Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Decompiler help please

92 views
Skip to first unread message

et...@whidbey.com

unread,
Nov 21, 2019, 8:35:22 PM11/21/19
to
I'm pretty bsure this is a real long shot. I have looked online but
have not yet found something that will help. So I ask here.
The EPROM reader/programmer arrived yesterday and it works fine.
The UV EPROM eraser should be here in a day or two. Amazon apologized
for the delay. So I can copy the EPROMs just fine but I would really
like the ladder programs so that I can make some changes to the
machine.
I understand that the decompiled ladder will noit have mnemonics
but maybe I can decipher it anyway.
The control in question is a FANUC 3T.
Thanks,
Eric

Clifford Heath

unread,
Nov 21, 2019, 9:10:01 PM11/21/19
to
How in heck do yu expect help with a decompiler when you haven't told
anyone what type of CPU it is?

Bill Martin

unread,
Nov 21, 2019, 9:33:17 PM11/21/19
to
It's even harder...ladder logic isn't a "cpu machine language" as such,
it's an interpreter running in some native cpu machine language...so
there would be a whole other layer of translation involved in order to
produce anything a human might be able to comprehend. Unless your
computer is made from relays...but why bother.

Fox's Mercantile

unread,
Nov 21, 2019, 9:33:26 PM11/21/19
to
Please try to pay attention.
"The control in question is a FANUC 3T."



--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

Clifford Heath

unread,
Nov 21, 2019, 11:23:03 PM11/21/19
to
That's the packaged controller.
It contains a CPU.
The CPU interprets some ladder codes.

It's highly unlikely that the information required to decode the EPROM
has ever been published. This sort of thing is usually trade secret.

If the CPU is known, and the interpreter ROM is read, there's some
possibility of using a decompiler.

whit3rd

unread,
Nov 22, 2019, 2:59:35 AM11/22/19
to
On Thursday, November 21, 2019 at 8:23:03 PM UTC-8, Clifford Heath wrote:
> On 22/11/19 1:33 pm, Fox's Mercantile wrote:
> > On 11/21/19 8:09 PM, Clifford Heath wrote:
> >> On 22/11/19 12:35 pm, et...@whidbey.com wrote:

> >>>     I understand that the decompiled ladder will noit have mnemonics
> >>> but maybe I can decipher it anyway.
> >>>     The control in question is a FANUC 3T.

> >> How in heck do yu expect help with a decompiler when you haven't told
> >> anyone what type of CPU it is?

> > "The control in question is a FANUC 3T."

A manual found online says the 3T master board uses an 8086, so it'll decompile
as an 8088 would (different bus, but same opcodes).

et...@whidbey.com

unread,
Nov 22, 2019, 12:06:38 PM11/22/19
to
Well, that's why I'm asking for help. I know hardly anything about
this stuff.
Eric

et...@whidbey.com

unread,
Nov 22, 2019, 12:07:48 PM11/22/19
to
On Thu, 21 Nov 2019 23:59:31 -0800 (PST), whit3rd <whi...@gmail.com>
wrote:
Thanks You. Now maybe I can find something to work. This stuff is way
out of my experience but with some guidance I'm hoping to be able to
decipher the ladder.
Eric

Paul Drahn

unread,
Nov 24, 2019, 8:33:08 PM11/24/19
to
Just be aware that sometimes trash is included in EPROMS to stop what
you are trying to do!

Paul

whit3rd

unread,
Nov 26, 2019, 2:35:11 PM11/26/19
to
On Sunday, November 24, 2019 at 5:33:08 PM UTC-8, Paul Drahn wrote:

> >>>>> On 22/11/19 12:35 pm, et...@whidbey.com wrote:
> >>
> >>>>>>     I understand that the decompiled ladder will noit have mnemonics
> >>>>>> but maybe I can decipher it anyway.


> Just be aware that sometimes trash is included in EPROMS to stop what
> you are trying to do!

Three things to look out for:
the ordering of the data bits can be scrambled; probably you can trace those to the CPU pins
the ordering of the address bits can be scrambled, and some may be
hard to trace to the CPU pins
the data and/or the address might go through inverting transceivers, which changes
the bit-is-a-one /bit-is-a-zero identity.

Tim R

unread,
Nov 27, 2019, 9:46:44 AM11/27/19
to
On Thursday, November 21, 2019 at 8:35:22 PM UTC-5, et...@whidbey.com wrote:
> The EPROM reader/programmer arrived yesterday and it works fine.
> The UV EPROM eraser should be here in a day or two. Amazon apologized
> for the delay. So I can copy the EPROMs just fine but I would really
> like the ladder programs so that I can make some changes to the
> machine.

I'm sure you know this but I've been bitten once and it may be worth mentioning.

I understand you're trying to back up your EPROM holding your program. We had a similar problem in a manufacturing plant, probably with an Allen Bradley PLC or similar. (it's been a while) The program went bad, we got the backup EPROM, and uploaded the saved program.

Except we didn't. The wording of upload and download was ambiguous and we accidentally saved the bad program from the machine onto the EPROM. We had one extra EPROM in the safe, got that out and ………… did exactly the same thing.

Everything was ladder logic and I came to hate those hardcopy printouts, but with enough time and patience we could type it back in. Ladder logic was theoretically understandable to the electricians but it violates every single concept in programming. Every line can refer to any other line - no modular programming, no single entry single exit, etc.

three_jeeps

unread,
Dec 4, 2019, 1:13:59 PM12/4/19
to
In this case, the 'river to the ppl' took a wrong turn.
with all due respect, FANUC 3T tells you almost nothing wrt what the OP is trying to do. One needs to open the box and find out what CPU is on the main board e.g. intel iX, ARM X, Atmel, etc. etc.
Knowing the CPU defines the ISA of the machine, and one can begin to figure out how to set up a logic analyzer to look at the machine instructions and execution. Configuring a small, simple ladder logic example or 5 and observing the instruction and address trace will provide some insight into what is being R/W to memory, EPROMS, etc. which can then be used to identify and decode the ladder login mnemonics used by the mfg.
From what I know, the approach taken to codify and execute ladder logic mnemonics is unique to the mfg. There are some general schemes/patterns used to execute the ladder program but again, every mfg has their own way of doing things.
If you really need to do things at this level, it is a LOT of work and a fair amount of guessing.
There have been a number of academic papers looking at various issues wrt Ladder Logic in general and in some cases, manufacturer specific details are sometimes contained in the paper. Perhaps googling may help you.
A JTAG connection and associated diagnostic software may be helpful as well.
Good luck
J


three_jeeps

unread,
Dec 4, 2019, 1:19:07 PM12/4/19
to
I just saw the reference to the 8086 processor. the good news is that there is lots of knowledge about this processor, including how to use a logic analyizer with it. I believe that looking at a static ROM dump of contents may not provide you with the insight for what you want to do. Looking at execution of the system would be more beneficial. But again, a LOT of work
J
0 new messages