DenYoNet - Data corruption on FTP transfers

Page 2/2
1 |

By Louthrax

Prophet (2076)

Louthrax's picture

23-06-2019, 13:19

So I had a reply from Yobi, who told me that the problem is that the W5100 is designed for 3.3v, and only 5v tolerent, and that this can cause data corruption. Darn ! I'll do some more tests (Yobi suggested that it might work on some MSX machines, but not on the Zemmix).

Wondering if it would be possible to have a hardware fix for that (something like that: ).

Other thing I observed in the denyotcp.asm source code is that it does not wait for command completion before initiating data transfer from the Wiz to the MSX. But I do not believe too much in this given the slow Z80 speed :)

By ducasp

Master (147)

ducasp's picture

23-06-2019, 13:33

https://forum.arduino.cc/index.php?topic=323384.30

It seems Eugeny faced this issue as well, perhaps he might help figuring out a good solution for it.... :P

By Louthrax

Prophet (2076)

Louthrax's picture

23-06-2019, 14:07

ducasp wrote:

https://forum.arduino.cc/index.php?topic=323384.30

It seems Eugeny faced this issue as well, perhaps he might help figuring out a good solution for it.... :P

Thanks, that's interesting. Might not be so simple... I guess all the lines (data & adresses) should be stepped down.

By ducasp

Master (147)

ducasp's picture

23-06-2019, 18:47

https://learn.sparkfun.com/tutorials/bi-directional-logic-le...

The major issue I see is that you probably would need to socket the w5100 in order to do that, and perhaps it wouldn't fit the cartridge case.... Or you could add the mosfets and cut pcb traces... Lots of work.

By Louthrax

Prophet (2076)

Louthrax's picture

23-06-2019, 21:23

Yeah, I think I'd probaly do that just after the cartridge pins, for all addresses and data lines. That would require something like 5 or 6 or these little chips wired and lots of cuts... Not sure I'll be brave enough to do that Smile

Page 2/2
1 |