Chip speed

Discuss hardware-related topics, such as development cartridges, CopyNES, PowerPak, EPROMs, or whatever.

Moderator: Moderators

Post Reply
noriaki_kakyouin
Posts: 34
Joined: Wed Apr 26, 2006 10:31 am

Chip speed

Post by noriaki_kakyouin »

When choosing chips (EPROM or Flash) what speed would you suggest? When ordering from this site I noticed that they have chips with speeds from 150ns to 45ns. What should I get?
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Post by kyuusaku »

Anything <=200ns will do.
User avatar
tokumaru
Posts: 12427
Joined: Sat Feb 12, 2005 9:43 pm
Location: Rio de Janeiro - Brazil

Post by tokumaru »

I see... so there is no such thing as a "too fast chip" then?

What exactly is this time by the way? Is it the time it takes for the chip to output the data after receiving the address or something like this?
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Post by kyuusaku »

Exactly, it's the time for the data on the outputs to be valid.
noriaki_kakyouin
Posts: 34
Joined: Wed Apr 26, 2006 10:31 am

Post by noriaki_kakyouin »

GAH! NANO SECONDS! I'm a dummy... nevermind me. Thanks for the help, gang.
Celius
Posts: 2158
Joined: Sun Jun 05, 2005 2:04 pm
Location: Minneapolis, Minnesota, United States
Contact:

Post by Celius »

OH! This probably explains why many of my devcarts didn't work ever! So, I'm a little confused here. Tokumaru says a chip can't be too fast, but you want anything Less than or equal to 200ns. Just to be clear, I can read, but do you want something Less than or equal to, or Greater than or equal to 200ns?
tepples
Posts: 22708
Joined: Sun Sep 19, 2004 11:12 pm
Location: NE Indiana, USA (NTSC)
Contact:

Post by tepples »

The smaller the number, the faster the chip. You want something 200ns or smaller (faster).
Celius
Posts: 2158
Joined: Sun Jun 05, 2005 2:04 pm
Location: Minneapolis, Minnesota, United States
Contact:

Post by Celius »

...Oh my god, I'm retarded. Thanks.
User avatar
Quietust
Posts: 1920
Joined: Sun Sep 19, 2004 10:59 pm
Contact:

Post by Quietust »

275ns or faster is sufficient for the CPU bus (i.e. PRG ROM, SRAM).
175ns or faster is needed for the PPU bus (CHR ROM, CHR RAM).
Quietust, QMT Productions
P.S. If you don't get this note, let me know and I'll write you another.
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Post by kyuusaku »

Well if we're getting technical 1/(21.47727MHz/4) = 186.27ns ;) Also does the PPU even expect data within a cycle? I believe most 200ns to still work, since my 200ns (maximum latency, yes I know) do.
User avatar
blargg
Posts: 3715
Joined: Mon Sep 27, 2004 8:33 am
Location: Central Texas, USA
Contact:

Post by blargg »

The speed rating of a chip is the maximum time until the data is valid under worst stated conditions. Like overclocking processors, the chips might perform better in some circumstances; one batch of seemingly too slow chips might work, another might fail. If you want reliable operation, pick parts whose specs meet your requirement.
User avatar
kevtris
Posts: 504
Joined: Sat Oct 29, 2005 2:09 am
Location: Indianapolis
Contact:

Post by kevtris »

Celius wrote:OH! This probably explains why many of my devcarts didn't work ever! So, I'm a little confused here. Tokumaru says a chip can't be too fast, but you want anything Less than or equal to 200ns. Just to be clear, I can read, but do you want something Less than or equal to, or Greater than or equal to 200ns?
I don't know... I think the memory devices CAN be too fast. 250ns is good for PRG, and 150ns is good for CHR usually. I wouldn't go below 90 or 100ns if you can help it (for either).

Bad things can start happening, like bus ringing and such (due to the beefier pin drivers on the faster chips), and they can suck more power, and make larger spikes which need more decoupling.

The worst thing that can happen would be some ground bounce, which could totally trash the data in a bunch of RAM locations if it happened when the address bus was changing.
/* this is a comment */
User avatar
blargg
Posts: 3715
Joined: Mon Sep 27, 2004 8:33 am
Location: Central Texas, USA
Contact:

Post by blargg »

The reason faster chips have higher power requirements is that they need faster transitions between logic levels. A slow chip can have a more gradual waveform since it doesn't need to be able to change as often, while a faster one needs a sharper one:

Code: Select all

            ______
         _-~
      _-~
____-~
   |--------| longer transition
       __________
      /
     /
____/
   |---| shorter transition
I've read that faster chips might also expect shorter transitions on inputs, so feeding them signals meant for slower chips might register as false transitions.
User avatar
Bregalad
Posts: 8056
Joined: Fri Nov 12, 2004 2:49 pm
Location: Divonne-les-bains, France

Post by Bregalad »

I often can buy chips about ~75 ns. And there is no problem. I don't think chips can be too fast, unless they're 2 ns or so, if they use some ultra-modern decoding technology on the bus with very old CPU.
Useless, lumbering half-wits don't scare us.
Post Reply