• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

A high-quality SATA PCI 2.5" hard card, to celebrate SATA's 20th birthday

macuserman

Well-known member
May already be accounted for, but has any consideration been given to the issue with the regulators where you end up having to bodge a Micrel 29150 onto the chinese cards to make them work with certain systems.


Seems worth investigating if going to the trouble of making a new card, surely there are other solutions besides that exact regulator but anyhow.
 

aperezbios

Well-known member
This is part of why we chose to _not_ use the generic Chinese cards as a starting point, even though it would have been easier, and less costly, to do. Those designs are just too cost-reduced, in our view. Here's the relevant section of the schematic from the reverse-engineered Adaptec ASH-1205SA power regulation circuitry.1673628105986.png

From the referenced thread/e-mail:

As a recommendation and alternatives:

1. A typical LDO will turn on at 1V. The 5V to 3.3V Micrel regulator will turn on quicker than the LDO from AMC. This allows the 1.8V control logic to be one before the 3.3 on the PCI bus. Using the Micrel regulator could gurantee current power sequencing operation.

2. Two 5V regulators can be used. One that supplies 5V to 3.3V and another than supplies 5V to 1.8V. The same type of regulator should be used to maintain the speed, but typically the 1.8V will be on before the 3.3V."
This design already employs recommendation number two, so we should be good. If we end up needing to make BOM changes in the future, it's really not a big issue.
 

macuserman

Well-known member
This is part of why we chose to _not_ use the generic Chinese cards as a starting point, even though it would have been easier, and less costly, to do. Those designs are just too cost-reduced, in our view. Here's the relevant section of the schematic from the reverse-engineered Adaptec ASH-1205SA power regulation circuitry.View attachment 51175

From the referenced thread/e-mail:


This design already employs recommendation number two, so we should be good. If we end up needing to make BOM changes in the future, it's really not a big issue.
Sweet! I'm super excited to see this come to reality! Don't mean to overstep I think we are all just excited about the possibilities.
 

mdeverhart

Well-known member
For some reason the link got mangled, it should be:

https://www.rabbitholecomputing.com/retro-sata-pci-hard-card-newsletter
 

CC_333

Well-known member
Indeed!

Not that it's needed, but just for curiosity's sake....

Is there any reason why it could/couldn't work on a Windows PC of similar vintage, provided it's equipped with the appropriate PC firmware?

c
 

dosdude1

Well-known member
Indeed!

Not that it's needed, but just for curiosity's sake....

Is there any reason why it could/couldn't work on a Windows PC of similar vintage, provided it's equipped with the appropriate PC firmware?

c
No option ROM is needed for a Windows machine (I believe). But if needed, you could always flash a standard Sil3112 PC option ROM.
 

aperezbios

Well-known member
No option ROM is needed for a Windows machine (I believe). But if needed, you could always flash a standard Sil3112 PC option ROM.
Correct, no ROM is needed under Windows, _if_ you don't need to boot from it, although you'd still need a suitable SiI3112 driver. The Silicon Image reference PC boot ROM is available via archive.org, from the original Silicon Image website.
 

CC_333

Well-known member
That's good!

This means that this card will also have much appeal for the vintage PC crowd, I would think.

c
 

Byrd

Well-known member
Love the M.2 slot and "hardcard" SATA 2.5" functionality. Will be keen to see how a mature, stable SATA chipset functions with certain brands of SSDs. Find it really hit and miss with SD/CF/SATA/M.2 compatibility in my Macs and PCs using generic adapters.
 

dosdude1

Well-known member
Love the M.2 slot and "hardcard" SATA 2.5" functionality. Will be keen to see how a mature, stable SATA chipset functions with certain brands of SSDs. Find it really hit and miss with SD/CF/SATA/M.2 compatibility in my Macs and PCs using generic adapters.
These cards are top-notch, and won't exhibit any issues you normally get with IDE to SATA adapters (and also solve a compatibility issue with some G4 machines present in the cheap Chinese Sil3112 cards). The main issue with those adapters comes from the fact that they use the horrible JMicron JM20330 IDE-SATA bridge IC, which has compatibility and reliability issues with many machines, and when it does work properly, is normally extremely slow (I've seen them top out at 33 MB/s on some machines). This is also one of the main reasons I designed and built my own custom 1.8" ZIF and 2.5" IDE SSDs, so I can finally have reliable and decently quick flash storage in mobile systems. For a desktop machine, a PCI SATA controller card such as this is the best option.
 
Last edited:

CC_333

Well-known member
Love the M.2 slot and "hardcard" SATA 2.5" functionality. Will be keen to see how a mature, stable SATA chipset functions with certain brands of SSDs. Find it really hit and miss with SD/CF/SATA/M.2 compatibility in my Macs and PCs using generic adapters.
That's a good point.

These cards are top-notch, and won't exhibit any issues you normally get with IDE to SATA adapters. The main issue with those adapters comes from the fact that they use the horrible JMicron JM20330 IDE-SATA bridge IC
I've heard about JMicron stuff being rather quirky and unreliable across the board, but of course, it's ubiquitous not because it works well, but because it's cheap and works well enough, which of course doesn't mean much.

It's probably too much, but something that would make this card super neat is if the firmware could be made open source somehow. Then it could be optimized and improved in ways that would be too non-trivial with the stock FW, couldn't it?

c
 

dosdude1

Well-known member
That's a good point.


I've heard about JMicron stuff being rather quirky and unreliable across the board, but of course, it's ubiquitous not because it works well, but because it's cheap and works well enough, which of course doesn't mean much.

It's probably too much, but something that would make this card super neat is if the firmware could be made open source somehow. Then it could be optimized and improved in ways that would be too non-trivial with the stock FW, couldn't it?

c
Unfortunately, all the available Mac ROMs are very much closed source, so there's not much that can be done with those outside of reverse-engineering. However with that said, I may some day write my own custom firmware for these cards, which of course I then would make open source.
 
Top