Help collecting pictures of CPU and PPU packages
Moderator: Moderators
Re: Help collecting pictures of CPU and PPU packages
Nice to have more RGB PPUs pictured! Do you happen to have any full-chip pictures of the RP2C03B, RP2C04-0001, or RP2C04-0004? They're cropped down on the wiki because the pictured chips have heatsinks on them.
I can supply images of B and D chips with heatspreaders, if those are interesting. I've never seen an A with a heatspreader, and finding a C seems next to impossible (mine pictured on the wiki was bought loose).
I can supply images of B and D chips with heatspreaders, if those are interesting. I've never seen an A with a heatspreader, and finding a C seems next to impossible (mine pictured on the wiki was bought loose).
- Individualised
- Posts: 309
- Joined: Mon Sep 05, 2022 6:46 am
Re: Help collecting pictures of CPU and PPU packages
Fascinating. Now what would be hilarious was if Ricoh parts were remarked as clones.loglow wrote: ↑Sat Sep 24, 2022 9:03 pm
Yes, both genuine (RP) and clone (UA) chips are very commonly re-surfaced and re-marked via Chinese gray market channels.
The re-marking is usually done with laser engraving while the original markings were usually done with white ink printing. I'm not sure if this is a failsafe test because I do have a batch of RP2C02H-0 marked chips that are laser engraved and they do appear to be genuine (non-clone color palette). It's possible they were re-marked anyway though.
I can absolutely confirm that clones are sometimes remarked as genuine. I purchased several dozen chips that were laser engraved with RP2A03E in oddly large text and all of them had the same lot code. They turned out to actually be a variety of UA6527 (12x, clone audio), UA6527P (16x), UA6527P (15x), and also I believe some genuine RP2A03 as well (12x, and with correct audio).
I've also often purchased clone chips that are re-marked clone chips! Again, the laser engraving is usually a giveaway, especially when the other markings on the chip (orientation markers, presence and/or size of circles, bottom markings) are different from each other but the (presumably made up) lot codes are the same. I've received chips marked UA6527 that were actually UA6527P, and even chips marked UA6527P with the 15x-style markings that actually have 16x clock dividers.
This is a bit of an aside, but does anyone know of a ROM that will definitively verify the identity and/or revision of the CPU/PPU that it's running on? I know that 240p Test Suite and Overclock Test can both identify the (divided) CPU clock speed, but I'm looking for an easy way to identify clone/genuine revisions in even more granular detail on otherwise unknown (aka obviously re-marked) chips.
Re: Help collecting pictures of CPU and PPU packages
Arbitrarily, I was assuming the pictures of heatspreaders weren't useful, but I'm open to arguments that they are worth including.
- Individualised
- Posts: 309
- Joined: Mon Sep 05, 2022 6:46 am
Re: Help collecting pictures of CPU and PPU packages
Is there any more context to the origins of this chip/what it exactly is?loglow wrote: ↑Sat Sep 24, 2022 9:11 pm A photo of a "2C02E" chip with a flat aluminum plate laminated to the top. The chip marking "2C02E" is stamped into the aluminum plate and there is no lot code; one corner of the plate is angled to indicate the orientation / pin 1. On the underside there are two somewhat larger-than-average circles molded into the plastic, and it says "JA" inside one of them and "19" inside the other. The chip has been tested and is indeed a working NTSC PPU that outputs a non-clone palette.
Re: Help collecting pictures of CPU and PPU packages
The RC2C03B is the one that theoretically has the slightly-wrong palette. Have you noticed this when you were using it?
Re: Help collecting pictures of CPU and PPU packages
Gosh, I'm almost surprised that I haven't encountered this yet...Individualised wrote: ↑Sun Sep 25, 2022 5:07 amFascinating. Now what would be hilarious was if Ricoh parts were remarked as clones.
Re: Help collecting pictures of CPU and PPU packages
I believe it originally came out of a genuine Famicom, but I'm not certain. Other than it functioning as an NTSC PPU with a non-clone palette, I'm not sure what else to add. I suspect that it's a RP2C02E-0, for lack of a better explanation. I'm not sure if a RP2C02E (non -0) even exists?Individualised wrote: ↑Sun Sep 25, 2022 3:41 pmIs there any more context to the origins of this chip/what it exactly is?
Re: Help collecting pictures of CPU and PPU packages
I don't have any of those, but I do have an RC2C05-03, which is currently missing a photo, so I'll take a picture of that.Fiskbit wrote: ↑Sun Sep 25, 2022 2:17 am Nice to have more RGB PPUs pictured! Do you happen to have any full-chip pictures of the RP2C03B, RP2C04-0001, or RP2C04-0004? They're cropped down on the wiki because the pictured chips have heatsinks on them.
I can supply images of B and D chips with heatspreaders, if those are interesting. I've never seen an A with a heatspreader, and finding a C seems next to impossible (mine pictured on the wiki was bought loose).
I also have a clean RP2C02D-0, so we can have a better photo than the current one that has a weird line through it.
I'm also curious about why an "RP2C03G" is listed on the wiki at all if it "probably does not exist" ...?
Last edited by loglow on Mon Sep 26, 2022 11:45 pm, edited 1 time in total.
Re: Help collecting pictures of CPU and PPU packages
Okay, just to complicate the UA6527P situation a little more, I discovered and tested a couple of these puppies today:
And... they have a /16 clock divider! They are marked with white ink and somewhat worn, so they don't appear to have been re-marked.
So this means that you can NOT necessarily determine the clock divider (/16 vs /15) solely based on the logo, in either direction.
All my other "new" (logo position) UA6527P chips have a suffix after the date (eg. -BS) and they all have /15 dividers.
I'm not sure how unusual these are in general, but I've only encountered two like this out of several hundred UA6527P chips.
And... they have a /16 clock divider! They are marked with white ink and somewhat worn, so they don't appear to have been re-marked.
So this means that you can NOT necessarily determine the clock divider (/16 vs /15) solely based on the logo, in either direction.
All my other "new" (logo position) UA6527P chips have a suffix after the date (eg. -BS) and they all have /15 dividers.
I'm not sure how unusual these are in general, but I've only encountered two like this out of several hundred UA6527P chips.
Re: Help collecting pictures of CPU and PPU packages
I'll check and take some photos with a palette generator for you.lidnariq wrote: ↑Mon Sep 26, 2022 11:13 amThe RC2C03B is the one that theoretically has the slightly-wrong palette. Have you noticed this when you were using it?
Re: Help collecting pictures of CPU and PPU packages
The line on the RP2C02D-0 is from the Famicom eject mechanism. I have a couple more D-0's somewhere that are likely in better shape, but I was sloppy and didn't add the console serials to my spreadsheet and have an unreasonable number of Famicoms they could be in. If you can provide a better photo, fantastic!loglow wrote: ↑Mon Sep 26, 2022 11:12 pm I don't have any of those, but I do have an RC2C05-03, which is currently missing a photo, so I'll take a picture of that.
I also have a clean RP2C02D-0, so we can have a better photo than the current one that has a weird line through it.
I'm also curious about why an "RP2C03G" is listed on the wiki at all if it "probably does not exist" ...?
I've removed the RP2C03G from the wiki. I think this may have been a typo by kevtris long ago. No one seems to have seen this chip and the few mentions online all seem to be from the same one or two sources.
Re: Help collecting pictures of CPU and PPU packages
I wonder if it's the revision instead. So far we've seen S, BS, and BA, and yours photographed above has no letter at all.loglow wrote: ↑Mon Sep 26, 2022 11:30 pm Okay, just to complicate the UA6527P situation a little more, I discovered and tested a couple of these puppies today:
[image of DIP40 with:
UMC
UA6527P
9032
626420]
And... they have a /16 clock divider! They are marked with white ink and somewhat worn, so they don't appear to have been re-marked.
[...]
All my other "new" (logo position) UA6527P chips have a suffix after the date (eg. -BS) and they all have /15 dividers.
Re: Help collecting pictures of CPU and PPU packages
Ok, so here's my catalog of all the variations that I've seen and tested. Make sure to click the UA6527P tab at the bottom.
https://docs.google.com/spreadsheets/d/ ... C0BEb4VNkw
Conclusions:
- Chips made in 1989 and earlier generally have the old logo and a /16 divider (no exceptions observed).
- Chips made in 1991 and later generally have the new logo and a /15 divider (no exceptions observed).
- Chips made in 1990 can't be determined by date alone.
- Revisions without a dash (eg. S, AC, BA, or no letter) generally have a /16 divider (no exceptions observed).
- Revisions with a dash (eg. -BS, -BT, -AS, -BA, -BG, -BM) generally have a /15 divider (no exceptions observed).
This was the sole -AS revision chip, which has a slightly different surface finish and printing quality on it:
Here's one of the two "no logo" chips:
I'll reply with photos of the other one because of an apparent attachment limit.
Re: Help collecting pictures of CPU and PPU packages
Here's the other "no logo" chip: