+JAC! Posted December 30, 2008 Share Posted December 30, 2008 Hi there, I am looking for the internal schematics of the ANTIC and GTIA. There are scans of them in ftp://ftp.pigwa.net/stuff/collections/nir...20Info/GTIA.PDF but they are not readable at all. Maybe somebody knows where these scan are from or who created them. Best regards, Peter / JAC! 1 Quote Link to comment Share on other sites More sharing options...
Heaven/TQA Posted December 30, 2008 Share Posted December 30, 2008 Hello Peter, finally arrived here... I mentioned several times your Visdom demo... ok...little bit offtopic but is the rastersplit in Visdom 2 done via PM overlays or via DLI midline changes? How did you discovered "HIP"-like mode in 1993? We discovered the "GTIA bug" 3 years later... Quote Link to comment Share on other sites More sharing options...
ijor Posted December 30, 2008 Share Posted December 30, 2008 (edited) I am looking for the internal schematics of the ANTIC and GTIA. There are scans of them in ftp://ftp.pigwa.net/stuff/collections/nir...20Info/GTIA.PDF but they are not readable at all. Maybe somebody knows where these scan are from or who created them. AFAIK there are no publicly available schematics for ANTIC, not even unreadable ones. The GTIA ones are included in the GTIA datasheet. The chipset datasheets were released some years ago by Curt/Atari Museum. I understand they are already unreadable in the original datasheet (in other words, the scans we have are excellent scans of the datasheet, the original datasheet had them bad already). Note that they aren't exactly unreadable. They are difficult to read, and some small portions are missing. But you can "read" most of them if you try hard enough, and it is also possible to "deduce" most of the missing sections. The ANTIC datasheet, at least in the way it was released, it doesn't include schematics. I don't know if they were never included in the datasheet, or they just were missing at some later point (a couple of other pages seem to be missing as well). Somebody mentioned here a few months ago, that MAME team is de-capping some chips. May be we should go that route. I'll gladly contribute for the costs of decapping, at least, ANTIC. But I suspect the costs mentioned wouldn't be nearly the same. Edited December 30, 2008 by ijor Quote Link to comment Share on other sites More sharing options...
Mathy Posted December 30, 2008 Share Posted December 30, 2008 Hello Peter Maybe Tomasz Piorek, developer of the VideoBoard XE, can fill in the gaps. VBXE is like an expanded GTIA. greetings Mathy Quote Link to comment Share on other sites More sharing options...
twh/f2 Posted December 31, 2008 Share Posted December 31, 2008 Hi there, I am looking for the internal schematics of the ANTIC and GTIA. There are scans of them in ftp://ftp.pigwa.net/stuff/collections/nir...20Info/GTIA.PDF but they are not readable at all. Maybe somebody knows where these scan are from or who created them. Best regards, Peter / JAC! Hallo Peter!!! Gut dich hier zu sehen unfortunately I can't say much to the actual topic. however, welcome onboard man!!! grüße!!! \twh Quote Link to comment Share on other sites More sharing options...
+CharlieChaplin Posted January 1, 2009 Share Posted January 1, 2009 Well, I have the following docs about Antic and GTIA: - one Antic PDF doc approx. 2.6 MB in length (most likely the one you mentioned) - some text and html files about Antic - three GTIA PDF docs named a) GTIA.PDF, length 1.1 MB, b) CGIA.PDF, length 4.3 MB and c)FGTIA.PDF, length 2.4 MB - two text files about GTIA Since most PDF docs are too long (even when zipped, guess why) to upload here, I will only attach the text and html files I have... e-mail me if you want all the rest... greetings, Andreas Koch. Quote Link to comment Share on other sites More sharing options...
Heaven/TQA Posted January 1, 2009 Share Posted January 1, 2009 Thanks Andreas for uploading the stuff... I just went through the GTIA PDF and it is interesting that it does not mention the MODE10 delay by 1 "half colourclock"... and it does not mention the PRIOR = 0 status. Quote Link to comment Share on other sites More sharing options...
+JAC! Posted January 1, 2009 Author Share Posted January 1, 2009 Hi everybody, Wow, that's what I call feedback. It's really nice to hear from you all this soon, so I'll do my very best to reply: @Igor: I was afraid somebody would write this. But also I did not expect to learn a new word: "decap" and to find this. Man, these guys are really crazy. Anyway, from what is written there this is not feasible from arbitrary logic but for special ROMs only. Somehomw it reminds me of my favourite slogan. @Heaven: I'll put the "FAQ" reading the Visdom demo(s) in a separate thread if that is OK. I got several requests in the past so maybe that will be interesting for other people, too The funny thing is that yesterday (!) I read by chance an article written by you in the ABBUC Magazin of 1995 in which you mention my first Visdom demo. It's really sad that I didn't have the time to reply back in those days. But now I'll take the chance! @Heaven II: Regarding the GR 10 GTIA bug, I remember that I once read that this is an actual bug that slipped throught the quality control due to a very tight time line back then. I assume it is the result from the additional lookup of the COL/LUM value from the color registers in this case together with logic that combines two ANTIC data packages into one (which is not the case for all non GTIA modes). It was said that the pixels jump back into the proper alignment in many cases if you heat up the chip a little. Well, even in case is is true, I am not goind to be the one who tries it with his own XL @Andreas: Thanks for the docs, but except for the nice timing overview, all this stuff is already present on my harddrive... @Mathy: Do you have a contact address of Tomasz Piorek? Quote Link to comment Share on other sites More sharing options...
Heaven/TQA Posted January 1, 2009 Share Posted January 1, 2009 Peter, you were one of my influences... esp. I love your ABBUC intros... your few releases were all technicly top edge... and even for a "non"-polish scener. I am remember the ABBUC intro with the big vector cube plus the ABBUC letters on the side... Quote Link to comment Share on other sites More sharing options...
+JAC! Posted January 2, 2009 Author Share Posted January 2, 2009 Hello Heaven, "Thanks for the flowers" as we say in Germany (don't know if this idomatic expression can be translated as it is, but I think you where in Germany back then, right?) You mean the ABBUC 34 intro. This is also still one of my favourites because it looks alot like the Amiga intros of that time. I'm planning to put all my stuff at Fandal site, there I'll add a slightly improved version (when you use the SIO patch in an EMU or a fast loader the loading screen can get messed up - or you miss the screen with the 16 shades logos, which is even worse ). BTW I just created the Visdom I & II FAQ topic, there I'll answer your question. Regards, Peter/JAC! Quote Link to comment Share on other sites More sharing options...
+CharlieChaplin Posted January 2, 2009 Share Posted January 2, 2009 Well, as said before I have some more PDF docs for Antic and GTIA - but since there is a 2MB upload size limit at atari-age I have no clue how to upload them. Afaik, PDF`s are already compressed, so zipping will not make a 2.6MB or 4.3MB PDF much shorter... But again, I can send you these PDF`s via e-mail if you want... -Andreas Koch. Quote Link to comment Share on other sites More sharing options...
Rybags Posted January 2, 2009 Share Posted January 2, 2009 Zip can sometimes knock a fair bit off a PDF. I'd also be interested in further Antic/GTIA docs, although I think I've got some of the Patent descriptions. There's a workaround for the 2 meg limit - just use Rar and split to around 1.9 meg sized segments. Problem is that Rar isn't a permitted filetype here, so I just Zip each one with no compression before uploading. Quote Link to comment Share on other sites More sharing options...
ijor Posted January 2, 2009 Share Posted January 2, 2009 But also I did not expect to learn a new word: "decap" and to find this. Man, these guys are really crazy. Anyway, from what is written there this is not feasible from arbitrary logic but for special ROMs only. Of course it is possible for arbitrary logic, it is done all the time, even for multi-layer chips. Only that is more complicated and more expensive. Actually, that web page is contradictory. I know it claims it is possible only for mask ROM. But many of the chips in their "in-progess" list are certainly not mask rom (neither they have mask rom). There seems to be some mistery around all that mame project. I made further googling some time ago. Somebody asked them more details in some forum, and if they could share the person/service doing the decap. They said no, and didn't provide much details. It doesn't matter too much what mame is doing or not. Decap for the Atari chipset is surely possible. I don't know how much it would really cost though. Quote Link to comment Share on other sites More sharing options...
ijor Posted January 2, 2009 Share Posted January 2, 2009 (edited) I'd also be interested in further Antic/GTIA docs, although I think I've got some of the Patent descriptions. I think we all (including Peter) have all the docs that Andrea has. According to the names they are the datasheets for Gtia, Cgia (Antic/Gtia combined) and FGTIA (Secam version of GTIA). Google and you should find them, and I think they are all actually here in Atariage already. It is extremely unlikely that somebody would have some "hot" Atari documentation without knowing. Those that have something not publicly available are probably very well aware, and intentionally don't publish it. As an example, Ken commented sometime ago that he is aware about several people having the MIO firmware sources, none of them published it. There might be a few exceptions, such as Bob recently publishing a couple of sources like DOS 2.5, he surely didn't mean to keep it for himself before. But that was a very rare exception, and he already knew what he had (he just didn't realize it was so "hot"). Edited January 2, 2009 by ijor Quote Link to comment Share on other sites More sharing options...
re-atari Posted January 2, 2009 Share Posted January 2, 2009 But also I did not expect to learn a new word: "decap" and to find this. Man, these guys are really crazy. Anyway, from what is written there this is not feasible from arbitrary logic but for special ROMs only. Decapping is a last resort for duplicating vintage PROMs/PALs/GALs, which (contrary to eproms) can be protected from being read out with a programmer. It requires some professional equipment, and a lot of know-how, time and effort to reverse engineer the function of such an IC, and even then success isn't always guaranteed. When a such a IC goes bad, replacements are hard to find, if at all. Decapping those IC's is basically the only way to definitively get at the logic equations burned into them, well, for a trained professional in computer hardware design, that is... Somehomw it reminds me of my favourite slogan. But (even more shocking), real computer programmers turn out to have humor too! re-atari Quote Link to comment Share on other sites More sharing options...
ijor Posted January 2, 2009 Share Posted January 2, 2009 Decapping is a last resort for duplicating vintage PROMs/PALs/GALs, which (contrary to eproms) can be protected from being read out with a programmer. It requires some professional equipment, and a lot of know-how, time and effort to reverse engineer the function of such an IC, and even then success isn't always guaranteed. I think the main issue is lack of the required equipment. If somebody would make the physical decap and would give us micro-photographs, then sooner or later we'll have reverse engineered schematics. It does require some know-how even then, but it is mostly about time and motivation. We are talking about tiny devices (in relative terms), not about some modern "monster" with transistors in the order of a billion. We also know quite some about those chips already, we don't need to reverse engineer from scratch. We have seen much harder and "bigger" reverse engineer tasks in the Atari scene already. Quote Link to comment Share on other sites More sharing options...
perry_m Posted January 2, 2009 Share Posted January 2, 2009 There was a guy in Hungary who reverse engineered the 6502 from photographs. I mentioned it in another thread: http://www.atariage.com/forums/index.php?showtopic=134099 His site is here, but it's in Hungarian: http://impulzus.sch.bme.hu/6502/letolt.php3 There is a paper describing his methods here, also in Hungarian: http://impulzus.sch.bme.hu/6502/download/tdk_2001.doc Someone did a partial translation of it here: http://www.downloads.reactivemicro.com/Pub...%20Engineering/ I took the paragraphs of the whole paper and ran them through several online machine translators. I could understand the gist of it. I have attached this translation. My strategy would be to first understand the Hungarian 6502 work and then attempt to reverse ANTIC using the knowledge gained, if we can get a photograph. From what I have read, sulfuric acid can decap the IC, and a good optical microscope and camera are all that is needed. It could be done by someone with access to a university lab. I found this lecture useful for understanding the theory behind NMOS logic: hungarian_trans.txt Quote Link to comment Share on other sites More sharing options...
ijor Posted January 2, 2009 Share Posted January 2, 2009 (edited) My strategy would be to first understand the Hungarian 6502 work and then attempt to reverse ANTIC using the knowledge gained, if we can get a photograph.From what I have read, sulfuric acid can decap the IC, and a good optical microscope and camera are all that is needed. It could be done by someone with access to a university lab. Hi Perry, May be we should start an "Atari decap" project, and I mean it seriously. Yeah, possibly somebody at an university lab can do it. Another option could be to contact that Hungarian guy (I guess we should be able to find some atarian on Hungary! ). And the last option would be to actually pay to a decap service, if we can raise the required funds. Btw, did you receive by email about the Pokey counters cycle mismatch? That was about at our last email exchange a couple of months ago, I didn't heard from you since then. Not complaining, just asking Edited January 2, 2009 by ijor Quote Link to comment Share on other sites More sharing options...
carmel_andrews Posted January 3, 2009 Share Posted January 3, 2009 well only solution is, contact George McLoed who designed both C and GTIA (infact the GTIA is named after him) that is if he is still around or still with us Quote Link to comment Share on other sites More sharing options...
ReactiveMicro.com Posted January 3, 2009 Share Posted January 3, 2009 (edited) Hello everyone. I found this thread via a Google Alert. Hope you don't mine me butting in. Let me reach in my pocket and ante up my two cents... There was a guy in Hungary who reverse engineered the 6502 from photographs. Correct. His name is Balazs Beregnyei in Hungary. His work you cited was from his College Theses about NMOS (I believe). I had been in touch with him quite a bit last year. Someone did a partial translation of it here:http://www.downloads.reactivemicro.com/Pub...%20Engineering/ That 'someone' is I. I worked with a friend of Balazs who knew English well and she helps us translate everything. Any questions we had we contacted Balazs to make sure the translation was correct. My strategy would be to first understand the Hungarian 6502 work and then attempt to reverse ANTIC using the knowledge gained. I have had a similar idea with a hobby project much like yours. Since I've done most of the research and work I'll share my thoughts and ideas to help save you guys a bit of time. If we can get a photograph. From what I have read, sulfuric acid can decap the IC, and a good optical microscope and camera are all that is needed. It could be done by someone with access to a university lab. Photographs would be essential to start with. I have a microscope that is used for this type of work and a digital camera attached to it. So pics are no problem. And yes, you will need to decap the IC to take pics of the die. But first I would have a question since I have no idea what IC you're actually dealing with. I assume it's some kind of custom ASIC? If not and it's some kind of PLD (HAL/PAL/GAL/CPLD/uC/etc) then there are services in China that I have used to recover JEDEC a file. It's not always cheap, but it's 99% effective. If it is an ASIC then let's continue. From what I've found it does look like an ASIC to me. Decapping would best be performed by a Company. I have used MEFAS.com in the past. It's like $75, and trust me - they do good work and if you really want to see what's on the die you REALLY need professional equipment to decap. The next issue is layers on an ASIC. I've found that more then three layers deep and you'll need to have at least the top layer removed so you can fully see the next and the one under that too. MEFAS.com does that as well, at about $350 per layer. Next issue would be to fully understand and be able to translate the silicon structures you find. I have yet to master this step. I do have a lot of resources though that will help, but have yet to read them all. If/when this time ever comes, I'd be more then willing to help so I too gain a bit of experience. I also like working with a group. What will really help is since there is that .PDF schematic, someone should be able to mostly verify what's found to the schematic and fill in some of the illegible parts. Of course the end result would be someone needs to build a macro version of what's found, connect it to the existing circuit, and verify what was found is correct. I found this lecture useful for understanding the theory behind NMOS logic: Yes, things like this would be very useful in helping you understand what's what. Of course if/when the time comes I'll post everything I have and we'll go form there. Thoughts/ideas? Henry ReactiveMicro.com Edited January 3, 2009 by ReactiveMicro.com Quote Link to comment Share on other sites More sharing options...
ijor Posted January 3, 2009 Share Posted January 3, 2009 I found this thread via a Google Alert. Hope you don't mine me butting in. Not at all, you are more than welcome. And yes, you will need to decap the IC to take pics of the die. But first I would have a question since I have no idea what IC you're actually dealing with. I assume it's some kind of custom ASIC? Yes, it is a custom ASIC. It uses NMOS technology, and the silicon process should be about the same as the 6502. We don't know the number of transistors, but I'd estimate they are not more than half the number of the 6502 ones. ...there are services in China that I have used to recover JEDEC a file. It's not always cheap, but it's 99% effective. I'm aware about those services, and I understand they can recover much more than a JEDEC. I understand that, if you want, they can perform full reverse engineering of chip (not just decap), or even a full board. The big question is the cost. Let's assume we want them to perform the decap, take micro-photographs (and optionally make a transitor schematics or netlist, and optionally make a gate level netlist). Do you have an idea of which kind of money are we talking aout? His name is Balazs Beregnyei in Hungary. His work you cited was from his College Theses about NMOS (I believe). I had been in touch with him quite a bit last year. Oh, very interesting! And how about asking him for help, or even possibly to do the job (or part of the job), not for free, but as a service? Decapping would best be performed by a Company. I have used MEFAS.com in the past. It's like $75, and trust me - they do good work and if you really want to see what's on the die you REALLY need professional equipment to decap. Again, very interesting piece of information! I understand that from what you are saying, that this cost is only for the physical and chemical decap? Would they optionally take the micro-photographs as well? If so, do you have an idea about the costs? Additionally, how is the whole legal issue handled. I'm not a lawyer, but I understand that before the DMCA this would be fully legal, now I'm not sure anymore. Not that I think the IP/Copyright holder would care. The question is if that company (MEFAS) cares. Do they ask you about what kind of chip it is? If you have rights, papers, authorization, etc? If/when this time ever comes, I'd be more then willing to help so I too gain a bit of experience. I also like working with a group. Well, as I was saying, I'm all for making an Atari Decap Project. I guess there are at least a few of us that would be interested. And certainly we wouldn't mind sharing and getting help from somebody outside the Atari scene. Quote Link to comment Share on other sites More sharing options...
ReactiveMicro.com Posted January 3, 2009 Share Posted January 3, 2009 Hello ijor. Yes, it is a custom ASIC. It uses NMOS technology, and the silicon process should be about the same as the 6502. We don't know the number of transistors, but I'd estimate they are not more than half the number of the 6502 ones. Okay, cool. Thanks for confirming that for me. Assuming it is about half the size of the 6502, then it would take about 2 months of work to recreate the schematic for the die. That is assuming we can learn pretty quickly how to read silicon structures. Of everything out there, NMOS is about the easiest. I'm aware about those services, and I understand they can recover much more than a JEDEC. I understand that, if you want, they can perform full reverse engineering of chip (not just decap), or even a full board. The big question is the cost. Let's assume we want them to perform the decap, take micro-photographs (and optionally make a transitor schematics or netlist, and optionally make a gate level netlist). Do you have an idea of which kind of money are we talking aout? The 'full' service where the do everything for you would run about $30k USD. I know of a place in the States that would do it though. About the same price too without the language barrier. Oh, very interesting! And how about asking him for help, or even possibly to do the job (or part of the job), not for free, but as a service? Well we could ask if we get stumped with something, but he's pretty busy and probably won't be too interested in helping. I've asked for help in the past and recieved about the same response. Really it's all labor from what I've been told. If you know what you're looking at, which is pretty 'easy' I've also been told, then all you need it the time and patience. Again, very interesting piece of information! I understand that from what you are saying, that this cost is only for the physical and chemical decap? Would they optionally take the micro-photographs as well? If so, do you have an idea about the costs? ***Note: If pics don't come up hit Refresh. Hotlink protection from this Host is a little rough. You may even be required to copy the URL in to a new Browser Window. Sorry about that. Well, I had this done: http://www.downloads.reactivemicro.com/Pub...ures/F11242.pdf I also had them take some pics of the fuse array up close: http://www.downloads.reactivemicro.com/Pub...1%20and%202.jpg Other pics of the array here: http://www.downloads.reactivemicro.com/Pub.../HAL%20Pictures Pics of what NMOS looks like: http://www.downloads.reactivemicro.com/Pub...%20fuse%201.jpg That little project was about $350-ish I believe. I'd have to check my records for the actual price though. It was under $400 I'm pretty sure. Anyway, the idea of that project was to recreate the HAL/PAL from the pics. I spent a lot of time trying to trace the pins back in to the array and never did complete it. About half way through the project I found an unsecure PAL that I was able to read. So it really didn't warrant the effort to continue. I did have a lot of fun though and learned quite a bit in the process. Additionally, how is the whole legal issue handled. I'm not a lawyer, but I understand that before the DMCA this would be fully legal, now I'm not sure anymore. Not that I think the IP/Copyright holder would care. The question is if that company (MEFAS) cares. Do they ask you about what kind of chip it is? If you have rights, papers, authorization, etc? Ha! For some reason I get asked this a lot. MEFAS won't care what we do, or ask them to do. It's not illegal to look at any IC or even reverse engineer it. It is illegal to try and copy and sell it if there's a patent. Is doing what we're thinking of doing illegal? No, in short. We're not looking to compete with the owner of the IP. Also, any patent there may have been is well since expired. So how could they sue and show any possible damage? Are we planning on trying to corner the dead 2600 market? Now if we some how manage to sell thousands of units and start making millions, then we may get a letter. But even then they would have a very hard time in court. But in short, for a hobby project that is going to emulate the IC - then what damage is there for a 20+ year old dead IC? Now if we try to clone a Core 2 Duo then we're going to get a call. Firmware has an unlimited copyright, unlike hardware. So in that case there may be an issue, but I can't see anyone asking you to stop selling ROMs from a dead PC. Only Apple has done that, but only with the early Macs. Well, as I was saying, I'm all for making an Atari Decap Project. I guess there are at least a few of us that would be interested. And certainly we wouldn't mind sharing and getting help from somebody outside the Atari scene. All I can offer is the use of my Microscope and some of my time. I'd mainly help out so I could learn a little more about silicon structures. With a little luck you guys will be right there with me and we'll all learn a bit in the whole process. If you want to email MEFAS and ask for pricing then I'll post my contacts info. He'll ask you to send him an IC and then he'll price it for you. If all you want is a decap then it'll be under $100. I assume someone either has a loose IC or a desoldering gun? If not, I could also pull one for you as I have all that type of equipment too. I'd recommend sending a live IC though. The cool thing is when they decap it you can still plug it in and use it afterwards. I always wanted to see what an IC looked like that was damaged from static though. Oh well. Back to work... Henry ReactiveMicro.com Quote Link to comment Share on other sites More sharing options...
ijor Posted January 3, 2009 Share Posted January 3, 2009 Hi Henry, You have PM Quote Link to comment Share on other sites More sharing options...
+CharlieChaplin Posted January 4, 2009 Share Posted January 4, 2009 Ok, I just noticed that the max. single upload space seems to be 10MB now. So here are the PDF`s and other docs I have for various A8 chips... only for those that are interested... there is still a lot of information missing... if someone has some more docs, feel free to upload them here, so I / we can add them to my / our collection... -Andreas Koch. Quote Link to comment Share on other sites More sharing options...
Bryan Posted January 5, 2009 Share Posted January 5, 2009 I can verify that GTIA's timing is buggy. I wrote a program many years ago that created the that Gr. 9/10 HIP effect using only graphics 9. I discovered that by flipping PRIOR a few times during the hblank, I was able to get a 1/2 pixel shifted version of it, so I alternated it on every other line then alternated the lines on the next frame. It looked great! Then I discovered that it didn't work on other XL's. Then I discovered that it worked on some other XL's if I heated up the GTIA. Then I gave up on it. (I didn't know at that time that mode 10 was already shifted) You had to be really careful counting up all the delays when designing LSI logic, and the GTIA was late getting out the door. I'm sure they (probably George McLeod) just ran out of time to fully debug it. Quote Link to comment Share on other sites More sharing options...
Like us on Facebook
Follow us on Twitter
AtariAge on Instagram
Watch us on YouTube
Chat on Discord
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.