Is that a JMicron (JM20330) chip? I think the issue might be the Master/Slave setting. Did you also try changing it from the DVD side to avoid conflicting?

I noticed there's a 2-pin jumper near the SATA port. Maybe worth testing if one of the pins lead to pin 33 (MSSEL). That pin controls whether the drive should appear as Master or Slave.

On the other hand, if the BIOS reports wrong UDMA level (like being restricted to UDMA2), you probably need to manually pull pin 34 (PDIAG) to the ground by adding a wire to connect pin 30 and 34.

PS: I'm interested in true 2-port (Master/Slave) adapters as well. They appear to be very rare.

Besides, IDE drives are still useful because almost all of them provide analog audio connectivity necessary for old OSes (like DOS) to play and hear CD audio. SATA drives do not have that anymore, requiring the use of "Digital CD audio".

I had to set the DVD as master and the SATA as slave on some machines to get both to work.

Also, if you value your IDE cables, it might be a good idea to break out (!) the small sides of the IDE cable socket.

What board (chipset) are you using this with?

LSS10999 wrote on 2024-06-30, 12:28:

On the other hand, if the BIOS reports wrong UDMA level (like being restricted to UDMA2), you probably need to manually pull pin 34 (PDIAG) to the ground by adding a wire to connect pin 30 and 34.

Would it make sense to disable UDMA for testing first?

I will take some photos of the adapter.
I will try the jumpers again but (unless I had a brainfart and tested the same positions twice LMAO), I am pretty sure i tested all combinations including cable select on the DVD.

These things exist in many shapes and forms on Amazon. It's a bit annoying that's not passive.

I like this in particular. Costs 20 euro ish.

However these are not exactly what I had in mind. I do like the aesthetics of ribbon cables........ (yes I am weird)...... and I don't know I can accept a vintage motherboard with SATA cables attached to it 😁

The next three pictures show an adapter that I only use if I can give up one IDE device. It does NOT connect 2 SATA to IDE, but can be used both ways for ONE device. (IDE->SATA OR SATA->IDE)
I have tried to find an adapter that would give you 2 SATA ports out of one IDE directly, but had no luck.
You can of course use a PCI controller.

Oh I see. Bummer. Now I notice the "TO DRIVE" and "TO COMPUTER" text.

LSS10999 wrote on 2024-06-30, 12:28:

if the BIOS reports wrong UDMA level (like being restricted to UDMA2), you probably need to manually pull pin 34 (PDIAG) to the ground by adding a wire to connect pin 30 and 34

Unfortunately, it will not work. I've tried that on many such chinese adapters, none of them worked at speeds faster than UDMA-2 on problematic motherboards/laptops. So don't waste your time.
The problem is somewhere deeper.
Also, I've measured the resistance between corresponding pins 30 and 34 on a real IDE laptop hard drive that works good in UDMA-5 mode on such laptops, and it is something about 500 MOhms! Almost no contact at all.

esasoft wrote on 2024-06-30, 15:47:

Also, I've measured the resistance between corresponding pins 30 and 34 on a real IDE laptop hard drive that works good in UDMA-5 mode on such laptops, and it is something about 500 MOhms! Almost no contact at all.

From what I read it's the "80-conductor" cable that's responsible for grounding the pin in question (34).

Normally these two pins (30 and 34) aren't connected on the disk side, though one can actually connect them to manually pull the pin to the ground so as to cheat the BIOS into thinking you're using a 80-conductor cable, thus enabling you to select a UDMA level higher than 2, up to the maximum supported by your controller.

Since I lurk into the Xbox community, my best recommendation is obviously gonna be a Startech.

That, and a chinese adapter with "VER:M03C" printed on it, which needs two resistors soldered to run at same quality as Startechs, and is based on the same IC used for the adapter in picture 1 post 2.

The M03C looks like this:

The missing resistors to be added to those are as following:

RH1 - 10k SMD
R4 - 100 ohm SMD

LSS10999 wrote on 2024-06-30, 16:02:

Normally these two pins (30 and 34) aren't connected on the disk side, though one can actually connect them to manually pull the pin to the ground so as to cheat the BIOS into thinking you're using a 80-conductor cable, thus enabling you to select a UDMA level higher than 2, up to the maximum supported by your controller.

This is all clear. But in the case of laptops, there are no wires, and the hard drive is directly connected to the motherboard. In fact, what is proposed to be done with the adapter is equal to connection of pins 30 and 34 on the HDD side.
It doesn't work, anyway.

I buy the ones that put the foam where the PCB touches the HDD to prevent shorts. I forget the brand.

Adapters using Marvell 88SA8052 (or derivative) usually work the best. Some of them have master/slave jumper but I've never tested them with cdrom on the same cable. It's my habit since the 90s not to mix HDDs and other kinds of drives on one cable, even back then it was sometimes troublesome.

There it is. Sorry for the late answer. JM20330. Is it good or bad?

Wes1262 wrote on 2024-07-08, 10:33:

There it is. Sorry for the late answer. JM20330. Is it good or bad?

I've several JM20330-based adapters and they all work without major issues. I think it also depends on the quality of other components used on the board.

I never found any Marvell-based adapters myself, so I'm not entirely sure about the differences compared to JM20330.

There was some investigation into these: SATA2PATA adapters experience

• Most STAT-->PATA bridges these days have Jmicron chips labeled JM330, JM20330, & "IDE/SATA DOF" (a die shrink or clone of the JM20330). They are usually 1 directional bridges. They either go PATA-->SATA or SATA-->PATA but not both. The performance is very good. They all perform about the same (within the testing margin of error).
• The other brand that you can find has chips labeled "SunPlusIT" . They seemed to be about 5-10% slower, and the adapters cost more, but they are usually bi-directional. The same adapter can do PATA-->SATA and SATA-->PATA
• I got one with a Marvell chip (88SA8052). They don't seem to be in active production anymore. It seemed to have roughly same performance as the Jmicron in the testing that I did. I've seen pictures of bi-directional Marvell bridges, but I don't have one

None of these devices did DMA2 (aka WDMA) for me, so they might not be a good fit for late 486 or Pentium systems. Consider an industrial CF or an SD bridge instead. Getting a device to do a particular ATA transfer mode involves interplay between the storage controller, the bios, the storage device, any device drivers loaded after boot, and sometimes the cable. Maybe I just didn't find the right combination. The Marvel chip did run for a couple seconds before crashing when I forced DMA2 mode on a promise 20630 VLB controller, but it still crashed every time.

Matze79 found a fm4-4383 bridge from Canon that has a chip labeled "bu19503kv". He said it worked well on older systems. I don't know much about it

I also hit a curiosity yesterday when connected a Jmicron bridge with a 40 conductor cable to I try to force UDMA2. Turns out my SATA bridge ignored PIN34 and tried to negotiate UDMA5 on both 40 conductor cables that I tried. I don't recommend running UDMA 5 on a 40 conductor cable. It was pretty clear when looking at the jagged speedsys storage performance graph that there were a lot of data transmission stops and restarts. Once cable was no good at all, the other was like 50% speed. Whatever the case, it wasn't healthy. So maybe avoid 40 conductor cables with SATA Bridges.

Here's the marvel adapter that I found.. It was pulled from an upgraded G3 Mac.

This is the vendor, but their site doesn't seem super healthy these days:

https://web.archive.org/web/20170224025120/ht … ts_2.php?gid=69

Looks like they used to have an adapter based on a Silicon Image SiI3811 which I've never seen before.

https://web.archive.org/web/20170224015843/ht … ts_2.php?gid=65

http://www.bitsavers.org/components/siliconIm … NAL_8-17-06.pdf

SATA to IDE or vice vesa is considered one drive each due to bios cannot do more than 1 devices due to bios design sees as one drive per SATA port same thing, for IDE you need two IDE to SATA adapters to drive two devices on motherboard's IDE port.

Cheers,

douglar wrote on 2024-07-08, 15:05:

This is the vendor, but their site doesn't seem super healthy these days:
<snip>

oh, they're still around:
https://www.amazon.com/Ableconn-IIDE-MSAT-2-5 … m/dp/B017VQT5YW
Be aware that I found issues with the "logo-less" chips... Either that, or my testing was way off... To be fair, that system I tested on had another issue I fixed recently after years of scratching my head (bad RAM).

LSS10999 wrote on 2024-07-08, 11:30:

Wes1262 wrote on 2024-07-08, 10:33:

There it is. Sorry for the late answer. JM20330. Is it good or bad?

I never found any Marvell-based adapters myself, so I'm not entirely sure about the differences compared to JM20330.

I know of Delock 62510 and Renkforce RF-4886682. Both seem to be common mostly in EU, however. I think Ableconn / Delock / Renkforce are all brands of one manufacturer since they have lots of identical products.
Or they source from the same Chinese factory 😀

Both adapters come with nice Master/Slave/CS jumper but their power supply is a bit of a weird choice.
Worked for me with PIO, UDMA2 and higher UDMA with integrated controllers and even UDMA6 on CMD680.
Haven't tested with MWDMA because support for that is problematic on all fronts (drivers, disks, controllers, all have their issues).
I've used Marvell-based adapters mostly for laptops and the compatibility is much better than with JMicrons. In regular PCs it might be less of an issue.

I found this one on ebay with the Marvell 88i8030-TBC chip on it, but I don't see any jumpers. Oh well it arrives I will see how it goes..