02885: safarir: [possible] CPU speed of 3 MHz seems suspiciously high and should be confirmed.

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
02885	Original Reference	Minor	Always	Jan 28, 2009, 03:43	May 22, 2011, 06:56

Tester	chowell	View Status	Public	Platform	MAME (Official Binary)
Assigned To	hap	Resolution	Fixed	OS
Status [?]	Resolved			Driver	phoenix/safarir.cpp
Version	0.129u2	Fixed in Version	0.141u3	Build	Normal
		Fixed in Git Commit		Github Pull Request #

Summary	02885: safarir: [possible] CPU speed of 3 MHz seems suspiciously high and should be confirmed.
Description	The safarir.c driver includes the CPU specification: MDRV_CPU_ADD("main", 8080, 3072000) /* 3 MHz ? */ The pcbinfo file for this game notes that the cpu is an 8080 and that the main crystal is 18 MHz, but it says nothing about the CPU clock speed. The standard 8080A could only run at 2 MHz (actually, 2.0833 MHz). The only 8080A version which was rated for 3 MHz (actually, 3.125 MHz) was the 8080A-1. It is possible that Safari Rally used this version, but this would make it the only 8080-based game in MAME to require one; all others ran the CPU at about 2 MHz or lower. Since the 3 MHz CPU clock has not been confirmed, it seems much more likely that the assumed CPU clock divisor is wrong. Another possibility is that the system used an 8085, since the standard version of that chip was rated at 3 MHz, but this would mean the CPU was misidentified in the pcbinfo file.
Steps To Reproduce
Additional Information
Github Commit

Flags	Possible
Regression Version
Affected Sets / Systems	safarir

Attached Files

No.03714 Luigi30 Tester Feb 1, 2009, 21:00 edited on: Feb 1, 2009, 21:00	This came up in kopunch.c, and it was determined that a way to tell 8080 source from 8085 source was the use of the sim instruction, which the 8080 didn't have. 1448: 30 sim 1449: 1D dcr e 144A: 20 rim 145A: 30 sim 145B: 1D dcr e 145C: 20 rim The 8080 doesn't have sim or rim, so it looks like the core should be an 8085A. The driver also says that the hardware is a precursor to the phoenix hardware, which also uses the 8085A.
No.07221 hap Developer Feb 17, 2011, 18:52	Suspiciously high clock issue was resolved somewhere in 0.129. It was also changed to i8085. However, PCB notes from japump, included in the driver src afterwards, mentions CPU:M5L8080AP. It's changed to that now, clock divider is assumed to be 12 (result: 1.5mhz) anyway, closing

No.03714

Luigi30

Tester

Feb 1, 2009, 21:00

edited on: Feb 1, 2009, 21:00

This came up in kopunch.c, and it was determined that a way to tell 8080 source from 8085 source was the use of the sim instruction, which the 8080 didn't have.

1448: 30 sim
1449: 1D dcr e
144A: 20 rim

145A: 30 sim
145B: 1D dcr e
145C: 20 rim

The 8080 doesn't have sim or rim, so it looks like the core should be an 8085A. The driver also says that the hardware is a precursor to the phoenix hardware, which also uses the 8085A.

No.07221

hap

Developer

Feb 17, 2011, 18:52

Suspiciously high clock issue was resolved somewhere in 0.129. It was also changed to i8085. However, PCB notes from japump, included in the driver src afterwards, mentions CPU:M5L8080AP. It's changed to that now, clock divider is assumed to be 12 (result: 1.5mhz)

anyway, closing