|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
I documented the Split Second PCB. The TMS1400 has a different pinout than the TMS1100- probably because the die is wider due to the additional ROM and had to be rotated to fit. I guessed at the signals, assuming the relative positions of the pins are mostly the same. So my assignment of the R0-R10 and O0-O7 signals could be reversed. The OSC pins are connected together and to a 100pF cap to Vss and a 24K resistor to Vdd. I couldn't measure the clock rate directly, but did measure a 1.5mS period on pin 11 directly after power on with "1" displayed, and a 2.2mS period on pin 25. www.seanriddle.com/splitsecond.txt
|
|
|
|
Joined: Mar 2006
Posts: 1,080 Likes: 7
Very Senior Member
|
Very Senior Member
Joined: Mar 2006
Posts: 1,080 Likes: 7 |
http://marc.info/?l=classiccmp&m=104965276917315&w=2 has tms1400 pinout (presumably?) That thread also HAD dark tower pcb pics, but they're long gone. Its possible the person who posted them back in 2001 still has them around somewhere. LN
"When life gives you zombies... *CHA-CHIK!* ...you make zombie-ade!"
|
|
|
|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
Excellent! So I got everything correct except I had reversed O7-O0. I updated my doc.
I'll wind up with a Dark Tower someday. Wildfire Pinball is on the way.
|
|
|
|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
I removed the top metal layer from Split Second and found some bit errors in my previous transcription: $2F4 from $30 to $40, $4DB from $FC to $EC and $909 from $78 to $68. I updated the files on my web site and uploaded the new pic. http://www.seanriddle.com/tms1400.html
|
|
|
|
Joined: Mar 2002
Posts: 1,371 Likes: 132
Very Senior Member
|
Very Senior Member
Joined: Mar 2002
Posts: 1,371 Likes: 132 |
Great work =) I should get started on emulating it in MESS soon. "I'll wind up with a Dark Tower someday. Wildfire Pinball is on the way." Hmm, according to the patent, Wildfire runs on an AMI S2150 MCU. http://www.google.com/patents/US4334679 (includes rom data). Dark Tower is likely a TMS1400.
|
|
|
|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
I also found an error on Electronic Detective: $64D from $DD to $CC. I updated the files on my web site and uploaded the new pic. www.seanriddle.com/0980.htmlThe Wildfire IC is an AMI, labeled C10641. Thanks for the patent. It looks like you can add external ROM- I wonder if you can program an EPROM to dump the internal ROM?
|
|
|
|
Joined: Mar 2002
Posts: 1,371 Likes: 132
Very Senior Member
|
Very Senior Member
Joined: Mar 2002
Posts: 1,371 Likes: 132 |
From glancing over the programming manual... I don't think so. The MCU always boots at bank 0 which is the internal ROM bank. Besides, like many other similar-to-TMS1000 MCUs of the time, every byte is an opcode: it doesn't have the concept of data bytes. ref: http://bitsavers.trailing-edge.com/pdf/ami/s2000/
|
|
|
|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
Page 2 of that manual says "built-in production test mode", but I don't see anything else in the manual about that. But bitsavers also has a data book with S2000 info that has more detail: http://bitsavers.trailing-edge.com/pdf/ami/_dataBooks/1979_AMI_MOS_Products.pdfThe ROMS input is tied high for internal ROM only, low for internal and external ROM, but can also be tied to the SYNC output to "override bank 0 with an external program" or tied to inverted SYNC to "verify internal ROM contents". That sounds promising!
|
|
|
|
Joined: May 2010
Posts: 1,051
Very Senior Member
|
Very Senior Member
Joined: May 2010
Posts: 1,051 |
I typed in the Wildfire code from the patent. It's in SREC format, so there is address and checksum info to help validate the bytes. www.seanriddle.com/wildfire4334679.txtwww.seanriddle.com/wildfire4334679.binIt's 1.5K, non-contiguous (0000-03FF, 0600-07FF), with 5 unused bytes. The Wildfire PCB connections to the microprocessor don't quite match the AMI S2150 pinout. Pin 1 is ground and 29 is +9v like the pinout shows, but +9v is also connected to pins 2 and 6- pin 2 is supposed to be D2 and pin 6 is an output. Also, pin 7 is not connected, but that should be A12, which is shown on the patent schematic. The PCB generally matches the patent schematic otherwise. I'll trace all the connections and see what's different between the chip used and the S2150 pinout. I'll see if I can dump the chip before I decap it.
|
|
|
|
Joined: Mar 2002
Posts: 1,371 Likes: 132
Very Senior Member
|
Very Senior Member
Joined: Mar 2002
Posts: 1,371 Likes: 132 |
hmm, Stiletto can you find out which other worth-to-emulate devices use the AMI S2000 series MCU? (I poked him on IRC so he knows I called him)
|
|
|
3 members (Cpt. Pugwash, 2 invisible),
193
guests, and
0
robots. |
Key:
Admin,
Global Mod,
Mod
|
|
Forums9
Topics9,331
Posts122,199
Members5,077
|
Most Online1,283 Dec 21st, 2022
|
|
These forums are sponsored by Superior Solitaire, an ad-free card game collection for macOS and iOS. Download it today!
|
|
|
|