What is happening here?

Discuss anything related to the VIC
matsondawson
The Most Noble Order of Denial
Posts: 343
Joined: Fri May 01, 2009 4:44 pm

Post by matsondawson »

Would it be possible to add an adapter to the cart port that exposes the VIA bus on via cycles for RAM1,2,3? I'm guessing you'd have to decode the RAM1,2,3 for the VIA cycles, and block any chips selects/RW coming from the CPU. Maybe about 5 ics worth?

Edit:
Wouldn't require any soldering then.
Of course that wouldn't give you all that extra colour ram... or would it?
User avatar
ral-clan
plays wooden flutes
Posts: 3702
Joined: Thu Jan 26, 2006 2:01 pm
Location: Canada

Post by ral-clan »

Occasionally I will view these amazing new VIC images on the internet and marvel....so your work is certainly appreciated, even if I haven't implemented it at the moment.
User avatar
Mike
Herr VC
Posts: 4816
Joined: Wed Dec 01, 2004 1:57 pm
Location: Munich, Germany
Occupation: electrical engineer

Post by Mike »

@Muzz73: In the corresponding thread in the Hardware section I posted some photos, where VFLI images are on display on a monitor. The newer posts feature examples which were rescaled from the test output of the converter to match the pixel aspect ratio correctly.

@16KVIC20 and @matsondawson: The RAM data and address lines which are directly connected to the VIC chip are not exposed to the cartridge port. There are three 8-bit buffers acting as timing barrier inbetween. These buffers separate CPU bus and VIC bus during VIC fetches. The colour RAM even gets separated from the normal (VIC) data bus with yet another analogue switch during that time.

When you want to make the 3K extra RAM accessible to the VIC chip, you must add it at the correct side of the buffers, and that is only possible on the mainboard.

@PaulQ and @rhurst: I also hope that others always remember that everyone here in Denial can contribute to the community. :)

@RobertBe: While the mod might also serve other uses (like expanding to a contiguous 30.5 K BASIC memory, i.e. 31231 bytes free), you see there's an NTSC 6560 depicted in the photo - a corresponding display routine is in the works. ;)

@ral-clan: That being said, you surely can imagine the anticipation, when the new display system was fired up the first time and hardware and software worked right together on the first try. :)
matsondawson
The Most Noble Order of Denial
Posts: 343
Joined: Fri May 01, 2009 4:44 pm

Post by matsondawson »

> When you want to make the 3K extra RAM accessible to the VIC chip, you must add it at the correct side of the buffers, and that is only possible on the mainboard.

The plan was to have an adapter you plug on the main board that also adapts to the cart slot. i.e. you plug onto the VIC side of the bus and the cart slot. You'd still have to open your machine, but you wouldn't have to solder it.
User avatar
Mike
Herr VC
Posts: 4816
Joined: Wed Dec 01, 2004 1:57 pm
Location: Munich, Germany
Occupation: electrical engineer

Post by Mike »

The internal RAM expansion does not only have to add connections (which could easily be done with some plug-in board contacting chip pins from above, or similar), it also needs to break connections on the mainboard.

My variant does this without the need to cut traces on the mainboard itself, rather the relevant chips are put into sockets - and where a connection has to be rerouted, the relevant pin of the chip is bent straight beforehand, and the connection is rewired.

TLovskog's variant does not replace the original chip select logic and RAMs, rather there's a PCB which plugs into the character ROM. Not all VIC-20's have this chip socketed, and it also requires to break traces on the mainboard and solder wires to other chip pins, so it's no better off.

As I wrote earlier, CPU bus and VIC bus are separated during VIC fetches. This is necessary, because the 6502 can't tri-state its address bus. When the bus buffers are enabled, the 6502 selects the direction of the data bus buffer over its CR/W line, and the address bus buffers are fixed in the direction CA -> VA. Even though the /RAMx signal lines actually reflect VIC fetches in the range $0400 .. $0FFF, that doesn't buy you much, the complete VIC address never reaches the cartridge port.

In my variant I especially took care, that the (now open) original /RAMx signals on the mainboard are tied to H-level. This is necessary for reliable operation of UD9 (which decides, whether CPU bus and VIC bus are to be separated) and to reliably disable external RAM expansions in that range. Otherwise, the open inputs of UD9 put a signal level of around 2.4 volts on the /RAMx lines, which is undefined for CMOS inputs!

I don't see any solution which implements this mod without some soldering work on the mainboard.
Vic20-Ian
Vic 20 Scientist
Posts: 1213
Joined: Sun Aug 24, 2008 1:58 pm

Post by Vic20-Ian »

Mike,

from my point of view this is a variation of inventor not understanding the market problem.

You have invented something and think it is wonderful. It solves all your display problems or needs etc.

I don't have the same problem or need therefore while I can appreciate it as a technical achievement, it has no commercial value to me.

By commercial value I mean I see no need to invest my time or any money in doing this as I would not use my Vic-20 to do this, much as I enjoyed seeing the demo pictures and was amazed.

I use my Vic-20 as a Vic-20, to play games and a bit of programming with it's delightful standard resolution.

I love my Simple, Friendly, Low resolution Vic-20 as it is and much more than I could love an internally hardware modified, Franken Vic.

This may seem at odds with my support for the Final Expansion 3. I use this as a multi rampack and for quick fix on favourite games really it is overkill for what I need but it is convenient to have the Ram and the SD2iec drive.

I appreciate your work and congratulate you but I just don't want to buy into it myself.

Keep going, don't stop just because we don't want to do it.
Vic20-Ian

The best things in life are Vic-20

Upgrade all new gadgets and mobiles to 3583 Bytes Free today! Ready
PaulQ
undead vic
Posts: 1967
Joined: Sun Jan 14, 2007 2:57 pm

Post by PaulQ »

Mike wrote:@PaulQ and @rhurst: I also hope that others always remember that everyone here in Denial can contribute to the community. :)
The simple expenses of money, time, and space it takes to acquire, maintain, and occasionally use a working Vic 20 computer in 2012 represents a contribution that most people would be unwilling to make.
rhurst
Omega Star Commander
Posts: 1369
Joined: Thu Jan 31, 2008 2:12 pm
Website: https://robert.hurst-ri.us
Location: Providence, RI
Occupation: Tech & Innovation

Post by rhurst »

... that most people ...
Meaning "regular" people, not us freaking geeks! :P
Any technology distinguishable from magic is insufficiently advanced.
https://robert.hurst-ri.us/rob/retrocomputing
User avatar
ral-clan
plays wooden flutes
Posts: 3702
Joined: Thu Jan 26, 2006 2:01 pm
Location: Canada

Post by ral-clan »

Mike,

You should consider doing a write-up about your new graphics mode and hardware modifications for Commodore Free magazine. A lot of people (more than on Denial, anyway) would read about it, be impressed by the images and perhaps try it.

Commodore Free magazine is always looking for articles, and your graphics mode would be a "meatier" and more significant article than they many reviews of old games that are often needed fill that magazine.

http://www.commodorefree.com/
Image Music I've made with 1980s electronics, synths and other retro-instruments: http://theovoids.bandcamp.com
User avatar
Mike
Herr VC
Posts: 4816
Joined: Wed Dec 01, 2004 1:57 pm
Location: Munich, Germany
Occupation: electrical engineer

Post by Mike »

Hi,
Vic20-Ian wrote:You have invented something and think it is wonderful. It solves all your display problems or needs etc. [...]
I think the display system is the main feedback and the most visible aspect a computer can show to the user, sound being second to that. The VIC-20 was designed around the VIC chip, and the charme of my mod lies, IMO, in that the VIC chip is kept, just some restrictions on the mainboard have been removed. :)

The VFLI mode was a natural follow-up on the experiments, that tokra and I conducted in 2010 which ultimately ended up in MAXIGRAFIK and the original FLI mode with 72x256 pixels. Tokra then decided to keep himself on the software route, while I wanted to combine the resolution of MAXIGRAFIK (208x256 pixels) with the colour resolution of FLI (8x1 pixel attributes, new background/border and auxiliary colour each raster). Without much thinking it was clear to me this could not be done with software alone.

A converter for the 72x256 FLI mode was already there, it was simply extended to VFLI's 208 pixels. The design of the display routine was inspired by another stable raster routine I had written a year earlier.

The internal 3K expansion had already been done by other people, wimoos for example also has such a modded VIC-20, and here's a heavily modded VIC-20. ;) The replacement of the colour RAM was new, though.

Tea can be brewn in different strengths. Likewise for the (internally) unmodified VIC-20, there are still MINIGRAFIK, MINIPAINT and the assorted tools available, which nicely work with both PAL and NTSC and just require an extra +8K RAM expansion as minimum. If that's more your cup of tea, fine. :)

And surely noone is going to blame you for your engagement with FE3. I remember you offered others to build a FE3 for them, and I guess that helped a lot people to get a FE3 who would otherwise have been deterred by its kit form.
ral-clan wrote:You should consider doing a write-up about your new graphics mode and hardware modifications for Commodore Free magazine.
That could be a nice follow-up to my contribution in Issue #38 indeed. At the end of the interview I was asked what I would have changed if I had the possibility to travel back in time and help redesign the VIC-20. Well, the error-free VIA is still somewhat beyond my reach, ;) but the 3K gap is now filled and the S-Video mod is also implemented in my VIC-20. :)

I'd want to implement the 2-prong variant hardware wise and NTSC display routine software wise first, though.
Last edited by Mike on Sat Dec 15, 2012 7:00 am, edited 1 time in total.
User avatar
RobertBe
Vic 20 Elite
Posts: 2304
Joined: Sat Jul 14, 2007 2:48 pm

Post by RobertBe »

Mike wrote:...the 3K gap is now filled and the S-Video mod is also implemented in my VIC-20. :)
Mike, am I wrong or does the S-video mod work also on a NTSC VIC-20?

Truly,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
User avatar
Mike
Herr VC
Posts: 4816
Joined: Wed Dec 01, 2004 1:57 pm
Location: Munich, Germany
Occupation: electrical engineer

Post by Mike »

Yes. ;)

The S-Video mod also works on a NTSC VIC-20. You will have to check your mainboard revision though, as the video circuitry is different between 2-prong and DIN models (same also applies to PAL).
User avatar
RobertBe
Vic 20 Elite
Posts: 2304
Joined: Sat Jul 14, 2007 2:48 pm

Post by RobertBe »

Thanks. I have both types of VIC-20s.

Truly,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
lordbubsy
Vic 20 Amateur
Posts: 45
Joined: Fri Nov 18, 2011 12:31 pm

Post by lordbubsy »

Mike wrote:Half a year ago, Eslapion donated a defective NTSC VIC-20 to tokra, who in turn sent it to me. In the last days, I found the time to repair it, and while being at it, implemented the first part of the mod as well. Incidentally, it is also a 2-prong VIC-20. The other two combinations, PAL 2-prong and NTSC DIN are equally possible. Expect results at the beginning of next year - here is a view of the 'proof of concept' work:
I had a defective PAL 2-prong board, which had to have a memory problem. (garbled screen) But those 11 18-pin RAM chips, really held me back repairing this mobo.
After reading this thread, I asked Mike for the wiring diagram.

So I went on de-soldering. It relay isn’t easy, and if you don’t care for those chips, you better cut the pins and remove them one by one.

As a substitute I used an "NEC D4364C 15L" RAM which is identical to the 6264.
After switching on, the garbled screen was gone, whoehoe! But there was no welcome screen either. Only a cyan border and white background. That could be a defective ROM. After replacing the BASIC ROM I had success!
I have "6655 BYTES FREE" :-)
So I can start to build a PAL VC-20.

A way to easily test 2114 RAM chips.

To test those eleven 2114 RAM chips, you can place them one by one in socket UE1, which is the color RAM.
I found out that three of them were broken.

To continue and do the VIC-20 extended color ram modification, I have to do another mod on my Commodore 128. The C128 has two 4416 (16K X 4) RAM chips as VDC video RAM. This can be expanded by replacing them with two 4464 (64K X 4) RAM chips. The two 4416 chips are exactly what I need for the VIC-20 extended color ram! Well, one of them.

It can take a while , but I’ll certainly pursue this.
Thanks community!

PS, I have no photos (yet), because it basically looks the same as in the initial post, only less neat. :oops:
8-bit assimilation
User avatar
Mike
Herr VC
Posts: 4816
Joined: Wed Dec 01, 2004 1:57 pm
Location: Munich, Germany
Occupation: electrical engineer

Post by Mike »

lordbubsy wrote:The two 4416 chips are exactly what I need for the VIC-20 extended color ram!
NO!!!

The 4416s are DRAMs, they have a completely different pinout and method to signal them.

What you need is a 16Kx4 SRAM, like the CY7C164 I used. Alternatives are the Hitachi HM6788P-35 and Intel P51C98-45. See this post for available BINs at ebay.
Post Reply