Hi,
und ist auch schon wieder in CB-Emu eingebaut - vielen Dank an 'hap' und Berger!

Grüße,
Franz

__________________
Meine Webseite: https://fhub.jimdofree.com/

Hallo zusammen,
bei meinen Turnieren ist mir aufgefallen, dass der Fidelity Avantgarde 2100 immer sehr schlecht abschneidet.

Ein Test ergab, dass wenn ich auf 500% Speed gehe, dieser nur max 230% anzeigt. Dies bedeutet natürlich einen Nachteil für dieses Programm. Nach ein paar Zügen wird sogar nur noch ~130 % max angezeigt.

Bei dem emulierten 6502 Prozessor kann ich auf meinem Laptop mit I7 Prozessor bei allen anderen Programmen problemlos auf 500% spielen.

Meine Frage ist also, kann das jemand bestätigen? Wenn ja, liegt bei dem Programm ein Fehler in der Emulation vor?

P.S. Ich spiele mit der Arena Oberfläche!

Yes, the CPU divider emulation used in several Fidelity drivers is slow.

If you want to run the chesscomputer unthrottled anyway, you could disable the divider completely, it's a configuration option.

hap, a short question just to be sure that I understand this CPU divider correctly:
in some situations (e.g. when it's the opponents turn and the program is not pondering) the chess program itself reduces the CPU frequency by a factor 2 (or 4, depending on the dip switch setting) to save battery power - is this correct?
And have these real devices only a dip switch for factor 2 or 4, or can this also be switched off completely?

__________________
Meine Webseite: https://fhub.jimdofree.com/

The CPU divider is from the address bus (in most cases, the CPU program counter). I assume it's due to slow EPROMs, similar how for example Mephisto Glasgow has wait states with DTACK.

No it's not a switch, it's a wire. The only way a user could change it is opening the chesscomputer and resoldering the wire or removing it.
(and maybe break the chesscomputer when the EPROMs can't handle the increased speed... so don't try it )

If you override these dividers in a real Fidelity chess computer, I can guarantee that it will crash or malfunctioning, even if you have installed faster EPROMs. There are many more changes that would have to be made. But when it comes to emulation... I don't have an opinion on that.

Best,
Gerardo

Oh, that's a completely different thing. So this means that the CPU frequency is switched from f to f/2 (or f/4) whenever the EPROM is addressed and back again to f when addressing the RAM?
But this had to be done for almost every program step (since the program is executed from the EPROM), so isn't his extremely inefficient?
Wouldn't it be better/simpler then to just use a slower CPU frequency all the time?
Well, I don't think that switching off the CPU divider would be any problem in the emulation - at least I've not yet run into any troubles doing it.

__________________
Meine Webseite: https://fhub.jimdofree.com/

MAME doesn't emulate the address bus at a low level, so yes the implementation is inefficient.
It is as you say, the CPU runs slower when it's accessing the EPROM area. Simply lowering the overall CPU freq is not the same thing.

Disabling it in emulation, the chesscomputer would be faster, I don't expect problems.

Neuer treiber: SciSys Chess Companion II und Explorer Chess. Danke an Sean Riddle fur ROM dump.
Even though the Explorer has less buttons, it's the same ROM as Companion II. Concord is also on this MCU, but other than aesthetics it is identical to Companion II.

This is the first chess computer with a Kaare Danielsen chess engine, and also the first Kaare Danielsen one that's ever emulated. It is the 4KB program.
Sean Riddle was finally able to dump the HD6301V1 MCU.
It's not known yet if or how HD6301Y0 (16KB ROM) can be dumped.

Hello,

The 6301Y seems to have an extra protection to prevent extraction of the rom.

Cordialy.