Hydra (RAM) cartridge?
Ym2413a
Posts: 630
Hey does anyone know when or if there will be a cartridge upgrade for the Hydra that has some external RAM?
I'm thinking maybe something fast enough to be used as Vram.
This would greatly free up a lot of HUB-memory if you could store tiles in a external Sram and only have to keep things like sprites and overlays in the HUB-RAM.
A bitmap or tile layer could be stored in such a large area and this would free up much memory in our 32k limit while providing more tile textures!
I'm sort of surprised the Hydra didn't have some sort of Vram on the board.
Maybe because of cost or another limit I'm not seeing at the moment...
I know about the pin-count problem and was thinking that latchs could be used to help save pins in such a design.
--> Example
Latch ADDRESS high, latch ADDRESS low. Read/Write! and Repeat process.
--Andrew Arsenault
I'm thinking maybe something fast enough to be used as Vram.
This would greatly free up a lot of HUB-memory if you could store tiles in a external Sram and only have to keep things like sprites and overlays in the HUB-RAM.
A bitmap or tile layer could be stored in such a large area and this would free up much memory in our 32k limit while providing more tile textures!
I'm sort of surprised the Hydra didn't have some sort of Vram on the board.
Maybe because of cost or another limit I'm not seeing at the moment...
I know about the pin-count problem and was thinking that latchs could be used to help save pins in such a design.
--> Example
Latch ADDRESS high, latch ADDRESS low. Read/Write! and Repeat process.
--Andrew Arsenault
Comments
Could you explain this in a little more detail? I'm gradually learning about EEPROM, as the Hydra bood explains a little about it, but I don't understand SRAM or latching.
Jim C
I know, but I sort of don't want to do anything that is outside of the Hydra standard.
I could mod the hack out of the hydra and turn it into something else.
But nobody would be able of running any of my code for them selfs unless they built the same mods as well.
So I was just interested if there was going to be any first party support for such a device.
Oh well...
Jim, Glad you asked.
SRAM is Static-RAM. or Static-Random Access Memory.
Latching is when you set a signal and it holds it after you've set it until cleared.
However, if you don't need bitmapped graphics then the 32K prop memory is more than enough to do very good tile based graphics with sprites. My included HEL engine with the hydra does tiles, 8 sprites, scrolling, 4 colors per tile at 16x16 tiles, and a lot of other cool features all in one cog and very little memory for example. But, if you do want a decent 256x192 like display with at least 16 colors per pixel then you will have to go to external memory archictures. But, 160x192 with 4 colors per pixel and a double buffer costs about 16K, if you add onto this a scanline palette table, then you can change palettes every scanline and get a pretty good display. Assuming, that you load your cogs with ASM code and you can blow away a lot of the code once its loaded, then you can free up most of you code space and use it as VRAM as long as you are NOT using SPIN code primarily.
So even with the lack of external SRAM, you can still get quick a bit done with the internal 32K with the right tricks, but there is no arguing you have to work for it [noparse]:)[/noparse]
Andre'
That's true.
It's just that I think 16 color bitmaps would be awesome for some programs, a layer with any pixel anywhere and color dithering would be really cool!
With the (hss) sound object, I worked months rewriting it and adding compression so that the programmer could have wavetable music and soundFX while keeping 80% of the chip open. I wanted the programmer to have a lot left over for the Vram tables.
Sometimes though I wish I could write mod music files with 28k of wavetable and not have to worry about the Vram. (lol)
Until then I'll stick to sub 2k tables and files.
Andre! I'd be so happy when you release such a card!
128k of RAM for Vram and/or Sample-Ram would make me a very happy person. ^^
*smiles*
(btw) You wrote a really awesome video object from the sounds of it. *I'll see it soon enough when I get my Hydra*
Thanks!
In interviews he said that the key was the use of 16k ROM in Solaris compared to the 8k ROM chip for Star Raiders.
Of course the issue there, as you have said, is that then you end up with something that is not standard. If you can build it cheap enough you could sell the carts to people though.
I think we should get a bunch of people to vote for features for a "Hydra Super Expander" cart (the HySEC). Everyone could give input into what they would want to see on the cart then agree on that as the standard. We might even be able to find someone to manufacture the thing for the folks that don't want to mess with that themselves.
--Chris
But, for now, let's see what people do themselves. And once I get a moment, I will design the card, take as many variables into consideration as possible and make as many people happy as possible, and manufacture it.
That said, I am always open for "wish list" items and suggestions, but remember COST COST COST -- it does no good if the card is $99! Unless of course it has another processor on it, and is like a "CPM" card or something in the same vein, but from the 21st century rather than the 80's
Andre'
Really I'd be happy just seeing a SRAM and a large EEPROM (or two 128ks). (though 256k is quite a bit of storage for the hydra if I do say so my self).
I'm big on cost too and view it more as a SuperCart. Sort of like the FX carts on the SNES or some of the later NES carts with advanced Memory Mappers.
Something cheap enough so that you could store a few good games on a handful without being rich.
I'm picturing a EEPROM, Sram and some glue logic for the Sram address latch.
The propeller is powerful. It can handle the rest pretty well.
I would love to see plug in 6502, Z80. 6809, AVR, PIC, Prop, 68000 cards as well.
Anyway, once I get clear from all this "biz" stuff I am stuck doing right now, I will get back to designing more stuff for the Hydra. The main thing right now is to get thousands of them into the wild, so if we do make a card, and say 10% of the owners buy it, then we sell a few hundred and its worth it. So key thing is try try and make the hydra a platform, so we have a standard to expand from.
Andre'