Catalogue of Arcade Emulation Software - the Absolute Reference

Valid XHTML 1.0! Valid CSS!



Large CAESAR Logo


0.37b11 [David Graves]


- With the correct clock speed of 12MHz for the 68000, garbage graphics remain over the Taito logo on startup. This seems to be a bug in the original which would have no effect if our timing was 100% right. The interrupt handling should be quite correct, it's derived straight from the schematics and PAL dump. The current workaround is to make the 68000 run at 13MHz. Lowering below 12MHz would work as well, and possibly be closer to the real reason (wait states slowing the CPU down?)

- The "FIRE!" arrows pointing to padlocks are not in perfect sync with the background scrolling. Should they?

- The quality of the zoomed sprites could probably be better. Drawing them as made by 16x8 tiles loses precision due to limitations of drawgfxzoom().

- Schematics show a OBPRI output to control sprite priority. This doesn't seem to be used however, and isn't hooked up. See othunder_TC0220IOC_w().

- TC0100SCN problem: text vs. bg0/1 offsets are wrong: first level wants bg0 4 further right. Cut screens (all?) want bg0 4 pixels further left. But the bg0 x scroll value is zero in both cases! (and the code setting it is a CLR, so there's no doubt it's meant to be). There are no set bits in the TC0100SCN ctrl regs which might be causing this. So I'm mystified. (Maybe it's related to game being ORIENTATION_FLIP_X ??)


- Driver based on F2 driver and Raine source.

- Hardware: Operation Thunderbolt operates on hardware very similar to the Taito Z system, in particular the game Spacegun. The lightgun hardware in these two (as well as the eerom and calibration process) looks identical. The game has 4 separate layers of graphics - one 64x64 tiled scrolling background plane of 8x8 tiles, a similar foreground plane, a sprite plane, and a text plane with character definitions held in ram. The sprites are 16x8 tiles aggregated through a spritemap rom into 64x64 zoomable sprites. The main difference is that Operation Thunderbolt uses only a single 68000 CPU, whereas Spacegun has twin 68Ks. (Operation Thunderbolt has a Z80 taking over sound duties, which Spacegun doesn't.)

- The game checks an external input called ROTARY in the schematics, not mentioned in the manual. The ROTARY input controls the separation between left and right players, the game adjusts the values written to the external volume controller depending on it. Possible values are: 111 (max) 011 (high) -01 (med) --0 (low). It's a rotary control so only one bit is supposed to be low.

- The keyboard leds I'm turning on are actually the gun solenoid outputs, which would rattle the gun while firing.

- The newer US set has a dip switch option to continue with only one coin even if coinage is set higher. I don't know if this is the only difference with the older set.


- 29th May 2001: David Graves streamlined the I/O handling in most of the Taito 68k based games, cleaned up several things and added default gun calibration values for Space Gun, Operation Thunderbolt and Under Fire.

- 25th March 2001: David Graves added state saving support to the Operation Thunderbolt driver.

- 23rd December 2000: Gerardo Oporto fixed dip switch settings in Operation Wolf / Thunderbolt and the dual screen games.

- 4th December 2000: David Graves added support for a third Taito TC0100SCN chip emulation.

- 1st December 2000: David Graves updated the TC0100SCN Taito chip emulation to support extra wide tilemaps, making Darius 2, Ninja Warriors and Warrior Blade playable.

- 11th November 2000: David Graves sent in preliminary Operation Wolf / Operation Thunderbolt drivers, however only Operation Thunderbolt works.