Finished PSU replacement on my X68k Expert, what next?

[eidis]

 Hi Famiac !

Things are getting complicated. Try powering on your X68000 for about five minutes with some game running and then try touching the top of the chips and see if any of them are abnormally hot.

Keep the scene alive !
Eidis

[famiac]

Just tried. Nothing was really hot at all.

Edit: that narrows it down i think. Pretty sure there's a bad solder joint on the 74als05, since i replaced that IC. So i'm going to reflow and check for continuity next weekend then i'll update you guys.

[famiac]

Could one of the gates be bad? I don't see anything wrong here. I also can't find all the test points.

anybody know where i can test these pins?
http://i.imgur.com/X5okXEj.jpg

[lydux]

Could one of the gates be bad? I don't see anything wrong here. I also can't find all the test points.

Humm... There are not many gates in video parts, mostly ASICs and RAM. However, it reminds me that the GVRAM pixels output is serially controlled by first the CRTC, then fed into a TTL gate combined with a video clock signal. The resulting signal control the GVRAM pixels ouput per scanline... If you have only graphics garbages (I mean human68k text work correctly), it might be the culprit.

I don't have EXPERT motherboard on hand, only a SUPER one but I guess it's similar. Can you see a 74AS02 between the big VRAM row and CYNTHIA chip ? (noted IC1 on mine) I would suggest you to change it.

anybody know where i can test these pins?
http://i.imgur.com/X5okXEj.jpg

I can't see it on my schematics yet, but if I remember right, this ALS05 is related to the invertion of SYNC outputs. This is the high-end part of the video generation. Your problems look like before.

[famiac]

Sure thing, thanks lydux. I thought it might be the sync inverter because in bomberman the graphics are correct but they scroll

Ordered the 74as02

[famiac]

Replaced 74as02, still having the same problem :/

[lydux]

Yes sorry, I was totally wrong... I though Bomberman title background was located inside GVRAM but I checked with XM6, and it's inside the PCG RAM (controlled by CYNTHIA).
This ALS02 is only used for GVRAM rendering.

So, TVRAM and GVRAM rendering parts are ok, as well as color palette lookup.

This leave the CYNTHIA part :



All parallel bus signals for fetching or writing PCG RAM data from cpu, or during rendering times can be excluded, as well as the image capture unit inputs. According to your memtest and screenshots, these signals look likes all goods.
Remaining signals are clocks :



I can't beleive of an HSYNC or VSYNC signals issue anymore. These are shared lines between all video chips and would create others troubles than just with PCG.
The only remaining ones (QEC, QDC, QCC, QBC and QAC) are coming from the OSCIAN video clock generator :



Unfortunatly, I still don't know many much about it. It deals with the 69MHz and 38MHz oscillators, and produce various clock signals according to some configuration inputs like high or low horizontal frequency resolution (15KHz vs 31KHz).
Actually, I can only advise you to check for continuity on all fuse blocks (FB29 --> FB37, might be noted diffently on your board but located near OSCIAN), as well as between the 5 OSCIAN and CYNTHIA clocks lines.

Of course, OSCIAN inputs can be involved (HRES0, HRES1, FH/L, MIX and HRL)...


Few questions :
- It unclears on your screenshots, but it looks like all sprites are doubled to different horizontal positions and are a bit translucent, right ?
- As you have an original monitor, can you try with a 15KHz game ?

[famiac]

Thanks lydux, i'll take a look in a bit. I will test with a 15khz game first and tell you what happens

[famiac]

Tested in 15khz; same problem happens

https://docs.google.com/file/d/0B14RXpId-XGTLTFPUlFRTjdxRFk/edit?usp=docslist_api

Checked all fuses. All OK for continuity

Connections between OSCIAN and CYNTHIA are checked and OK for continuity

[lydux]

Ok, thanks for posting a video ! I can better see the problem. I were not able to understand the Argos no Senshi garbages, it's clear now !

I tried to load Puyo Puyo in XM6, stopping emulation during attract screen and give a look into the PCG RAM  content :



This is exactly garbages that appears on your screen !
That means CYNTHIA keeps fetching the whole PCG RAM during rendering instead of lookup against it according to the sprite buffer content.
But it also does the right stuff as I can see the brown joystick and the blue button appearing in your video at the correct horizontal position. This excess of work might explain the scrolling.


The question now is why does it do this ?...
Investigating...

[famiac]

Lydux, i appreciate your work so very much. If you collect games i'd like to send you a gift for your time.

[famiac]

Alright guys, while i patiently wait for some word from lydux, i'm thinking maybe we can fix the other motherboard i have sitting around. It is fully recapped, however when i power the system on, i can no longer turn it off with the soft power switch, the floppy drives accept disks and eject them on demand but do not read them, afaik there is no video out, and when i try to turn off the console with the power switch on the front, only the green led turns off/on depending on the power switch position, the fan keeps spinning.

[eidis]

 Hi Famiac !

Please have a look at the following article:

Soft-Power troubleshooting and repair
http://gamesx.com/wiki/doku.php?id=x68000:soft_power_switch_troubleshooting_and_repair

Keep the scene alive !
Eidis

[famiac]

I don't think there's anything mentioned about no video out... The x68k itself doesn't boot. How do i test those power signals? Multimeter?

[lydux]

Hi Famiac,

Sorry the long delay, I still need some free times to look into your CYNTHIA problem (next week should be good). As I said, it's a complex part in which I don't get everything. I need to understand some signals first. At least, we now have a good idea where the problem is.

It is fully recapped, however when i power the system on, i can no longer turn it off with the soft power switch, the floppy drives accept disks and eject them on demand but do not read them, afaik there is no video out, and when i try to turn off the console with the power switch on the front, only the green led turns off/on depending on the power switch position, the fan keeps spinning.

However, I might help you on with this one without involving testing anything on my side. At least, 3 of my X68000 failed with this exact same situation (all were recovered). These symptoms implied a general core part failure on early boot where the processor stop for some reasons, generally because of a device that does not answer correctly during its initialization.
Also, I found some months ago that ACE models suffers from a bad design in the interrupts processing part, which suddenly put the whole machine in such state. So, which models is this board ?

First, make sure there is no broken pcb traces on I/O board. Battery leaks often corrode and brake traces, turning the x68k into this state.

If all are ok, you'll have next to analyze the processor status ; in power-on condition.
Here is the CPU pinouts :

Code Select Expand


           HD68000
       +------\/------+
    D4 |1  .        64| D5
    D3 |2           63| D6
    D2 |3           62| D7
    D1 |4           61| D8
    D0 |5           60| D9
   /AS |6           59| D10
  /UDS |7           58| D11
  /LDS |8           57| D12
  R//W |9           56| D13
/DTACK |10          55| D14
   /BG |11          54| D15
/BGACK |12          53| GND
   /BR |13          52| A23
   VCC |14          51| A22
   CLK |15          50| A21
   GND |16          49| VCC
/HALT |17          48| A20
  /RST |18          47| A19
  /VMA |19          46| A18
     E |20          45| A17
  /VPA |21          44| A16
/BERR |22          43| A15
/IPL2 |23          42| A14
/IPL1 |24          41| A13
/IPL0 |25          40| A12
   FC2 |26          39| A11
   FC1 |27          38| A10
   FC0 |28          37| A9
    A1 |29          36| A8
    A2 |30          35| A7
    A3 |31          34| A6
    A4 |32          33| A5
       +--------------+


With a voltmeter, put the black lead on any metal ground part. And with the red one, give me voltage measurement of the following pins :

General :
  * CLK (pin 15)
  * /RST (pin 18)
Processor condition :
  * /BERR (pin 22)
  * /HALT (pin 17)
Bus arbitration :
  * /BR (pin 13)
  * /BGACK (pin 12)
  * /BG (pin 11)
Interrupt processing :
  * /VPA (pin 21)
  * /IPL2 (pin 23)
  * /IPL1 (pin 24)
  * /IPL0 (pin 25)

The next step depends of the result.

[famiac]

The PCB is an Expert model. My other X68000 boots with the IO board so i assume if i use the same one there are no IO board problems right? The daughterboard is fully recapped and should be working.

[famiac]

Wow i feel awful right now. I reassembled it with the other motherboard, (i put a gold pin socket on top in the cpu socket to fit xellent30s) and it booted and played graphics properly!

Looks like the socket was bad. But guess what? I removed the cputo fit in xellent30s but forgot to unplug x68000!!! Now i think i ruined my xellent30s and my pc because it won't boot it stays stuck at the red light no matter what i do

Edit: it boots with cpu now, but not with xellent30s. The socket is very finicky. attempting to boot with xellent30s unit gives me the original problem (no video/soft power)

[eidis]

 Hi Famiac !

Check if there is a fuse close to the xellent30s socket.

Keep the scene alive !
Eidis

[famiac]

Thanks for the tip eidis. I have exams now so i won't be able to check until the weekend.

[famiac]

Ok there seems to be a problem with the xellent30s. Eidis, i can't find the fuse you're talking about.

It's the problem where the led goes green but the x68k doesn't boot or turn off

Edit: some more lovely news: one of the pins on the pin header broke, and i can't find a replacement part.

I don't have the time or knowledge to repair this xellent30s unit, could i send it to someone for repair? I'm about to break down after so much time and money spent

[BlueBMW]

Thats a real bummer .  Can you take a picture of the damage?

[famiac]

Just one pin on the pin header broke, and the rest of the damage is either non-existant or invisible

EDIT: i found a replacement header. Can i sent it along with the xellent30s to someone for repair?

Here are some voltages for you lydux

Tested voltages :
Xellent connect to cpu Socket

  * CLK (pin 15) 1.625v
  * /RST (pin 18) 4.85v
Processor condition :
  * /BERR (pin 22)   4.85v
  * /HALT (pin 17) 100mv
Bus arbitration :
  * /BR (pin 13) 4.85
  * /BGACK (pin 12) 4.85
  * /BG (pin 11) 4.2v
Interrupt processing :
  * /VPA (pin 21) 4.85v
  * /IPL2 (pin 23) 4.85v
  * /IPL1 (pin 24) 4.85v
  * /IPL0 (pin 25)4.85v

Socketed hd68hc on xellent30s

General :
  * CLK (pin 15) 1.7v
  * /RST (pin 18) 4.85v
Processor condition :
  * /BERR (pin 22) 4.85v
  * /HALT (pin 17) 78.8mv
Bus arbitration :
  * /BR (pin 13) 4.85v
  * /BGACK (pin 12) 4.23v
  * /BG (pin 11) 4.85
Interrupt processing :
  * /VPA (pin 21) 4.85v
  * /IPL2 (pin 23) 4.85v
  * /IPL1 (pin 24) 4.85v
  * /IPL0 (pin 25) 4.85v