MMC5: Revision history

2 April 2024

• curprev 14:4814:48, 2 April 2024‎ Ben Boldt talk contribs‎ 46,918 bytes +220‎ →‎Scanline IRQ Status ($5204, read/write): Clarified that the scanline does not reset from a scanline interrupt.
• curprev 09:0709:07, 2 April 2024‎ Bregalad talk contribs‎ 46,698 bytes +192‎ Move the split mode info to the split mode registers. Separates frame and blanking data access in the table.
• curprev 07:3907:39, 2 April 2024‎ Ben Boldt talk contribs‎ 46,506 bytes +198‎ →‎Scanline Detection and Scanline IRQ: updated a couple things having to do with scanline counter resetting.
• curprev 07:2507:25, 2 April 2024‎ Ben Boldt talk contribs‎ 46,308 bytes +353‎ →‎PPU Data Substitution Enable ($2001 = PPUMASK): Added info about scanline counter.

1 April 2024

• curprev 20:5120:51, 1 April 2024‎ Bregalad talk contribs‎ 45,955 bytes +66‎ →‎Internal extended RAM mode ($5104): Mention we were talking about mode %01, mention the open bus.

30 March 2024

• curprev 04:2804:28, 30 March 2024‎ Ben Boldt talk contribs‎ 45,889 bytes +251‎ →‎Internal extended RAM mode ($5104): Improved connection between PPU address and CPU address for extended RAM data.

29 March 2024

• curprev 22:5522:55, 29 March 2024‎ Bregalad talk contribs‎ m 45,638 bytes 0‎ →‎Internal extended RAM mode ($5104): My previous edit screw up the order of the info, this edit fixes it.
• curprev 13:0513:05, 29 March 2024‎ Ben Boldt talk contribs‎ m 45,638 bytes +7‎ →‎Vertical Split Mode ($5200)
• curprev 12:5912:59, 29 March 2024‎ Ben Boldt talk contribs‎ 45,631 bytes +849‎ →‎Vertical Split Mode ($5200): Improved description, fixed problems where I was saying that the MMC5 substitutes pattern data.

28 March 2024

• curprev 08:2908:29, 28 March 2024‎ Bregalad talk contribs‎ 44,782 bytes +32‎ Oops, better sorting of the info between footnote and extended attribute mode
• curprev 08:2308:23, 28 March 2024‎ Bregalad talk contribs‎ 44,750 bytes −112‎ →‎Internal extended RAM mode ($5104): Avoids to tell the same thing twice, removed the incorrect info of extended attribute being 4-screen mirrored when it is actually single-screen mirrored

27 March 2024

• curprev 03:4503:45, 27 March 2024‎ Ben Boldt talk contribs‎ m 44,862 bytes +89‎ →‎MMC5A
• curprev 03:3703:37, 27 March 2024‎ Ben Boldt talk contribs‎ 44,773 bytes +429‎ →‎MMC5A Registers: Corrected errors related to register $5207.
• curprev 00:3200:32, 27 March 2024‎ Ben Boldt talk contribs‎ m 44,344 bytes +7‎ →‎Nametable mapping ($5105)
• curprev 00:1900:19, 27 March 2024‎ Ben Boldt talk contribs‎ 44,337 bytes +759‎ →‎Internal extended RAM mode ($5104): Updated description detailing what happens when assigning ExRAM as nametable when extended attributes are enabled.

26 March 2024

• curprev 21:5821:58, 26 March 2024‎ Ben Boldt talk contribs‎ 43,578 bytes +71‎ →‎Scanline IRQ Status ($5204, read/write): Pin 92 low disables scanline IRQs.
• curprev 21:5421:54, 26 March 2024‎ Ben Boldt talk contribs‎ 43,507 bytes +355‎ →‎8x16 PPU Sequence Monitoring Enable ($2001 = PPUMASK)
• curprev 21:3521:35, 26 March 2024‎ Ben Boldt talk contribs‎ 43,152 bytes +319‎ →‎NES internal state monitoring: Corresponding update to description of $2001. I need to check if scanline IRQs get disabled this way as well.
• curprev 21:2421:24, 26 March 2024‎ Ben Boldt talk contribs‎ 42,833 bytes +135‎ →‎Internal extended RAM mode ($5104): Added PPUMASK info for extended attribute mode.
• curprev 21:2121:21, 26 March 2024‎ Ben Boldt talk contribs‎ 42,698 bytes +101‎ →‎Vertical Split Mode ($5200): I found that vertical split mode is disabled by PPUMASK monitoring, similar as 8x16 sprite mode.
• curprev 02:2002:20, 26 March 2024‎ Ben Boldt talk contribs‎ 42,597 bytes +115‎ →‎Vertical Split Mode ($5200): More details, things that automatically disable split mode.
• curprev 02:1402:14, 26 March 2024‎ Ben Boldt talk contribs‎ 42,482 bytes +40‎ →‎Vertical Split Mode ($5200): Pin 92 low disables split mode.
• curprev 01:5301:53, 26 March 2024‎ Ben Boldt talk contribs‎ m 42,442 bytes +9‎ →‎Internal extended RAM mode ($5104)
• curprev 01:5301:53, 26 March 2024‎ Ben Boldt talk contribs‎ 42,433 bytes +346‎ →‎Internal extended RAM mode ($5104): I must have goofed something up before because I absolutely am not getting nametable data to work in mode %11. Tested and added split mode info.

25 March 2024

• curprev 22:1122:11, 25 March 2024‎ Ben Boldt talk contribs‎ m 42,087 bytes +474‎ →‎Vertical Split Mode ($5200): Additional cleanup.
• curprev 21:0521:05, 25 March 2024‎ Ben Boldt talk contribs‎ 41,613 bytes +411‎ →‎Vertical Split Mode ($5200): Clarifications as I am reading and remembering how this works. No new information in this edit.
• curprev 19:5419:54, 25 March 2024‎ Ben Boldt talk contribs‎ 41,202 bytes +130‎ →‎Fill-mode color ($5107): Fill mode color vs. pin 92 low when using extended attributes with fill mode.
• curprev 19:4619:46, 25 March 2024‎ Ben Boldt talk contribs‎ 41,072 bytes −278‎ →‎Configuration: removed incorrect statements that were caused by a test setup issue. My test waited for in-frame before doing these writes, so the test itself got disabled with pin 92 low.
• curprev 19:1219:12, 25 March 2024‎ Ben Boldt talk contribs‎ 41,350 bytes +13‎ →‎Internal extended RAM mode ($5104): additional note clarifying that extended ram writes via ppudata does not become available when pin 92 is driven low. I.e. pin 92 low is not just pretending to be in a different mode.
• curprev 19:0619:06, 25 March 2024‎ Ben Boldt talk contribs‎ 41,337 bytes +84‎ →‎Internal extended RAM mode ($5104): pin 92 low prevents write access to $5C00
• curprev 08:3508:35, 25 March 2024‎ Ben Boldt talk contribs‎ 41,253 bytes +320‎ →‎Fill-mode color ($5107): added more fill mode info.
• curprev 03:2403:24, 25 March 2024‎ Ben Boldt talk contribs‎ 40,933 bytes +201‎ →‎Internal extended RAM mode ($5104): Writing during v-blank in modes 00,01 is not writing zeros for me. It either doesn’t write or it corrupts it.
• curprev 02:2002:20, 25 March 2024‎ Ben Boldt talk contribs‎ 40,732 bytes +41‎ →‎Nametable mapping ($5105)
• curprev 02:1502:15, 25 March 2024‎ Ben Boldt talk contribs‎ m 40,691 bytes 0‎ →‎Internal extended RAM mode ($5104)
• curprev 02:1202:12, 25 March 2024‎ Ben Boldt talk contribs‎ 40,691 bytes +466‎ →‎Internal extended RAM mode ($5104): Details added for $5104 register. Testing revealed that mode 11 does work as a read-only nametable.
• curprev 01:3601:36, 25 March 2024‎ Ben Boldt talk contribs‎ 40,225 bytes +64‎ →‎Scanline IRQ Status ($5204, read/write): pin 92 info

24 March 2024

• curprev 21:1421:14, 24 March 2024‎ Ben Boldt talk contribs‎ 40,161 bytes +137‎ →‎Scanline IRQ Status ($5204, read/write): In-frame bit clarification. I think we knew this already but I did test to confirm the h-blank. I did a test program that reads $5204, ROL*3, write to $4016, then watched OUT0 with a scope.

31 December 2023

• curprev 03:1603:16, 31 December 2023‎ Ben Boldt talk contribs‎ m 40,024 bytes 0‎ →‎PRG Bankswitching ($5113-$5117): Made example select ROM since 10 was larger than possible for RAM.

30 December 2023

• curprev 00:3900:39, 30 December 2023‎ Ben Boldt talk contribs‎ 40,024 bytes +6,269‎ →‎PRG Bankswitching ($5113-$5117): Bankswitch register clarification. No new info; just trying to explain it better.

18 January 2023

• curprev 00:0300:03, 18 January 2023‎ Rainwarrior talk contribs‎ 33,755 bytes +1‎ Internal extended RAM section is not part of MMC5A

17 January 2023

• curprev 23:4823:48, 17 January 2023‎ Rainwarrior talk contribs‎ m 33,754 bytes +11‎ quick link to MMC5A from lead
• curprev 23:4723:47, 17 January 2023‎ Rainwarrior talk contribs‎ 33,743 bytes +173‎ →‎MMC5A Only: create an MMC5 section for the article to hoist it up as a visible heading and make the division a little clearer
• curprev 23:4323:43, 17 January 2023‎ Rainwarrior talk contribs‎ 33,570 bytes +168‎ mention MMC5A's unused feature status in lead

27 October 2022

• curprev 03:4203:42, 27 October 2022‎ Ben Boldt talk contribs‎ m 33,402 bytes −254‎ →‎Scanline Detection and Scanline IRQ: Removed the last paragraph about reset detection, which was already explained farther up.

5 October 2022

• curprev 21:3421:34, 5 October 2022‎ Rainwarrior talk contribs‎ 33,656 bytes +53‎ →‎Vertical Split Scroll ($5201): note that SL mode is unused

9 August 2022

• curprev 21:2121:21, 9 August 2022‎ Ben Boldt talk contribs‎ m 33,603 bytes −2‎ →‎Internal extended RAM mode ($5104): Removed unnecessary "1=" from table.

27 June 2022

• curprev 13:5813:58, 27 June 2022‎ Ben Boldt talk contribs‎ m 33,605 bytes −43‎ →‎Upper CHR Bank bits ($5130): cleanup
• curprev 03:4403:44, 27 June 2022‎ Ben Boldt talk contribs‎ 33,648 bytes −50‎ →‎Fill-mode tile ($5106): I remembered asking about this in the forum. I updated this with what I learned there.
• curprev 03:3203:32, 27 June 2022‎ Ben Boldt talk contribs‎ m 33,698 bytes +3‎ →‎Fill-mode color ($5107): fixed a typo
• curprev 01:3101:31, 27 June 2022‎ Ben Boldt talk contribs‎ 33,695 bytes +1,641‎ Reorganized and clarified information about extended RAM mode.

4 February 2022

• curprev 17:2717:27, 4 February 2022‎ Ben Boldt talk contribs‎ 32,054 bytes +203‎ →‎Write: Add warning to register $5208.

17 December 2021

• curprev 21:2421:24, 17 December 2021‎ Lidnariq talk contribs‎ 31,851 bytes +253‎ →‎CHR Bankswitching ($5120-$5130): another try
• curprev 19:3919:39, 17 December 2021‎ Lidnariq talk contribs‎ 31,598 bytes +271‎ →‎CHR Bankswitching ($5120-$5130): try to do a better job of explaining why 8x16 sprite mode is fun

8 November 2021

• curprev 17:0517:05, 8 November 2021‎ Quietust talk contribs‎ m 31,327 bytes 0‎ →‎Sound ($5000-$5011): don't forget $5015
• curprev 16:1316:13, 8 November 2021‎ Lidnariq talk contribs‎ 31,327 bytes +14‎ →‎Sound: include range on this page

19 January 2020

• curprev 23:0623:06, 19 January 2020‎ Lidnariq talk contribs‎ 31,313 bytes +12‎ →‎Write: remind reader that M2 = CPU cycle. also edit last sentence to be a little more fluid

17 November 2019

• curprev 11:2011:20, 17 November 2019‎ Rainwarrior talk contribs‎ m 31,301 bytes −10‎ →‎PRG-RAM configurations: fix grammar, remove unnecessary "mirror of"
• curprev 10:4710:47, 17 November 2019‎ Rainwarrior talk contribs‎ 31,311 bytes +31‎ →‎PRG-RAM configurations: maybe the miscellaneous notes should come after the main description and table
• curprev 10:3010:30, 17 November 2019‎ Rainwarrior talk contribs‎ m 31,280 bytes +10‎ →‎PRG-RAM configurations: ExROM link
• curprev 10:2910:29, 17 November 2019‎ Rainwarrior talk contribs‎ 31,270 bytes −754‎ →‎PRG Bankswitching ($5113-$5117): realizing my edits were a bit redundant to the previous heading, removing and incorporating it
• curprev 10:2410:24, 17 November 2019‎ Rainwarrior talk contribs‎ m 32,024 bytes +39‎ →‎PRG-RAM configurations: editing, additional quick note about 16K games
• curprev 10:1010:10, 17 November 2019‎ Rainwarrior talk contribs‎ 31,985 bytes +1,050‎ →‎Emulation quirks with commercial games: This advice is too difficult for me to parse. Trying to reorganize this, making a table to try and illustrate the situation.

9 October 2019

• curprev 05:5705:57, 9 October 2019‎ Zeromus talk contribs‎ 30,935 bytes +3‎ →‎Scanline Detection and Scanline IRQ
• curprev 05:5605:56, 9 October 2019‎ Zeromus talk contribs‎ 30,932 bytes +234‎ →‎Scanline Number IRQ ($5203)

19 February 2019

• curprev 23:0923:09, 19 February 2019‎ Ben Boldt talk contribs‎ 30,698 bytes +49‎ →‎Scanline Detection and Scanline IRQ: Added that scanline 0 detection auto-acknowledges IRQ.

15 February 2019

• curprev 23:5523:55, 15 February 2019‎ Ben Boldt talk contribs‎ 30,649 bytes +241‎ →‎Scanline Detection and Scanline IRQ: Important clarifications to system reset detection, showing that it is identical to disabling scanline IRQ.
• curprev 23:2823:28, 15 February 2019‎ Ben Boldt talk contribs‎ 30,408 bytes +288‎ →‎Scanline Detection and Scanline IRQ: Added info about reset detection
• curprev 22:1822:18, 15 February 2019‎ Ben Boldt talk contribs‎ 30,120 bytes +156‎ →‎Scanline Detection and Scanline IRQ: Clarification, writing to $2001

11 February 2019

• curprev 20:5820:58, 11 February 2019‎ Ben Boldt talk contribs‎ 29,964 bytes −59‎ →‎Scanline Detection and Scanline IRQ: Sorry, I was wrong about NT vs AT, my test had an error, I confirmed it doesn't matter.
• curprev 20:2720:27, 11 February 2019‎ Ben Boldt talk contribs‎ 30,023 bytes +11‎ →‎Scanline Detection and Scanline IRQ: Updated scanline IRQ occurs on PPU cycle 4 of the scanline. Also adding info that the matching fetches must not be from AT range.

9 February 2019

• curprev 15:1115:11, 9 February 2019‎ Ben Boldt talk contribs‎ 30,012 bytes +207‎ →‎Scanline Detection and Scanline IRQ: Clarify, some things that clear the 'in frame' flag do not clear the scanline IRQ.
• curprev 08:0608:06, 9 February 2019‎ Ben Boldt talk contribs‎ 29,805 bytes +763‎ Tweaked some things based on recent observations.

30 January 2019

• curprev 13:1713:17, 30 January 2019‎ NewRisingSun talk contribs‎ m 29,042 bytes +1‎ →‎Scanline Detection and Scanline IRQ
• curprev 13:0913:09, 30 January 2019‎ NewRisingSun talk contribs‎ 29,041 bytes +472‎ Revised description of the scanline detector

20 January 2019

• curprev 19:0319:03, 20 January 2019‎ Ben Boldt talk contribs‎ 28,569 bytes +20‎ Added pinout link to infobox.

18 January 2019

• curprev 06:2106:21, 18 January 2019‎ Lidnariq talk contribs‎ m 28,549 bytes +74‎ →‎Vertical Split Scroll ($5201): MMC5 doesn't keep you from setting the fine vertical scrolling differently across left-and-right split screen
• curprev 06:0206:02, 18 January 2019‎ Lidnariq talk contribs‎ m 28,475 bytes +11‎ →‎Operation Notes: an ellipsis is a bit chatty. so is use of "you"
• curprev 05:4605:46, 18 January 2019‎ Lidnariq talk contribs‎ m 28,464 bytes −9‎ →‎PRG-RAM bankswitching quirks: clarify
• curprev 05:4305:43, 18 January 2019‎ Lidnariq talk contribs‎ m 28,473 bytes +12‎ →‎CHR Bankswitching ($5120-$5130): fix internal link

12 January 2019

• curprev 18:3118:31, 12 January 2019‎ Tepples talk contribs‎ m 28,461 bytes +29‎ →‎See also: add references element for cite.php

11 January 2019

• curprev 21:3221:32, 11 January 2019‎ Ben Boldt talk contribs‎ 28,432 bytes −64‎ Updated, confirmed in thread that split screen mode is used in the ending of Bandit Kings of Ancient China.

4 January 2019

• curprev 05:5805:58, 4 January 2019‎ Ben Boldt talk contribs‎ m 28,496 bytes +36‎ Added category "Mappers with cycle IRQs", technically only MMC5A but nonetheless.

3 January 2019

• curprev 20:2020:20, 3 January 2019‎ Lidnariq talk contribs‎ 28,460 bytes +4‎ →‎Configuration: Split out monitor registers. While the monitored copy at $2000 and $2001 are definitely part of configuration, the later ones may not be, and it's obfuscatory to include it there.
• curprev 08:4008:40, 3 January 2019‎ Ben Boldt talk contribs‎ m 28,456 bytes +1‎ →‎Registers: Moved note to include all PPU registers.
• curprev 04:0004:00, 3 January 2019‎ Ben Boldt talk contribs‎ m 28,455 bytes −2‎ typo fix
• curprev 02:0702:07, 3 January 2019‎ Ben Boldt talk contribs‎ 28,457 bytes +638‎ →‎Registers: Added newly found sniffed registers

31 December 2018

• curprev 02:5802:58, 31 December 2018‎ Lidnariq talk contribs‎ m 27,819 bytes −4‎ →‎8x16 mode enable 2 ($2001): edit-o
• curprev 02:1602:16, 31 December 2018‎ Lidnariq talk contribs‎ m 27,823 bytes +1‎ →‎CHR Bankswitching ($5120-$5130): braino
• curprev 02:1502:15, 31 December 2018‎ Lidnariq talk contribs‎ 27,822 bytes +70‎ →‎CHR Bankswitching ($5120-$5130): backref
• curprev 02:1302:13, 31 December 2018‎ Lidnariq talk contribs‎ m 27,752 bytes +6‎ →‎8x16 mode enable 2 ($2001): forgot to close cite
• curprev 02:1302:13, 31 December 2018‎ Lidnariq talk contribs‎ 27,746 bytes +739‎ →‎Configuration: Technically these registers are part of the MMC5

25 December 2018

• curprev 02:3902:39, 25 December 2018‎ Tepples talk contribs‎ 27,007 bytes +314‎ →‎Emulation quirks with commercial games: Relying on 64K fallback doesn't help repro makers, per lidnariq

22 December 2018

• curprev 19:0519:05, 22 December 2018‎ Lidnariq talk contribs‎ m 26,693 bytes −36‎ →‎CHR select $5120-$512B: Both "LSBit from register ignored" and "upper bits from register ignored" behaviors exist. Only VRC6, Mapper 90, MMC1, and MMC5 can select multiple different banking layouts. VRC6 PRG, m90 PRG & CHR, MMC5 CHR ignore MSbit.
• curprev 15:2215:22, 22 December 2018‎ Bregalad talk contribs‎ m 26,729 bytes −5‎ →‎CHR Bankswitching ($5120-$5130): Misspelling fixed.
• curprev 14:0314:03, 22 December 2018‎ Bregalad talk contribs‎ 26,734 bytes +401‎ Makes PRG and CHR tables more alike to eachother, add warning about CHR bankswitching being in actual size. Remove note about nonexisting registers.

21 December 2018

• curprev 22:4422:44, 21 December 2018‎ Lidnariq talk contribs‎ 26,333 bytes +148‎ →‎Emulation quicks with commercial games: typo. Explain what "Uncharted Waters" actually needs, because "PRG-RAM banking" is underspecified.
• curprev 22:3922:39, 21 December 2018‎ Lidnariq talk contribs‎ 26,185 bytes −2‎ →‎PRG-RAM bankswitching quircks: typoes
• curprev 22:3822:38, 21 December 2018‎ Lidnariq talk contribs‎ m 26,187 bytes −25‎ →‎PRG Bankswitching ($5113-$5117): grammar, typos, fluidity
• curprev 22:3722:37, 21 December 2018‎ Lidnariq talk contribs‎ 26,212 bytes −65‎ →‎PRG Bankswitching ($5113-$5117): VRC6's two PRG bankswitching registers specify offset as multiples of 16 KiB and 8 KiB. So remove false generalization.
• curprev 21:3821:38, 21 December 2018‎ Bregalad talk contribs‎ 26,277 bytes −3,120‎ →‎PRG Bankswitching: Removes duplicate information, numbers are in hex ($ sign) and stop talking about nonexisting registers $5110-$5112