Page 1 of 1

Reverse engineering the 6522 VIA chip?

Posted: Wed Mar 28, 2018 9:07 pm
by dreamseal
I have some experience (on a hobby basis) in reverse engineering the logic of NMOS silicon die shots. I was wondering if anyone would be interested if I were to make a start on reversing the logic of the 6522 VIA chip from the images on the following web pages:

http://visual6502.org/images/pages/MOS_6522.html

http://siliconpr0n.org/map/mos/6522/dlyr1_mit20x/

The second link is of an image with the metal layer removed, so it shows only the diffusion, polysilicon, and contacts.

Something similar has already been done with the 6502 (see the home page of visual6502.org if you're not familiar with it).

If people are interested, then what I would do is start a new topic in an appropriate section (not sure which section to put it in at this stage) and I would add a post for each small section that I reverse, with an image of that section of the die shot with the layers traced around in different colours for the different materials, and also a logic diagram of the reversed logic for the small section of the die shot shown. John D McMaster made a start on tracing around the layers, but he has since abandoned the effort. I asked him about this and he said he likes to focus on producing die shots rather than doing the reversing itself. He says there are far more people prepared to reverse the logic once a die shot is produced than there are people able to decap and image a silicon chip.

The MOS 6522 VIA chip was used in a number of 8-bit machines, so such a forum topic and series of posts would be of interest to a forum devoted to any of those machines, but due to my new found love of the Oric, I would like to create them here, if people think that the defence-force forums is an appropriate place for it.

If the consensus is yes, then I will add that to my list of Oric-related projects (which seems to be growing each week at the moment). What section would be best to create such a topic in? I was thinking of the "Technical questions" section, but technically its not a question (although perhaps I'd be posing a question of "what's inside the 6522?" and then would set about trying to answer that question). If not the "Technical questions" section, then maybe General is the next best. It is quite a technical and hardware related topic though.

Re: Reverse engineering the 6522 VIA chip?

Posted: Thu Mar 29, 2018 5:56 am
by NekoNoNiaow
My, that would be quite an undertaking, I hope you are assisted with good vision analysis software during this task. ;)

Do not take my opinion as representative of the community in any way (I only infrequently dabble in Oric stuff) but although I find the idea of reverse engineering the 6522 chip fascinating, from an Oric perspective it would make more sense to decap/photograph the chip responsible for video production (I forgot its designation) since this could allow to exploit it in yet unforeseen very practical ways.

For example, if its address generator could be manipulated this would potentially pave the way for double buffering techniques in HIRES mode (one can dream!). ;)

Re: Reverse engineering the 6522 VIA chip?

Posted: Thu Mar 29, 2018 6:41 am
by dreamseal
I quite agree, but unfortunately there isn't yet a die shot of the HCS10017 that I know of. Decaping and imaging isn't something I'm able to do myself. I would quite happily switch to the HCS10017 when a die shot does become available.

Re: Reverse engineering the 6522 VIA chip?

Posted: Thu Mar 29, 2018 6:53 am
by Dbug
dreamseal wrote: Wed Mar 28, 2018 9:07 pm What section would be best to create such a topic in? I was thinking of the "Technical questions" section, but technically its not a question (although perhaps I'd be posing a question of "what's inside the 6522?" and then would set about trying to answer that question). If not the "Technical questions" section, then maybe General is the next best. It is quite a technical and hardware related topic though.
As long as it is in the Oric Hardware subsection of the forum, it does not really matter.
"Technical questions" can also be thought as "Technical Answers To Questions You Did not Even Know You Were Wondering About" :)

Re: Reverse engineering the 6522 VIA chip?

Posted: Sat Jul 14, 2018 9:45 am
by dreamseal
dreamseal wrote: Thu Mar 29, 2018 6:41 am I quite agree, but unfortunately there isn't yet a die shot of the HCS10017 that I know of. Decaping and imaging isn't something I'm able to do myself. I would quite happily switch to the HCS10017 when a die shot does become available.
On the topic of reversing the HCS10017 Oric ULA chip, if someone were thinking of funding an effort to have this chip decapped and photographed, then would the chips available on ebay be the genuine, original chip?

https://www.ebay.co.uk/sch/i.html?_from ... 7&_sacat=0

The descriptions claim that they are New Old Stock, so in theory these chips should be fine, right? They are relatively inexpensive as well. The expensive bit would be the decapping and imaging.

Re: Reverse engineering the 6522 VIA chip?

Posted: Sat Jul 14, 2018 10:33 am
by iss
IMO, all chips on eBay with sellers from Bulgaria are fine. They are remains after Pravetz 8D production was stopped. I don't see any reason someone to put effort and resources to fake exactly Oric's ULA expecting to sell no more than 2-3 pieces :).

EDIT: If needed I can send some pieces "as sacrifice on the altar of science" :D.

Re: Reverse engineering the 6522 VIA chip?

Posted: Sat Jul 14, 2018 11:25 am
by dreamseal
Do we know, did the Pravetz 8D have the same supplier for the ULA as for the French & British built Oric machines? I assume CDI (whoever they were) created the chips for both the original Oric branded and licensed clone machines?

Edit: I recall now already seeing what CDI stands for. It was California Devices Inc, right? @Godzil mentioned that in one post, and a quick Google search confirms this from The Register article of the Oric being 30 years old.

Re: Reverse engineering the 6522 VIA chip?

Posted: Sat Feb 18, 2023 11:02 pm
by dreamseal
For those who have been wondering if I made any progress on the 6522, I didn't end up starting really, as I got sidetracked on to the HCS10017 chip at the time. I did also work on a few other chips unrelated to the Oric after that, such as the SAA5050 (used by the BBC Micro). But the good news is that someone else has now reversed/dissected the 6522. If you weren't aware of it yet, the following forum discussion on 6502.org covers the details:

http://forum.6502.org/viewtopic.php?f=4 ... hilit=6522

Not sure if anyone has looked closely at the schematic that was reversed, but it is now out there for all to see.