Dead Falcon resuscitation

Troubles with your machine? Just want to speak about the latest improvements? This is the place!

Moderators: Mug UK, Zorro 2, spiny, Greenious, Moderator Team

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 3:54 pm

So I am currently trying to bring another Falcon board back from dead. I have assembled all missing ICs and got "unsupported" on LCD on power on. So far I have identified three issues.
The first one was pretty easy to fix - two data lines were 0 ohms against VCC line, which I have traced down to faulty resistor array (P7 I guess). After I have replaced it, this is now fixed. Honestly, this was first resistor network I have ever seen fail like this, except of course those mechanically damaged.
Second issue was cold solder joint on a ferrite bead, resulting in dead clock and CPU getting hot very quickly. Resolved easily.
The other issue I see is a bit of a puzzle to me. It appears that impedance between VCC and GND is unexpectedly low - just 35.6 ohms. I would expect few hundreds of ohms, maybe less, but certainly not this "next to nothing". Its not short connection either, I mean my PSU handles this load pretty well, I am reading nice 5.1 volts all across the board when powered.
I wonder what could cause such an odd impedance on power rail. Its certainly not any of the e-lytes, they are now all changed and I have been reading same odd 35.6 ohms value even when the board was stripped off them all. It cant be cracked ceramics either, since they either go short, or open, nothing in between.
I thought someone has been messing with the board and changed maybe one of the SMD decoupling caps for resistor by mistake, but it does not seem to be the case either.
Anyone seen anything like this before?
Last edited by mpattonm on Wed Sep 05, 2018 4:09 pm, edited 1 time in total.

User avatar
troed
Atari God
Atari God
Posts: 1432
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Dead Falcon resuscitation

Postby troed » Wed Sep 05, 2018 4:02 pm

Why do you think that's odd? IIRC a regular ST has 50 Ohm or so.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 4:03 pm

For real? Hm...
Edit: I have pulled out Mega 1 and it reads just below 600 ohms, once caps charge.

User avatar
troed
Atari God
Atari God
Posts: 1432
Joined: Mon Apr 30, 2012 6:20 pm
Location: Sweden

Re: Dead Falcon resuscitation

Postby troed » Wed Sep 05, 2018 4:12 pm

I usually measure resistance powered off though. What did you have before the caps charged?

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 4:18 pm

Of course it is on un-powered unit. But caps still "draw" from multimeter affecting the readout, it takes a while before the value on true rms mm settles. It goes from 100 up to final 600 ohms on Mega, while it just locks right up to those 35.6 on Falcon - like the caps were not even there.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 4:53 pm

Anyway, I have pulled the probes out, this is what I see on CPU pin 6 (CLKIN). I guess that is why the bird seems dead as doornail.
You do not have the required permissions to view the files attached to this post.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 5:21 pm

Got nice 8MHz clock on CPU now. BERR high, halt HIGH, reset circut all right. Still no video. I am hooked up :)

User avatar
Arne
Captain Atari
Captain Atari
Posts: 461
Joined: Thu Nov 01, 2007 10:01 am

Re: Dead Falcon resuscitation

Postby Arne » Wed Sep 05, 2018 5:41 pm

Now it would be the right time IMHO to connect a LA to the data/address lines on the internal expansion port and see if the CPU accesses ROM.
Image

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 7:35 pm

Well this was quick test. DATA12, DATA13 and DATA14 are in short against each other too? WTF has happened to this board in the first place?!
Last edited by mpattonm on Wed Sep 05, 2018 8:33 pm, edited 1 time in total.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 8:32 pm

Some background: when I first got this board, its half way empty, quite a lot of ICs missing: CPU, FPU, DMA, TOS, both ACIA chips, all 74244s around joypad ports and finally all 5 GALs.
Now, all chips were taken off clean, w/o brutal force. Just one pad under DMA was missing and one pad under CPU loose. Obviously, someone was trying to save it, rather then salvage its chips. I somewhat anticipated somebody just used this board as source for components, but it makes a more sence now - all these chips are on databus.
So guys - whos' work is it? :)

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 10:02 pm

The short is cleared. I see a non-stop strobing on address lines, but only second or so on data lines. Still no video :O

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Wed Sep 05, 2018 10:15 pm

Funny, I do not see same line strobing on Videl address line pins. Nothing is obviously fed to DAC, of course no video.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Thu Sep 06, 2018 1:36 pm

Another intermittent issue in CLK32 generator circuit. Now the CLK32 signal is stable, but seems odd to me. Can anyone please tell me if this is how COMBEL/VIDEL master clock (CLK32) in PAL system should look like?

Looking at the schematics I am inlined to say yes to my own question, but then, NTSC system has CLK32 generated by crystal oscilator and that has very different output levels.
You do not have the required permissions to view the files attached to this post.

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Fri Sep 07, 2018 5:25 pm

I am clueless.
I have inspected all components I could they all seem fine now. I have changed failed data line pullup resistor array and 32.084988M crystal resonator (as it was drifting a lot! between restarts) for generic 32.000M for now, since I do not have PAL version on hand. Anyway it should be close enough to wake the system up, maybe except for the video.
All clocks are sane. BERR, HALT ok, RESET OK. All high.
After power on or reset, all data lines and address lines strobe for a second, ROM4 signal strobes and then everything go silent. No bus activity.
CPU clock remains down at 8MHz. Same happens with and without Falcon test cartridge on and two different memory modules, one stock 4MB, one new 16M SIMM version, verified OK.
The Falc definitly does not read from cart ROM, judging from ROM4 activity LED on a cart.
CPU is new, and I have probed all the got damn pins, all soldered properly. Same goes for SDMA.
FPU is not assembled.
RTC chip replaced, new one is on the socket, the NVRAM should be cleared by grounding its pin 12 but I have no chance to verify.
ACIA chips new as well, not that it would be important at this point.
All five GALs are verified and known to work in another Falcon. I have bought an original ROM for insane sum of money to rule out EPROM incompatibility. Also I have downloaded its contents (its plain Hitachi HC27C4096 by the way) and verified against TOS image file, all fine.
Somebody please help.

User avatar
frank.lukas
Hardware Guru
Hardware Guru
Posts: 1638
Joined: Tue Jan 29, 2008 5:33 pm
Location: Germany

Re: Dead Falcon resuscitation

Postby frank.lukas » Fri Sep 07, 2018 5:41 pm

Have you a Atari Falcon Test cartridge for the Romport ?
fancy Atari Musik anDA Dance "Agare Hinu Harukana" 1998 ATARI http://www.youtube.com/watch?v=JX10fxb5eYE

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Fri Sep 07, 2018 5:55 pm

Yes, the problem is this Falc does not boot from it. Same beavior with or without cartridge in the port.

User avatar
frank.lukas
Hardware Guru
Hardware Guru
Posts: 1638
Joined: Tue Jan 29, 2008 5:33 pm
Location: Germany

Re: Dead Falcon resuscitation

Postby frank.lukas » Fri Sep 07, 2018 5:57 pm

What about with the Test Cartridge and a Null Modem cable and a serial console (9600Baud8N1)on a PC or Mac ?
fancy Atari Musik anDA Dance "Agare Hinu Harukana" 1998 ATARI http://www.youtube.com/watch?v=JX10fxb5eYE

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Fri Sep 07, 2018 6:02 pm

When the bus goes silent, all data lines and address lines remain up, so it seem to me notning holds the bus.
One more observation: if I pull TOS ROM out, the unit does not even strobe ROM4 and does not acess the cartridge at all :o

mpattonm
Hardware Guru
Hardware Guru
Posts: 331
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: Dead Falcon resuscitation

Postby mpattonm » Fri Sep 07, 2018 6:04 pm

frank.lukas wrote:What about with the Test Cartridge and a Null Modem cable and a serial console (9600Baud8N1)on a PC or Mac ?

I do not have the USB to serial cable on hand. And while I will get it on Monday, I do not think I will see anything on serial port. See I am quite sure it does not read a code from cart ROM. At all.

User avatar
frank.lukas
Hardware Guru
Hardware Guru
Posts: 1638
Joined: Tue Jan 29, 2008 5:33 pm
Location: Germany

Re: Dead Falcon resuscitation

Postby frank.lukas » Fri Sep 07, 2018 6:20 pm

I think when the CPU are running and a little bit of RAM are there the test cartridge give something out over the serial port.
fancy Atari Musik anDA Dance "Agare Hinu Harukana" 1998 ATARI http://www.youtube.com/watch?v=JX10fxb5eYE

User avatar
Gaiyan
Hardware Guru
Hardware Guru
Posts: 337
Joined: Tue Jun 29, 2004 3:39 pm
Contact:

Re: Dead Falcon resuscitation

Postby Gaiyan » Fri Sep 07, 2018 6:25 pm

frank.lukas wrote:What about with the Test Cartridge and a Null Modem cable and a serial console (9600Baud8N1)on a PC or Mac ?


I actually have an ST for this. I've created a cart that contains an executable ram disk which, after warm reset, boots to the desktop with several utilities. One is a VT52 emulator that I use all the time to diagnose dead boards. It's pretty cool.
Image

User avatar
dhedberg
Atari Super Hero
Atari Super Hero
Posts: 820
Joined: Mon Aug 30, 2010 8:36 am
Contact:

Re: Dead Falcon resuscitation

Postby dhedberg » Fri Sep 07, 2018 7:55 pm

Gaiyan wrote:
frank.lukas wrote:What about with the Test Cartridge and a Null Modem cable and a serial console (9600Baud8N1)on a PC or Mac ?

I actually have an ST for this. I've created a cart that contains an executable ram disk which, after warm reset, boots to the desktop with several utilities. One is a VT52 emulator that I use all the time to diagnose dead boards. It's pretty cool.

Ah, sounds useful! Would you consider sharing an image of this disk?
Daniel, New Beat - http://newbeat.atari.org. Like demos? Have a look at our new Falcon030 demo MORE.

User avatar
frank.lukas
Hardware Guru
Hardware Guru
Posts: 1638
Joined: Tue Jan 29, 2008 5:33 pm
Location: Germany

Re: Dead Falcon resuscitation

Postby frank.lukas » Sat Sep 08, 2018 6:06 am

You can use also a normal terminal program like Rufus or Connect97 and so on ...
fancy Atari Musik anDA Dance "Agare Hinu Harukana" 1998 ATARI http://www.youtube.com/watch?v=JX10fxb5eYE

User avatar
dhedberg
Atari Super Hero
Atari Super Hero
Posts: 820
Joined: Mon Aug 30, 2010 8:36 am
Contact:

Re: Dead Falcon resuscitation

Postby dhedberg » Sat Sep 08, 2018 10:03 am

Ah, OK. I'll try and set something up just to see what you get out from it. Never tried it.
Daniel, New Beat - http://newbeat.atari.org. Like demos? Have a look at our new Falcon030 demo MORE.

mikro
Hardware Guru
Hardware Guru
Posts: 1780
Joined: Sat Sep 10, 2005 11:11 am
Location: Kosice, Slovakia
Contact:

Re: Dead Falcon resuscitation

Postby mikro » Sat Sep 08, 2018 11:21 am

You can even boot the ST/E language disk on a real ST/E where this Terminal VT52 ACC is present (strangely, it's missing on the Falcon).


Social Media

     

Return to “Hardware”

Who is online

Users browsing this forum: lp and 14 guests