JP86SS

iTrader: (1)

I've been searching the internet and coming up with a few neat things. A guy named Perry Fischer (sp?) had released a couple to public domain with notes of having the ability to watch (and graph) registers during step or run execution. "sim68" was the program name. It was made around 92-94' DOS program.
The ability to step through the source code as it is executed is really easier to follow when trying to learn the steps.
The registers can be viewed as they are changed. Just neato in my book. I haven't located any other newer versions, the e-mail came back as a bad link.
Anybody have a favorite 68HC11 simulator for PC ?

MTPFI-MAF

I have one I downloaded Some where but the link dosen't work any more So I was just going to E-mail you the Program to Use it as you see fit, not sure if it is the same one you have but the name is THRSim11, The problem is that it is 16Meg and I can't e-mail it. The post for it is on TGO somewhere,.

EDIT:
here is the Link to the post https://www.thirdgen.org/techbb2/sho...hc11+Simulator

and here is the link to the site THRSim11 Page it is on http://www.hc11.demon.nl/thrsim11/info.htm

JP86SS

iTrader: (1)

I couldn't use that link(dead) but a quick search found it.
Looks pretty neat. I've been playing with it during lunch.
The add on stuff looks interesting. I have to figure out how to make it sequence properly.
It may be because its the demo version with limited ram simulation.
Looks to be worth the 50 bux anyway. I may give that a try if I don't find anything better.
I like the idea of a software bench rather than hardware (at this point anyway) The hardware bench will take me too long to do what I want. I had dreams of using analog boards to simulate the real world signals but this looks easier to manipulate and I can see the registers.
Neat.

JP86SS

iTrader: (1)

Just some FYI,
I did purchase this package and am quite happy with the actual working version. The memory won't run our programs until you buy the license, will only do small chunks of code.
You can setup signals to feed the inputs and have scopes to see them or any other location with custom windows, etc.
The program does everything except dis-assemble. There is an editor and compiler built in that works well. Doesn't like the ".byte" lables on data so I'm commenting and changing everything in my AUJP as I go.
I'll keep everyone up on this when I get a full simulation up and running as a software ECM bench.

JPrevost

cool stuff. I actually got a copy of that software from my friend who is an EE. They still teach the 68hc11 at OSU for low level stuff, I thought that was interesting. Too bad they never have time to deal with ALL this code. They just do stupid little analog/digital stuff and a few outputs. Nothing fancy like GM did.

89350STRO

Try this one, I think you'll like it much better than sim68. It was written by some students at the Milwaukee School of Engineering. Make sure you use version 1.67 it will step through the code, I haven't been able to get version 1.68 to step through the code. Here is the link. It is called WOOKIE!! It is an acronym for Wireless Object Oriented Kindly Interfaced Emulator.

http://www.msoe.edu/eecs/ce/ceb/resources/

89350STRO

JP86SS

iTrader: (1)

I actually tried WOOKIE but could get it to work with my limited attention span. After a few days at it I went back to SIM68. Sim was a little crude but was letting me step through the code.
When I first tried the THRSim, I was impressed with the add ons for signal generation that would let me setup VSS at whatever frequency and watch memory values with the scope indicator. I was hopeful that I could setup all of the inputs to simulate the operation of the ECM.
I broke down and spent the $50 to get the full version and it lets me setup the memory to run the full code. It steps or runs it fully with all of the ports configurable on the fly for changing the inputs. (up to 100K Hz sine, tringle, sawtooth, no square wave though) I haven't gotten to the serial port stuff yet but it also has a watch window for displaying the output of the data. The real problem with the demo version is it really doesn't let you use enough memory for our application.

Thanks for the link although I'm kinda tied to the THRSim11 at the moment anyway. (remember I have a short attn span )
Look, a bird!

Let me know if you have much success with Wookie. I'll let you know how it goes with this program too.
Jp

junkcltr

iTrader: (1)

I thought about giving this a try some time ago, but decided it wasn't worth the money.......without ever trying out the demo version. Maybe I should have tried it.

The GM MCU has a different register set for controlling SPI, SCI, timers, COP, and other stuff. Did you make external models for the surrounding I/O chips in the ECM?
As an example, the 730 reads an A/D by setting up register $400X and then writes and reads the SPI interface..........I don't see a true Motorola 68HC11 simulator handling this correctly.

Please explain. It is much easier to debug on the PC than firing up an ECM on the bench in my opinion.

J

junkcltr

iTrader: (1)

I just tried Wookie V1.70..............It hung a rod with my S19 file.

J

JP86SS

iTrader: (1)

I have ran my S19 on the full version, the demo won't let you open up the memory to load the whole thing.
the program kept throwing me messages that if I change the config register (which the AUJP does) that said "If you do strange things, strange things may happen".
Kind of funny but was just telling me that the program had reassigned the registers other than the std configuration.
It runs as far as I can tell. Gets through all of the memory clearing and configuration stuff.
I'm getting hung up on the SWI commands.
The simulator loops at the command waiting for the software interrupt and I don't know what triggers it and how it gets cleared.
I think its the input signals or DRP that is not counting up to indicate the engine is trying to start.
I've given the interrupt and the program continue, although it bombs out if I don't clear it or reset it at the right times.
I'm sure its something I'm not doing correctly.
When I do "manually" give the signal the program does continue to execute but get lost in the stack because of the interrupt timing.
That's as far as I have gotten but am still in the middle of FULLY commenting the AUJP so I can follow what is going on.
I'm not even close to giving up yet.

junkcltr

iTrader: (1)

The simulator is always going to act strange when it does I/O reads and memory stuff because the ECM Motorola microcontroller is not the same as an off-the-shelf Motorola 68HC11 which the simulator was designed for.
So you will strange behaviour with interrupts, reading I/O, reading/writing memory.

The simulator would be good for the guys on the board that would like to figure out the large displacement with large injector
"problem" in the $8D code (730 ECM). They are maxing out the fixed point math algorithms, (the 16x16 multiply). The MCU uses the center of the result......and they need more bits to the left. They could just manually enter the values (MAT, MAP, BPW, etc) into the registers that the MCU stores after reading the A/D.
Anyway, the simulator is great for that kind of stuff. I tried the Sim68 simulator and it looked good. BUT *none* of the 68HC11 simulators will correctly execute the GM code without customizing the register set, memory space, and I/O space.

J

JP86SS

iTrader: (1)

Do you have any specifics you could share on this. i don't know a ton (or an oz. for that matter) about the hardware and chip differences you mentioned. Maybe I can persuade the writer to code in some of these varations. He seems pretty open to suggestions and has responded to all of my question so far.
I'm going to send him a request to ensure compatibility with the 8192 baud rate already. A few more specific requests may be able to let him evaluate the problems of the std configuration vs what we need to really run the code.

JP86SS

iTrader: (1)

This is the configuration screen of the simulator.
I'm sure my settings are not all correct. my ram1 was just setup for full range. take a look at the pic, I can't seem to get it to paste into the post.

junkcltr

iTrader: (1)

Unfortunately, I don't have any info worth posting on the MCU differences right now. What I mean by that is that I see all kinds of reads & writes to memory and registers that do not make sense when reading the standard Motorola 68HC11 datasheet.

I have starting documenting what I *think* is correct in terms of address space for memory and I/O. I have not tested all of the bits settings in all of the control registers using actual hardware so I would not post info that is probably wrong in many cases.

I have to get my ECM bench set up going so that I can verify the documentation I have created from reading my dissasembly file. With the bench setup, I can write little 5 - 10 line pieces of code that will test what I think is correct.

Once I get the bench running I will post info, but I have gotten hung up messing with a 95 LT1 PCM code/computer/wiring harness........so I am not sure when I will get around to testing out this stuff. As a start to get your feet wet, I recommend checking out the GMECM site. It has info about the "P4" computer (registers & memory). That is the term used for the 68HC11 hybrid Motorola MCU used in the 165 & 730 ECMs I believe.

J