PROJECT : A TRS80 MC10 emulator on the ZX Spectrum

[Dr beep]

I am working on a loadroutine.

I did a blockload of my SHOGUN-game and I was able to play my own game, although slower than on MC10.

After the load I will release the first version of the emulator with games to play.

Next version will get more translated ROM-routines and if you have found bugs I will try to fix them.

[Dr beep]

A programfile on the MC10 has bytes that mimic headersound.
I can't load a game like on the ZX81 emulator.

I will make a 8K+ virtual tape memory and will then load from that virtual tape

[Dr beep]

I rearrangent the program so I now have 8K+ for a virtual tape, room for 8K RAM in the emulator
and still 8K free to translate the ROM.

[Dr beep]

The MC10 uses blocks to load a program, there is no 1 block of code that loads a game like the ZX Spectrum does.
This also means that each file haas a lot of overhead and a 8K program can be over 10K in size.

This means that I will be short of memory when I make a virtual tape memoryblock.
I will make a convertertool like on SP-2-ZX81 and then make a loadoption with just 1 block to load.

[Dr beep]

The convertertool is working, a loader must be added but I did a first load with a memorydump.

Next thing was a bug in the IF with a string comparing, so that is next to solve.

[Dr beep]

The bug was in an opcode I thought was solved, but it needed a different test.

Now the loader and then a first release?????

[Dr beep]

Another bug found.

rotate on the MC10 also does a test on the result, a set of Z when result is 0.

The Z80 doesn't alter z-flag

[Dr beep]

A new bug appears but I know that some opcodes are not complete, so maybe do that first

[Dr beep]

Solved a bug but found out that a fix of a bug causes a new bug in a routine that worked due to NOT getting an altered value, so next bug to find.

[Dr beep]

Previous bug again solved. Some weird behaviour that I could only find by stepping on a pc-emulator of the MC10.

Next bug is in the RND-routine.
Problem: when seed is different my emulator and pc-version will not run the same code.

I have managed to set the rnd-seeds to a default value so the code will be the same each run.

[Dr beep]

So yesterday I found the RND bug and a game played well. Today I was working on release of v1.00 when I added a game and that game gave another error, so back to debugging before releasing v1.00

[Dr beep]

Version 1.00 has the debugger active which cost some speed. I translated the routine at F160 and the program
1FORF=0TO255:PRINTF;:NEXT is 14 sec faster already.
So next routines will speed up more.

[Dr beep]

I already translated mutliple used routines, but the next is a large one but it will definitely add a lot of speed (expectation).
The "evaluate a an expression" is used in almost every command. So up to translating that routine.

[Dr beep]

The evaluation routine uses some FP-routines I had not translated.
So first all FP-routines and now up to the evaluationroutine.

[Dr beep]

Changes in the latest 2 version made a game crash, so back to earlier version, but latest change gave 20% speedup in the basic.

[Dr beep]

I have decided to stop recoding the ROM and release the emulator as it is now.

I hope to have the emulator ready before end of the week.

[zx_if1]

hi dr. beep
How about a little suggestion?
adapt the emulator so that it can recognizes the ula+
If there is ula+ orange appears then
if not only dull yellow

Another little idea: will there be a Sam version in the future?

[Dr beep]

hi dr. beep
How about a little suggestion?
adapt the emulator so that it can recognizes the ula+
If there is ula+ orange appears then
if not only dull yellow

How do you recognize the ULA and can it be solved with a single attribute value?
That is how the colour is set.

Another little idea: will there be a Sam version in the future?

I am thinking of such, but my SAM knowledge is really rusty.
On a SAM the memory could be used as on the MC10 so all changes in memorylocation can be skipped and it will
get some extra speed, besides the Z80B processor.

[dust hill resident]

Hello,

This is a really cool project and I have a lot of respect for your work, thank you for sharing it and the development journey with us.

Also, thank you for bringing the TRS80 MC10 to my attention, I had never heard of it before, and I've just spent the past week or so reading about it and playing with it in emulation, it's a really nice and fun little computer.

- dhr

[kokkiklhs]

Hi Dr beep, thank you again for this new brilliant piece of software of yours!
Astonishing achievement, superb coding skills!!!
You are always full of nice surprises!!!

Well, one question:
I have downloaded and tested the 1.12 version, plays nicely and the software included loads fine! But how about converting some more .c10 files? I have tried with the assorted utility for the speccy but I got confused, couldn't figure out how to use it properly...
Do we need to fill a specific RAM area with the data, using a Spectrum emulator and then save it to .tap? I am using Spectaculator, which lacks that feature!
Can you please explain the exact procedure?
TIA

[Dr beep]

Here is a video how the conversion works.

[kokkiklhs]

Oh, thank you! That's exactly what I was suspecting, so I guess I'll quit Spectaculator and download ZXSpin instead!

[Dr beep]

Oh, thank you! That's exactly what I was suspecting, so I guess I'll quit Spectaculator and download ZXSpin instead!

Or find out how Spectaculator works with tap-files.

I can do the same with other emulators. In my video how 48 KRAAK works I use Z80.