Author Topic: Interpreting uPD765 FDC result bytes  (Read 10743 times)

0 Members and 1 Guest are viewing this topic.

Offline JonB

  • 6128 Plus
  • ******
  • Posts: 741
  • Country: gb
  • Liked: 203
  • Likes Given: 46
Re: Interpreting uPD765 FDC result bytes
« Reply #125 on: 13:27, 31 May 15 »
Interesting. I noticed that WCK isn't getting a signal, but the datasheet on CPCWiki says:

Quote
WCLK (Write Clock)
The WCLK input sets the data write rate to the FDD. It is
500 kHz for FM, 1 MHz for MFM drives, for 8 MHz operation
of the FDC; 250kHz FM or 500 kHz MFM for 4 MHz
FDC operation.
This signal must be input for read and write cycles
WCLK’s rising edge must be synchronized with CLK’s
rising edge, except for the uPD765B.

So this sheet is implying I need a valid WCK signal for read operations. The fault here is with IC7486 which divides the system clock down to what I assume is 250khz and ensures it is aligned.

 [ Invalid Attachment ]

Where it says 500khz, that is as measured, and it's a fairly clean square wave (so the right hand flipflop and intervening gates seem to be working). But at the 250Khz point, there is something unexpected. It is a LOW signal (0v with a bit of noise). Pins 10 and 12 are HIGH as expected; and the 2Mhz signal at pin 13 and 1Mhz signal at pin 3 are both present and correct (in phase), so I think the second flipflop (on the left) has failed. The part is on order...
« Last Edit: 13:39, 31 May 15 by JonB »

Offline JonB

  • 6128 Plus
  • ******
  • Posts: 741
  • Country: gb
  • Liked: 203
  • Likes Given: 46
Re: Interpreting uPD765 FDC result bytes
« Reply #126 on: 13:40, 31 May 15 »
And here is some scope porn...


Yellow is the 2Mhz trace.
Cyan is the 1Mhz trace.
Magenta is the 500Khz trace (5v so it's a bit flatter than the others which are @2v).



Offline JonB

  • 6128 Plus
  • ******
  • Posts: 741
  • Country: gb
  • Liked: 203
  • Likes Given: 46
Re: Interpreting uPD765 FDC result bytes
« Reply #127 on: 14:11, 02 June 15 »
Hmm, I think I got that wrong. The expected trace at pin 9 of the 74LS74 is a 500khz pulse with width of ~200ns, like this:


 [ Invalid Attachment ]


..which is what I got when I replaced the part. Serves me right for not reading the data sheet fully. Oh well.


Anyway, having replaced the chip I reconnected the PC and started Hyperterm. Lo and behold...


Code: [Select]
PHILIPS P2000C     63K CP/M 2.2


KEYBOARD UK/NL, VIDEO UK/NL, PRINTER EMPTY


A:640K-FL1 B:640K-FL2


A>dir
A: CPM61    COM : CPM62    COM : CPM63    COM : CBIOS61  COM
A: CBIOS62  COM : CBIOS63  COM : ABASIC   COM : CHESS    COM
A: CONV     COM : ZUID1        : ZUID2        : ZUID3
A: ZUID4        : allard   BAS : STAT     COM : UTIL     COM
A: WSMSGS   OVR : WSOVLY1  OVR : WS       COM : WS       INS
A: UTIL     MSG : CPYDSK   COM


YES!!!

Offline Bryce

  • The Hardware Guy.
  • Supporter
  • 6128 Plus
  • *
  • Posts: 11.033
  • Country: wf
  • It's not broken, it just hasn't been fixed yet.
    • index.php?action=treasury
  • Liked: 3906
  • Likes Given: 405
Re: Interpreting uPD765 FDC result bytes
« Reply #128 on: 15:27, 02 June 15 »
Congratulations, and as a bonus you got to learn a bit about your new scope too :)

Bryce.

Offline JonB

  • 6128 Plus
  • ******
  • Posts: 741
  • Country: gb
  • Liked: 203
  • Likes Given: 46
Re: Interpreting uPD765 FDC result bytes
« Reply #129 on: 19:58, 02 June 15 »
Now just the terminal board to fix.  ;D


Thanks for your help & advice!