a new firmware. It has got a 7x2 pin header labeled J0, J1, ... J6 and a 5x1 pin header labeled 07,46,47,48,49. Did anybody already figure out how these relate to the headers of the standard gotek?

[nretro]

Here it is. Looks similar enough... tough it is probably better to check first.... Thanks for looking into it.

[nretro]

Is there a schematic for the usual gotek? I also have two of those, but it would be easier to check if I would not have to reverse engineer those first. I mean the zero hypothesis would be that they did not change the headers.

[nretro]

Just for the fun of it, I checked and found out, that the lower pin of the 1x5 connects to the same pin of the processor. So there is hope...

[nretro]

46 (2nd from below) connects to the third pin on the upper side of the CPU (from left). Usual gotek connects this to the 6th pin on the upper side.
47 connects to the 2nd pin on the upper side of the CPU.
48 connects to the 3 digit display where it says 3v3.
49 connects to the upper most pin on the left side of the cpu.

[nretro]

On the 1x5 header there is NRST, and then CPU pins 46-49: PA13, VSS2. VDD2, PA14. PA9 and PA10 (USART1) are not connected at all. Both methods to flash the firmware use USART1. Is there an alternative or do i have to solder a connection to the CPU?

[nretro]

It does seem to support some kind of flashing via USB. When both buttons are pressed at startup, the display shows U00 -> 126 -> F01. But I could not find any suitable files on the internet or on the supplied CD.
EDIT: No, seems to be a formatting utility. Writes an "IBM 6.0" MBR and a partition table with one partition spanning the whole USB stick.

[drdpj]

PA13 & PA14, vdd and gnd will let you program it with an stlink v2 adapter.

…

On Wed, 29 May 2019, 22:09 nretro, ***@***.***> wrote: It does seem to support some kind of flashing via USB. When both buttons are pressed at startup, the display shows U00 -> 126 -> F01. But I could not find any suitable files on the internet or on the supplied CD. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#232>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AATU2PCLRCHCNJJX44NRQJLPX3WJNANCNFSM4HQGI2AQ> .

[keirf]

Yes the 1x5 is a programming header. I've never used the jtag interface but I think I have the adapter in my box of bits and it's supposed to be fairly straightforward.

Can you work out the other header?

[nretro]

The 2x6 is almost the same as in the usual gotek, only left and right got switched:
J0-J2 right PA0
J3-J6 right VSS
J6 left PIN4 (of floppy connector)
J5 left 37 PC6 (Jumper JA)
J4 left 27 PB1 (Jumper JC)
J3 Left 16 PA2 (Jumper JB)
J2 left PIN10
J1 left PIN12
J0 left PIN16

I ordered a PL2303HX a while back. Is that likely to work or do I need another adapter?

[keirf]

The PL2303HX is a serial adapter and would program via USART1, which is not connected to a header on this Gotek. Similarly, you need access to PA9 to strap it high to program via the USB-DFU method.

So, the only method available here is via the SWD-JTAG interface on the 1x5 header. For this you need an ST-LINK v2 adapter. You can get clones on Ebay for a pound or so from China, or for a little more if you want quicker delivery.

I have one or two stashed somewhere although I've never used them, but I am happy to work out a method on Linux at least, following one of the online guides such as: https://github.com/rogerclarkmelbourne/Arduino_STM32/wiki/Programming-an-STM32F103XXX-with-a-generic-%22ST-Link-V2%22-programmer-from-Linux

[nretro]

Success.

1. Connect GND and +5V to power connector of gotek
2. Connect SWDIO to header pin 46 (PA13)
3. Connect SWCLK to header pin 49 (PA14)
4. Connect header pin 07 (NRST) to GND
5. Execute (I used OpenOCD 0.10.0 on Debian unstable)
   openocd -f /usr/share/openocd/scripts/interface/stlink-v2.cfg -f /usr/share/openocd/scripts/target/stm32f1x_stlink.cfg
6. In a different shell execute "telnet localhost 4444". From now on we use the telnet session
7. type "reset halt" and immediately disconnect header pin 07. The reply should be something like

target halted due to debug-request, current mode: Thread
xPSR: 0x01000000 pc: 0x080061d4 msp: 0x200006c0

8. "stm32f1x unlock 0" was needed on this board to enable writing to the flash bank. ("flash banks" displays all available banks)
9. type
   " flash write_image erase <your_path>/FF_Gotek-v2.12.hex"
10. type "reset run"
11. disconnect st-link.

I still have to test if it actually works. Up to now I only checked that it says "F-F" on the display.

[nretro]

OK... and does work as expected loading an adf on an Amiga. I'll be busy with other things over the next couple of weeks, but at some point I'll also try to add display and rotary mods. I'll post the results.

[keirf]

Will be interesting. Especially the rotary pins may not have a header. Will have to see if there's anything to solder to other than mcu pins directly.

[penfold42]

Worst case, use the JTAG pins or the SPI flash pins

[keirf]

I have linked to this ticket from the wiki: https://github.com/keirf/FlashFloppy/wiki/Gotek-Compatibility#models-to-avoid

I will close this ticket now, but we can add to it with further info in future (eg. if we work on support for rotary encoder).