Question of commodore users re the vic chip and other things

[carmel_andrews]

Mainly relating to the vic chip, according to the famous commodore history book 'home computer wars' The chip was finished early 1978 (which puts it ahead of the Atari 8bit chipset) but tramiel made a booboo by getting MOS (later Commodore Semiconductor Group/GSG) to try selling/flogging the chip to rival games/computer hardware companies

 

It wasn't till late 1979/early 1980 when tramiel decided to get commodore using it's own technology into it's own computers

 

Any reason why commodore didn't try trumping Atari (and the A8 chip set) and getting a machine with this VIC chip tech out in 1978 (bearing in mind that the first CBM system that was going to use the VIC chip was some updated version of the Pet system (project named T O I)

 

Looking at the VIC chip itself, was it basically a commodore take on Atari's T I A chip (i.e a combined sound/graphics chip)

 

Did commodore do two versions of the vic chip, i.e a combined sound/graphics chip (as used in the vic 20/vc 20 and vic 1001) and also just a graphics chip

 

Lastly is it true that the launch of the sinclair zx80/zx81 finally convinced tramiel and some of the commodore management that commodore should be aiming at the low end of the computer market (i.e why it went for the vic system), namely the sub 500 usd/ 300 usd computer market, the above mentioned book seems to infer and imply that the sinclair zx80 is what got commodore entering the low end computer market

[kool kitty89]

Mainly relating to the vic chip, according to the famous commodore history book 'home computer wars' The chip was finished early 1978 (which puts it ahead of the Atari 8bit chipset) but tramiel made a booboo by getting MOS (later Commodore Semiconductor Group/GSG) to try selling/flogging the chip to rival games/computer hardware companies

 

It wasn't till late 1979/early 1980 when tramiel decided to get commodore using it's own technology into it's own computers

 

Any reason why commodore didn't try trumping Atari (and the A8 chip set) and getting a machine with this VIC chip tech out in 1978 (bearing in mind that the first CBM system that was going to use the VIC chip was some updated version of the Pet system (project named T O I)

From what I've read, the VIC itself was never specifically aimed at a color PET machine, and regardless the VIC-20 broke significantly from the PET's market with a push for general mass market accessibility for average consumers. The canceled 6564 was the chip intended for TOI and 6562 intended for the Color PET. (both more ambitious designs including 80 column text in the 6564)

Both required fast SRAM (DRAM was too slow at the time) which was too expensive to implement in any significant amount for the time. The VIC-20 did end up using only SRAM onboard, but that was almost 3 years later and I believe it was slower SRAM as well. (plus Commodore had apparently amassed and overstock of the 512kB 4-bit SRAM chips at the time which made an attractive option for the VIC-20)

 

As for the VIC chip itself, it appears that the late 70s 6560 incarnation had more limited features than the revised chip implemented in the VIC-20. If nothing else it expanded the color palette and enhanced the sound. (I'm not sure on the details, but if I had to guess: it seems likely that the primary 8 colors were the original set though I'm not sure what the sound would have been)

 

Looking at the VIC chip itself, was it basically a commodore take on Atari's T I A chip (i.e a combined sound/graphics chip)

Not really other than the very vaguest sense. The VIC is a lot more than TIA and a lot different: TIA is very simple and limited and requires a lot of CPU overhead and precise timing to generate the display (racing the beam) though it does have hardware sprites and a rather large color palette. The VIC has actual hardware character graphics modes and generates the display without the CPU intervention required by TIA, though it lacks hardware sprites and has fewer colors available to select from.

 

It might have been useful, but it's hard to tell... if they were aiming at the same goal as the VIC-20 bug 2 years earlier, it would have been even more cut-down, both with the older VIC chip and likely even less onboard RAM. It would depend on how BASIC was set-up among other things (if built-in and using 1.5 kB automatically at boot, that could be pretty limiting, but they could have probably managed that differently too), or more specifically how much RAM was in the system and how much available for use. If only SRAM could be used, that would be a major factor too as DRAM could have been a major cost savings. If they had to go with SRAM for a 1979 machine and as low-cost as possible, it likely wouldn't have had more than 2 kB onboard (4 kB would have been pushing it), though with DRAM it probably could have afforded 8 kB pretty reasonably, perhaps 4 kB to push for even lower cost. (you have other added component costs for DRAM use though) And CBM did already have the advantage of vertical integration and relatively simple chipset compared to the A8 (or the non integrated hardware of the Apple II) The general capabilities would be far less than the contemporary Atari 8-bit line, though arguably better than the Apple II (depending on how much the original VIC was missing from the 1981 incarnation -even with 8 colors and 1 channel sound it would have had some advantages if the general display modes were comparable) and with the simple chipset and vertical integration of manufacturing, it likely could have undercut the 400's price point by a good margin. Though with the case of 2 kB of RAM, that would make tape based software pretty impractical and even 4 kB would be tight for that (let alone if 1.5 kB were monopolized by BASIC), but carts were in common use that anyway, though a good deal more expensive than cassettes. (in the case of the VIC-20 much of the game library in particular was carts anyway though)

 

Did commodore do two versions of the vic chip, i.e a combined sound/graphics chip (as used in the vic 20/vc 20 and vic 1001) and also just a graphics chip

I need to look into it more but it seems there were 3 chips contributing to the design: the original 1977 6560 VIC chip, the 6562 intended for the color PET, and the 6564 intended for TIO with the original VIC being the primary basis for the 1981 chip but with some enhancements in terms of sound circuitry derived from the 6564 and added color from the 6562. (I'd gotten the impression that the original VIC at least had some form of sound generation)

 

Lastly is it true that the launch of the sinclair zx80/zx81 finally convinced tramiel and some of the commodore management that commodore should be aiming at the low end of the computer market (i.e why it went for the vic system), namely the sub 500 usd/ 300 usd computer market, the above mentioned book seems to infer and imply that the sinclair zx80 is what got commodore entering the low end computer market

Perhaps, though I'd gotten the impression that fears over Japanese developments also contributed to that. The ZX80/81/TS-1000 were all TOO low-end to really be useful, and a compromise between what atari did with the 400 and something on the level of the Sinclair machines would have fit much better. If they tried to push that back in '79/80, such a machine would likely have been more like $400 (maybe closer to 300 depending just how cut back it was), so a good deal cheaper than the 400, but still far more useful than the Sinclair machines. (particularly for games)

[Mayhem]

The rhetoric you will hear generally is that Tramiel was worried about low cost machines from Japan hurting his computer business in the same way the calculator market had been affected. I believe the Vic20 was already in development before the ZX80 was available, it launched in Japan in September 1980 and was deliberately targeted at being a low cost, entry level home computer.

[carmel_andrews]

Sorry, made a booboo, it should have read 'question FOR commodore users' not question OF

Edited October 3, 2010 by carmel_andrews

[kool kitty89]

The rhetoric you will hear generally is that Tramiel was worried about low cost machines from Japan hurting his computer business in the same way the calculator market had been affected. I believe the Vic20 was already in development before the ZX80 was available, it launched in Japan in September 1980 and was deliberately targeted at being a low cost, entry level home computer.

 

Yep, not just Calculators either but also Adding Machines and Typewriters before that. So Tramiel had more than a little warning over the Japanese.

 

It is a bit ironic that the C64 in some ways facilitated the Japanese dominating the video game market in the sense of the C64 catalyzing the North American crash and related shift of several manufacturers to computers (failed attempts by Coleco and Mattel -albeit Coleco may have had a chance if that full system bundle hadn't been the only form factor -ie had a bare bones console version of the Adam been available at prices closer to the C64 as well as not making the printer and possibly dual tape drives integral to the desktop form factor as well, but the Aquarius didn't have a chance and Atari had good hardware with established software in their computers but they had a lot of other problems...).

 

Commodore also seemed to manage their entrance into the European market a good deal better than other US companies.

Edited October 4, 2010 by kool kitty89

[supercat]

I wonder how the chip-enable speed of SRAMs compared with the random-access speed? I wonder if it would have been practical to wire a machine such that the LSB of the address selected between two or four RAM chips, and arrange to access both of them one after the other in the same cycle? That would seem to ease the timing requirements.

[kool kitty89]

I wonder how the chip-enable speed of SRAMs compared with the random-access speed? I wonder if it would have been practical to wire a machine such that the LSB of the address selected between two or four RAM chips, and arrange to access both of them one after the other in the same cycle? That would seem to ease the timing requirements.

I know it was mentioned that the 6562 and 6564 both required high-speed SRAM, but did the 6560 VIC require SRAM or was SRAM used due to availability and related similicity/cost reasons? (CBM had been using SRAM in all the early model PET computers iirc which is where the surplus used for the VIC came from I'd think) It would likely be slower SRAM too, not particularly high-speed for the time, albeit obviously more expensive than comparable DRAM chips and without other timing/latency issues. (but simpler to interface and using less board space unless the circuitry for DRAM control was consolidated into another chip on the board -so cheaper in that respect, especially with an existing stockpile available and the relatively small amount of RAM used onboard the VIC-20) Hmm, actually since the VIC-20 used only 512 kB 4-bit SRAM chips (10 of them) that would actually be a significant amount of board space going to those chips especially given the higher pin count than DRAM (non-multiplexed address lines), so that stockpile of SRAM chips really must have been a huge factor if the VIC could work with DRAM.

 

Did the VIC-20 RAM expansions use SRAM or DRAM?

[Chris++]

I'm nearly positive that they used SRAM, because the chips merely added a straight 64k (to replace the prior 16k).

[supercat]

I know it was mentioned that the 6562 and 6564 both required high-speed SRAM, but did the 6560 VIC require SRAM or was SRAM used due to availability and related similicity/cost reasons?

 

DRAM requires more complicated circuitry to address. The VIC-20 used ten 1Kx4 SRAM chips for main memory, plus one for color data.

 

Did the VIC-20 RAM expansions use SRAM or DRAM?

 

The VIC-20 output chip-select signals for three 1K blocks from $0400 to $0FFF, and for 8K blocks at and above $2000. This would allow convenient use of six 1Kx4 SRAMs for a 3K expander, or one 8Kx8 SRAM for each 8K in larger expanders.

[kool kitty89]

I'm nearly positive that they used SRAM, because the chips merely added a straight 64k (to replace the prior 16k).

Huh? 16k and 64k... do you mean for the RAM expansion on the VIC?

 

 

 

 

I know it was mentioned that the 6562 and 6564 both required high-speed SRAM, but did the 6560 VIC require SRAM or was SRAM used due to availability and related similicity/cost reasons?

 

DRAM requires more complicated circuitry to address. The VIC-20 used ten 1Kx4 SRAM chips for main memory, plus one for color data.

Yes, so that would be a factor in terms of cost/complexity and/or development time/cost. (in the latter case that would include the potential for integrating DRAM interface/refresh circuitry into the VIC chip itself or another custom chip used to save board space)

 

Didn't the Atari 8-bits (and 5200), Apple II, TRS-80, CoCo, and C64 all use DRAM exclusively?

Edited October 4, 2010 by kool kitty89

[Rybags]

All the Ataris used DRAM. Exceptions to that would be some 3rd party expansions and stuff like RAM carts.

 

I didn't know VIC used SRAM... wasn't it like 3 times the price or something at that time?

[kool kitty89]

All the Ataris used DRAM. Exceptions to that would be some 3rd party expansions and stuff like RAM carts.

 

I didn't know VIC used SRAM... wasn't it like 3 times the price or something at that time?

 

It was my impression that SRAM was more like 8 to 10 times the cost of contemporary DRAM (more silicon, lower production volume, and more pins per chip -double the address lines as DRAM uses multiplexing), but that's for sheer component costs of the RAM chips themselves so there's more to it than that. There's the added circuitry needed for DRAM (control and refresh timing) which adds more component cost, board space, and development time. (mainly short-term factors, with DRAM paying off a lot more in the long run -especially with the control/refresh circuitry integrated into a single chip or even consolidated into another custom chip on the board)

There's also the issue of general performance of SRAM (in terms of speed and access latency) compared to contemporary DRAM making it the only option for some applications, especially back then -vs later high-performance DRAM based options. (the really fast/high performance SRAM chips would be even more expensive too)

Those reasons made SRAM popular in game consoles and embedded systems (simplicity and performance), though in some cases the use is a bit odd as added circuitry should have been fairly minimal for DRAM and added performance wasn't needed. (like the main RAM in the ColecoVision, SG-1000, or Master System -opposed to the MSX's DRAM and especially given the VDP already is using DRAM and the Z80 supports DRAM refresh)

 

In the VIC's case you had the existing supply of SRAM chips commodore had, then the plans as a relatively short-term product, and especially emphasis on short development time to get it on the market as soon as possible. (the color RAM would almost certainly be SRAM in either case though)

Edited October 5, 2010 by kool kitty89

[Chris++]

Granted, the 64k expansion was only available from a third-party company. That went for the 32k cart as well, for that matter. If memory serves (ha), the 64k cartridge came with software, so that the user could determine which section of the expanded RAM he wanted to enable. This feature was included because some software wouldn't work in tandem with the expansion. (The 32k cart simply had a switch on it.)

[supercat]

The Colecovision's VDP included DRAM control circuitry for its own DRAM. To use DRAM for the main memory would have required entirely separate DRAM control circuitry. The "refresh counter" on the Z80 saves a little bit of DRAM-control logic, but using a DRAM will still require a fair amount of extra circuitry. Hooking an SRAM up to a processor circuit often requires only one or two gates; using DRAM will require an address multiplexer along with some sequencing circuitry. Certainly not impossible, but not particularly easy.

[kool kitty89]

The Colecovision's VDP included DRAM control circuitry for its own DRAM. To use DRAM for the main memory would have required entirely separate DRAM control circuitry. The "refresh counter" on the Z80 saves a little bit of DRAM-control logic, but using a DRAM will still require a fair amount of extra circuitry. Hooking an SRAM up to a processor circuit often requires only one or two gates; using DRAM will require an address multiplexer along with some sequencing circuitry. Certainly not impossible, but not particularly easy.

 

Yes, so it would be a trade-off in board space/complexity, development time/resources (for circuitry consolidated into a single chip -or even added to another custom chip), and using much less RAM. (if speed wasn't an issue to exclude DRAM that is) If you could settle for a small enough amount of RAM, the advantages of DRAM would decrease significantly. (in the VIC's case I'd guess it was mainly due to time constraints and availability, as even with full non-consolidated DRAM support circuitry there could be a big cost savings switching to 2 kB DRAM chips, perhaps enough to even allow 8x 14kx1-bit chips to be used and still save cost, but using 4kx4-bit chips for 8kB would make sense too and still be 8k to the VIC's 5)

Albeit, from that perspective, going to higher density SRAM chips would also be advantageous if still more expensive. (ie using 4kx4-bit chips)

[carmel_andrews]

Re: RAM issue, the reason why CBM used the ram it did in the vic 20 is because MOS (later CSG) had a bloated inventory of certain types of memory left over from their calculator days, so they used it in the vic 20/vc 20 and vic 1001 (dunno about their ram expansions though)

 

Source...look at the book i referred to in the orig. post

[kool kitty89]

Re: RAM issue, the reason why CBM used the ram it did in the vic 20 is because MOS (later CSG) had a bloated inventory of certain types of memory left over from their calculator days, so they used it in the vic 20/vc 20 and vic 1001 (dunno about their ram expansions though)

 

Source...look at the book i referred to in the orig. post

Hmm, so MOS/CBM had been using 1kx4-bit SRAM chips in their calculator chipsets?

 

I'd thought it also had to do with SRAM being used in the PET machines. (unless that was also due to the calculator overstock)

[carmel_andrews]

http://www.bombjack.org/commodore/books.htm

 

click on the one 'the home computer wars' to download, mr KK, it's all in there

 

I did hear something that they designed some parts of the pet around some of their calculator designs

 

As well as the TOI, the first commodore machine (pre vic 20) to using the vic chip

[kool kitty89]

http://www.bombjack.org/commodore/books.htm

 

click on the one 'the home computer wars' to download, mr KK, it's all in there

 

I did hear something that they designed some parts of the pet around some of their calculator designs

 

As well as the TOI, the first commodore machine (pre vic 20) to using the vic chip

 

Umm, I don't see "the home computer wars" anywhere on that page.

[carmel_andrews]

sorry wrong section

 

 

http://www.bombjack.org/commodore/books-generic.htm