Dankjewel voor het toevoegen van Bavaria in Revelation =) I'm sure it will make users happy. I also mentioned in a previous post that I changed the chess drivers MAME source code license to from GPL to BSD3 (Bavaria emulation was part of it). Assuming you got the info from MAME, I'm glad it was of use to you and as I said, the license change is exactly the reason for this, so modern chesscomputer companies can refer to it without requiring to release the source code.

The protocol is indeed very simple, when I wrote the code, I was actually a little surprised at how straightforward it is. Much better than Tasc's piece-recognition board version.

Franz,

I implemented the Tasc board based upon the SB with the induction coils. So i had to use specific timing, which is quite problematic as i can set the speed for the complete emulation as well. This was a nice implementation but very difficult. I assume you used the SB with the digital coding. If so, we have different approaches to the tasc emulations I now realise that coding with Mame as a basis makes life somewhat easier.

When will you implement the Novag Robot Adversary under Mame ..

Regarding the Bavaria, your analysis was mentioned in a post on schachcomputer.info. I tried to implement it a couple of years ago. But no succes. Combining your tips and the source of the IO driver from Richard gave me the clue. The 3 us delay for the busy flag gave the needed info / push. I simply did not wait enough here. And then the emulation does not detect the board from software. I changed the wait to over 3uS, or to be precise, minimal 1 interrupt routine further.

PS, how did you analyse the Bavaria ? Did you use a real bavaria board ? Or did you analyse the assembly code ?

__________________
Grusse,
Ruud Martin

'hap' is our expert for MAME hardware drivers, not me.

I don't know much about the Tasc Smartboard emulation, Sandro wrote that. I think the one he emulated is the 1st SB30 version, the one with 81 leds.
*edit* actually, looking at the code, it looks more like SB20, but with a workaround to control 81 leds instead of 64.

https://github.com/mamedev/mame/blob...ivers/tasc.cpp
https://github.com/mamedev/mame/blob...smartboard.cpp

For Bavaria, I simply 'listened' to reads and writes while the WIP emulation is running, and tried to reverse engineer them. Sometimes I won't even need to look at the disassembly. I don't have any chess hardware at home. When I need more info, I can derive it from good quality pcb photos. I also got a lot of help from Berger with the chesscomputer emulations, doing hardware tests and he even made schematics for some.

If I had the Novag Robot roms, I can give it a try. Or does it have an MCU with embedded ROM like the Savant? That one wasn't easy to dump, we had to send it to an expert.

And indeed, I'm not Franz.

Hallo Volker!
Ok, die neue Version ist online.

Ich habe die 36MHz bei allen WM-Versionen hinzugefügt und außerdem auch beim 'TM London 68040 (experimental)' den Frequenzbereich noch um 36/50/60 MHz erweitert (falls jemand auch an einem Vergleich 68030 vs. 68040 interessiert ist).

Außerdem habe ich bei den (im letzten Update in MessExtra eingebauten) Mephisto Modular 24MHz-Geräten (Almeria/London/Lyon/Portorose/Vancouver) jetzt die Auswahl auch auf 24/30/36 MHz erweitert, sodaß nun eigentlich auch der 'TM Almeria' (mit 30MHz) einstellbar ist (sofern das Programm im ROM tatsächlich das gleiche sein sollte). Das gilt aber nur beim Aufruf dieser Modelle direkt in MessExtra oder MessChess (nicht mit CB-Emu!) - daher heißen diese Modelle dort jetzt auch '(Turbo)' (anstatt '(24 MHz)' wie in CB-Emu).

Jetzt ist aber Schluß mit Sonderwünschen ...

Grüße,
Franz

Hi Michael (?, ich kenne mich nicht so gut mit alle namen )
Thanks for the info. yes, mame also has a good debugger.
I have the novag Adversary roms, these are somewhat older and i had to experiment to read them. The emulation is quite complex, as i completely emulated all motors to give a display on the Revelation II with the moving arm. Loads of signals and all kind of protection mechanism regarding the arm not going ballistic

__________________
Grusse,
Ruud Martin

Yup, my name is Michaël.
I can imagine that the robot arm was complex to simulate =)

Looking at Berger's photos here: https://sites.google.com/site/proyec...ez/novag_robot , the rest of the chesscomputer looks relatively simple. Z80 controls the robot arm, no MCU that I expected.

Hallo Franz,

ich danke Dir für die Umsetzung der Konfiguration zusätzlicher Taktfrequenzen für die Mephisto WM-TM!

MfG
VOLKER

Jetzt ist aber Schluß mit Sonderwünschen ...



Hallo Franz,

wie oft hast du das schon geschrieben.😉

VG Heiko

__________________
https://www.schachcomputer-online-museum.de

Neu: Berger hat Excel Mach IV roms ausgelesen, und es ist jetzt emuliert in MAME.