Compact Flash the Same as MultiMedia/Secure Digital Cards?

[Dale Stewart]

Hi

Is Compact Flash the same protocol, etc for programming as the MultiMedia/Secure Digital cards with the Compact Flash library?

Thanks

Dale

[mipedja]

Data in MMC and SD cards are stored in sectors in a similar way, but protocol is not the same. In the case of CF, data transfer is parallel, and in the case of MMC and SD serial (it can be simple SPI).

[Dale Stewart]

Hi,

Thanks for your advice.

Data in MMC and SD cards are stored in sectors in a similar way, but protocol is not the same. In the case of CF, data transfer is parallel, and in the case of MMC and SD serial (it can be simple SPI).

So I guess being parallel data transfer it is a lot faster than serial?

Cheers

Dale

[Steve]

Logically, that would make sense wouldn't it. So can someone explain to me why Serial ATA (SATA) is faster than the parallel IDE/ATA bus?

[Bob Lawrence]

You can read about it i here:

http://www.serialata.org/docs/serialata ... nology.pdf

[Steve]

Thanks for that. It appears that serial transfer can actually be faster than parallel. An SD card may be better.

Of course, the fastest drives available at the moment can only move data off the platters at around 55MB/Sec so a 100MB/Sec ATA/IDE bus can keep up with it. With RAM prices so low silicon disks may be the way to go.

[mipedja]

I think that CF should be faster than SD cards.
CF card can work in different ways. Usually for connection to a microcontroller is used true IDE mode. For faster applications memory mode should be used.

[Dale Stewart]

Hi everyone

I am a newbie, so forgive me for the dumb questions

Is USB and firewire serial?

Are they faster than IDE?

I ask this because I know that newer PICs have USB 2.0 on chip, and maybe SD?MM RAM could be faster? Any ideas?

Cheers

Dale

[LGR]

Is USB and firewire serial?

Yes, but the question is more complicated than is one faster than the other. For any given physicall medium, parallel will also be faster than serial. Think of an 8-lane highway instead of 1-lane. But, there are tremendous variations in the capacity of the individual lanes. As a practical matter, serial/parallel doesn't, by itself, mean anything regarding speed.

I ask this because I know that newer PICs have USB 2.0 on chip, and maybe SD?MM RAM could be faster? Any ideas?

The problem with this is that USB is inherently a master/slave topology, and the PICs can only be slaves. The only devices out there than can be masters are computers. Maybe it is possible to write master protocol code for the newer USB-enabled PICs, but this was never intended by microchip, and will be unsupported. Physically, I don't see why it can't be done, but it would require a lot of research.

What mipedja was driving at is that the bottleneck isn't the interface, it is the device itself. There is no point in trying to increase the speed of the interface if the chip can't keep up. They could have written the CF code for memory mode, but it wouldn't have made a difference.

What are you trying to do that requires such high speeds, audio? CF should work for that. If you are trying to do something extremely fast, like video, or waveform capture, you need to look at the entire system.

[Dale Stewart]

Hi

Thanks to all who answered my Q's.

I was just curious from a newbie's point of view for projects later after I have done some basic stuff first.

Cheers

Dale