Maker Pro
Maker Pro

Using PLD or FPGA for ISA bus board with DMA

G

Guest

Jan 1, 1970
0
I'm planning to build an ADC board with onboard buffer with DMA transfer
from onboard buffer to computer's memory.
Where can I find an ISA bus board design example with DMA functionality
using PLD or FPGA ?
What would be the max paractical speed to transfer data from ISA bus to PC's
memory?
 
K

KR Williams

Jan 1, 1970
0
I'm planning to build an ADC board with onboard buffer with DMA transfer
from onboard buffer to computer's memory.
Where can I find an ISA bus board design example with DMA functionality
using PLD or FPGA ?

ISA is trivial. A simple web search should give you all the
information you need. What sort of DMA do you want? ...classic
ISA DMA, busmaster? Both are possible, but both rather rot. If
you really need DMA support, I'd take the plunge and go PCI.
It's not all *that* hard, if you use the bridges and drivers that
are available (of course you didn't mention you budget either).
What would be the max paractical speed to transfer data from ISA bus to PC's
memory?

With a modern PC, I wouldn't expect more than 500kB/s, *maybe*
twice that. ISA sucks rocks and no one cares about giving it a
hand.
 
A

Assaf Sarfati

Jan 1, 1970
0
I'm planning to build an ADC board with onboard buffer with DMA transfer
from onboard buffer to computer's memory.
Where can I find an ISA bus board design example with DMA functionality
using PLD or FPGA ?
What would be the max paractical speed to transfer data from ISA bus to PC's
memory?

ISA was never designed for DMA, unless you consider the 8-bit 8237 on the
motherboard. Both ISA and the 8237 are very much obsolete, and even if
present on current chipsets, they have probably not been verified for
correct operation for generations (chip generations, that is).

Unless entirely impossible, I'd suggest designing a PCI board; there are
many interface chips available, and most of them include some sort of DMA.
Best of all, the PCI bus was designed to use DMA as its basic data-transfer
method (it's not called DMA; it's called Master transaction in PCI-speak).

Designing a board around a PCI interface chip not too bad; the chip makers
have lots of guides, app notes and possibly even evaluation cards available;
PCI is doable using 2 signal and 2 power planes - even (if you are very
careful) only 2 signal planes. OTOH, you can't wire-wrap or protoype one
very easily.

If going the PCI route, I'd recommend learning at least the PCI basics,
since it's very different from ISA, or any single-processor async bus
(which ISA basically is).
 
Z

Zak

Jan 1, 1970
0
Assaf said:
ISA was never designed for DMA, unless you consider the 8-bit 8237 on the
motherboard. Both ISA and the 8237 are very much obsolete, and even if
present on current chipsets, they have probably not been verified for
correct operation for generations (chip generations, that is).

I remember the Adaptec 1542 (busmaster DMA on ISA) to be problematic
with PCI chipsets.

Also, DMA over ISA is limited to 16 megabytes of reachable memory.

Programmed I/O over dual ported RAM (DMA to memory on the card, that is)
might be the easy way out, and probably fast enough.


Thomas
 
K

KR Williams

Jan 1, 1970
0
ISA was never designed for DMA, unless you consider the 8-bit 8237 on the
motherboard. Both ISA and the 8237 are very much obsolete, and even if
present on current chipsets, they have probably not been verified for
correct operation for generations (chip generations, that is).

THere are three forms of ISA DMA. The 8237 DMA registers only
address one. ISA can also busmaster (HOLD/ HOLDA/, IIRC). IBM
had a series of modem/sound cards that did ISA busmastering. It
worked rather well, though the drivers had to copy into buffers
below the 16MB line.
Unless entirely impossible, I'd suggest designing a PCI board; there are
many interface chips available, and most of them include some sort of DMA.
Best of all, the PCI bus was designed to use DMA as its basic data-transfer
method (it's not called DMA; it's called Master transaction in PCI-speak).

I wouldn't disagree with this assessment. ISA is dead, and may
it stay dead. ;-)
Designing a board around a PCI interface chip not too bad; the chip makers
have lots of guides, app notes and possibly even evaluation cards available;
PCI is doable using 2 signal and 2 power planes - even (if you are very
careful) only 2 signal planes. OTOH, you can't wire-wrap or protoype one
very easily.

You won't meet spec without the internal planes. It might work,
but might not. I'd never go below 2S2P for a PCI card, which
throws it out of the range of the hobbyist.
 
A

Assaf Sarfati

Jan 1, 1970
0
KR Williams said:
THere are three forms of ISA DMA. The 8237 DMA registers only
address one. ISA can also busmaster (HOLD/ HOLDA/, IIRC). IBM
had a series of modem/sound cards that did ISA busmastering. It
worked rather well, though the drivers had to copy into buffers
below the 16MB line.

Actually, you had 2 x 8237 chips: one was wired for 8-bit, one for 16-bit
(even addresses) transfers. The two chips were cascaded (don't remember
which was master and which slave); one 8-bit channel was used for DRAM
refresh and one for floppy. Since the 8237 has only 16-bit address,
there was an external 8-bit register to get 24-bit address (up to 16 MB,
can't cross 64-KB physical address boundary in one DMA transaction).
You could also program a 8237 channel to be bus arbiter only, using your
own hardware to create bus cycles.

However, none of this was widely used, and even when there was no PCI,
many ISA DMA cards wouldn't work in all "PC Compatible" systems. IIRC,
most of these high-performance cards has a default programmed-I/O mode
(for example: basic ATA-controller PIO mode), which was the one used
by most users.

I am not at all sure that modern chipsets still support all these
modes. After all, who in his right mind will use an ISA bus-mastering
network or tape interface card TODAY? why waste verification CAD cycles
to test these modes? I am not even sure if modern OSes still read/write
floppy using DMA, and using DMA to refresh DRAM is dead as the Dodo.

I wouldn't disagree with this assessment. ISA is dead, and may
it stay dead. ;-)


You won't meet spec without the internal planes. It might work,
but might not. I'd never go below 2S2P for a PCI card, which
throws it out of the range of the hobbyist.

I was assuming a hobbyist card, so he wouldn't care about failure rate in
a 10,000-unit production run; one unit will most probably work - most PCs
leave a reasonable noise-margin error.

You may also be able to get a prototyping board with the PCI interface
ready for use and a prototyping area in the local-bus side.
 
Top