Tracing old 68K code

AFAIK a logic analyzer will be passive, but if you have enough inputs to fully monitor the bus and if you have fancy enough software it'll do disassembly on the fly. For than era of micro, doing that sort of task, it may be better than an emulator.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

What is the device exactly? Perhaps a keyboard for example...

As to your original question, many good logic analizers will read the op codes/data along with the CPU, and display disassembled code (with data) on the screen. You can set them to trigger at the point there is a problem (assuming you know when the problem happens).

Why do you think the software has failed? Seems to me that it is a hardware problem?

Which CPU?

John

68K Macintoshes contained a ROM micro-debugger that processed simple commands through a text window or a serial port. Developers could point the NMI handler to it when MacsBug wasn't suitable. I think the micro-debugger was part of a Motorola development kit so there's a chance that this computer has it in the ROM too.

Google considers all of this to be obsolete information that shouldn't be returned in searches. Good luck.

--
I won't see Google Groups replies because I must filter them as spam

Try something "weird" like (linear) mapping (of) RAM to a video display..

68000, fortunately DIP64 which allows easy probing since it's not socketed.

--
Thanks,
Fred.

Not at liberty to tell.

No, it's the HW that fails. But not understanding all the process details from the schematics (a pair of undocumented FPGAs are involved), I need to have some way to BKP and step execute the SW.

The POST returns some cryptic error message, but then the systems seems to work accurately and within limits... Anyway this is a critical system and the customer at least needs to understand what's happening, and preferably have it fixed, which should be easy when understanding the fault.

--
Thanks,
Fred.

???

--
Thanks,
Fred.

Too bad, as likely there is someone here who may have experience. For example, I've some experience with Kurzweil keyboards (powered by a

68K) and unless this is a custom one-off device I'm sure you will be better off if you could tell...

What's the message?

The 16900A is a pretty current LA and can do lots and lots of fancy things. (discontinued last year or early 2010)

I don't want to buy the pattern generator board which would have allowed lots of possibilities, but having poked a bit into the LA documentation there seems to be a good solution:

The LA have a trigger output that's generated on the fly according to the fancy conditions you setup. It also have one (slow) 8 bit port for target control, like driving a reset, NMI,... The LA can also be totally controlled through VBA (yeah it's an XP box).

I guess that all that, combined with a small PAL/CPLD connected to the uP board and with some minimalist track butchering can solve my pb quite easily and with reasonable handling.

--
Thanks,
Fred.

es.

Greetings... I must admit I know nothing about your system... However, I was trained in electronics during the time when they taught us to how to fix things at what we used to call the component level.. If the unit is somewhat old.. more than 10 years old, consider resoldering the connections around the power supply... and even where you have connectors to long cable lines.. You could have a bad soldering connection... this can happen if the equipment is somewhat old... Larry ve3fxq

I'd verify the software is still OK (correct eprom contents). Next you could verify the hardware (memory and peripherals) with a small test program. If the digital section is OK then the problem must be somewhere in the analog section. With old hardware you can expect anything.

I doubt a logic analyzer will do much good until you can pin-point an error and set a trigger condition.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

Pickup an old Fluke 9000A and a 68K pod? I'd think you'd be better off to get an old Nohau emulator.

Cheers

AKA as the video card trick.

Another technique is a pair of DAC one for address and one for data to drive a scope in the XY mode. Does anybody know of a DSO that does XY mode?

Doesn't every DSO do that? I never came across a DSO that can't do XY mode.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

Lemme see if i can get my TDS544A to do it. Maybe i really need a nice Instek or Rigol?

I have a TDS510A and it can do XY for sure. According to the TDS510 manual you must press Display -> format (bottom menu) and select XY mode from the side menu.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

display..

So i went and downloaded the users manual, dug through it a bit and it is right where you said, i checked it on my scope and ther it is. Thanks, no need to buy a portable for that infrequently used capability. I do have reasons to buy a portable though.