Switched power for IDE-to-SATA adapter from mobile bay PCB

[BonzaiDuck]

Assume one has a port for a 12V circuit and device. A "+" pin and a ground pin.

You want it to be a 5V port.

Isn't it possible to build either a small PCB or just a wired assembly to convert 12V to 5V?

And maybe someone knows the answer to this question: What components? Is there a circuit diagram?

 

[corkyg]

Maybe this should have been asked on the HT Forum? What does it have to do with M&S? Anyway, how much current is involved?

 

[ArisVer]

Great for cars. Haven't looked at it thoroughly but it looks simple to make.

http://www.instructables.com/id/12v-to-USB-adapter-12v-to-5v-transformer-great-/

 

[BonzaiDuck]

Maybe this should have been asked on the HT Forum? What does it have to do with M&S? Anyway, how much current is involved?

You'll find maybe a couple recent threads on IDE, IDE hot-swap bays/caddies and IDE-to-SATA adapters in the mix.

There is a 12V plug and pinout accessible at the bay PCB of my mobile rack/hot-swap assembly. It is now fairly certain that this is there to power a 12V fan for certain bays fitted with the PCB, and fairly certain it is 12V. It is also very likely -- subject to verification -- that the key-switch on the front of the caddy (or bay and whichever) controls power to that pinout.

So I'm working with someone who may wire together a 3-leg 5V regulator device for me. And he's asking the same question: "What is the current requirement?"

To be succinct -- a 5V and ground from that 12V plug should enable me to turn off an Ide-to-SATA adapter with the key-switch the turns off the power to the IDE HDD.

This resolves a minor inconvenience that requires a drive to be connected to the IDE-to-SATA adapter at boot-time, when the adapter is power-cycled and enumerated by the HDD controller BIOS. Otherwise, there is either (a) a delay in enumerating the drives on the controller, or it is stuck in an endless loop hoping to find a drive, but can't.

So without switching off the adapter, I have to insert a drive and turn on the keyswitch before I reboot the system. Once the OS has presented the desktop, I can otherwise "safely remove" the drive, turn the keyswitch off, and "uninstall" the adapter (identified still with the drive name and model) from the Device Manager.

With a switched adapter controlled by the keyswitch, I wouldn't have to enter device manager to "uninstall;" and I wouldn't have to leave a drive connected at restart or reboot.

 

[Elixer]

Right, you need a three terminal voltage regulator, and you need to know the max current consumed to be able to rig up a resistor of the correct value to do this.
In your case, I don't think this is a very good idea, and could cause issues.

If I recall correctly, there should be a time out value in BIOS for IDE devices, usually, promise controllers (most common) are set to timeout at 30secs, but, you should be able to disable this controller in BIOS. For a BIOS that don't have this, just hook up a DVD drive or whatever, and that would be good enough.

 

[denis280]

Right, you need a three terminal voltage regulator, and you need to know the max current consumed to be able to rig up a resistor of the correct value to do this.

+1 and ex: a 1k resistor will make a single blinking led to work on a 12volt battery

 

[BonzaiDuck]

Right, you need a three terminal voltage regulator, and you need to know the max current consumed to be able to rig up a resistor of the correct value to do this.
In your case, I don't think this is a very good idea, and could cause issues.

If I recall correctly, there should be a time out value in BIOS for IDE devices, usually, promise controllers (most common) are set to timeout at 30secs, but, you should be able to disable this controller in BIOS. For a BIOS that don't have this, just hook up a DVD drive or whatever, and that would be good enough.

Well, the controller is PCI_E card about which I posted in another thread. The adapter is connected to the controller, and the drive is connected -- via caddy and bay -- to the adapter. I'll have to look into it.

But -- figure. Some tech-expert(s) have had great luck in powering these adapters so that the keyswitch on the caddy/bay turns off the power together with the drive. And most of the adapters -- maybe all of them -- really just use the +5V and ground wires of a floppy 4-pin power plug off the PSU. The "wiring mod" required identifying the solder points on the bay PCB, and soldering the 5V and ground to them. Risk and mess with that, at least for me. But what would be better than providing the 5V from a little white port on that PCB, if the keyswitch turns off the power to it? You add in the voltage regulator to get the proper voltage. I hesitate to imagine there could be complications.

If it's done right, it shouldn't really be a problem. In fact, the way it works now isn't a problem either. [And I still need to look at the BIOS of the controller card.]

 

[RU482]

there are a ton of ready made circuits that you can buy off ebay just like this. cheaper than you can buy the parts for, usually

 

[BonzaiDuck]

there are a ton of ready made circuits that you can buy off ebay just like this. cheaper than you can buy the parts for, usually

Yes. I'm corresponding with a tech-veteran in UK, and my long-time techie friend in Virginia. UK came back with a link to such a device -- which you'd only prepare with the appropriate plugs to connect. I'm going to let VA put it together, and even offered him some money, but he insists doing it for free. I still think I should pay him for the part or Parts.

Right now, we're trying to get an answer to that question: The adapter's current requirements. UK's link shows a device for 3A. I suspect that would still work -- amply. The reseller says it "gets a little warm" when pushing the full 3A. So we'll take our time with this.

Like I said before -- this is all about getting extra value for a total of $350 in drives, caddies and the bay purchased six years ago. One of the drives only had 9 hours on the Smart odometer -- I bought all three of the 500GB drives for backup, so nothing to be ashamed of.

It all works fine the way it is, with the caveat about having a drive connected at boot-time. Hopefully this will just be the icing on the cake -- and such a trivial cake it is. Of course, if it works properly, I won't need to enter Device Manager to "uninstall" the drive so I can swap in another and have it automatically recognized. If the adapter is powered down by the keyswitch, then powered up by turning the keyswitch back on, it will do the same thing.

I'm also pretty sure the DIY mod only poses a risk to the adapter -- not the controller, not the drive-bay/caddy, not the drive. I still have a handful of those things. I can even test the wiring mod with a $2 Chinese knockoff adapter before I use it with the $17 unit.

 

[Cerb]

Right, you need a three terminal voltage regulator, and you need to know the max current consumed to be able to rig up a resistor of the correct value to do this.
In your case, I don't think this is a very good idea, and could cause issues.

In anyone's case, it's not a very good idea. Few 3-terminal regs are going to work well, and reliably, their performance will suck, if not put right at the device, even if they're OK, and by the time it's put together, BonzaiDuck could have just replaced the rack, or bought a decent 12V PSU.

Instability and/or poor performance could well mean killing drives, or any other controller chips in the way, via excess ripple and surges, and/or data corruption from voltage sags.

3-terminal regulators that can work at high speeds are either sensitive, and likely to become unstable, in a custom implementation, work poorly at lower than ideal frequencies (some won't even have specs that go below AM, same as the old regs don't have specs going that high--check page 35 of the datasheet from the reg in the instructables article, for example: it will make substantially noise towards 100kHz, and pass any coming in to your attached device, which is why the PSRR graphs for them usually stop at such low frequencies, as the downslope gets significant), or both. Mere inches of wiring can be the difference between working well and not working, too.

Modding the system as mentioned in the OP would probably be the safest and easiest thing to do. Else, get a premade PC-friendly converter board.

 

[BonzaiDuck]

In anyone's case, it's not a very good idea. Few 3-terminal regs are going to work well, and reliably, their performance will suck, if not put right at the device, even if they're OK, and by the time it's put together, BonzaiDuck could have just replaced the rack, or bought a decent 12V PSU.
Instability and/or poor performance could well mean killing drives, or any other controller chips in the way, via excess ripple and surges, and/or data corruption from voltage sags.
3-terminal regulators that can work at high speeds are either sensitive, and likely to become unstable, in a custom implementation, work poorly at lower than ideal frequencies (some won't even have specs that go below AM, same as the old regs don't have specs going that high--check page 35 of the datasheet from the reg in the instructables article, for example: it will make substantially noise towards 100kHz, and pass any coming in to your attached device, which is why the PSRR graphs for them usually stop at such low frequencies, as the downslope gets significant), or both. Mere inches of wiring can be the difference between working well and not working, too.
Modding the system as mentioned in the OP would probably be the safest and easiest thing to do. Else, get a premade PC-friendly converter board.

I'm studying your response, but am a bit confused, and wonder if you understand what I'm attempting to do. In the second para, you say "Modding the system as mentioned in the OP would probably be the safest and easiest thing to do." The first para is filled with cautions that I'm not sure apply, but I'm trying my best to understand. I could post pictures, but I'd either have to take them and upload, or copy from e-mail correspondence.

Here's a link to similar "Addonics" devices that more or less describe what I'm trying to mod into a StarTech hot-swap bay and IDE-to-SATA adapter:

http://www.addonics.com/products/saturn_hot-swap-kit.php

Startech and other similar IDE-to-SATA adapters are powered by the ubiquitous 4-pin floppy-drive power plug from the PSU; PSU's still include the patch cable for conversion of the standard 4-pin Molex. The adapters only use the 5V (red) and a black ground.

A 4-pin Molex feeds my hot-swap bay to power the drive and forward fans in the caddies. This fits the bay PCB shown in the Addonics pictures (the PCB built-in to the bay -- not the "IDE extender" or the two types of adapters (IDE-to-USB or IDE to sATA). There is a matching PCB in each caddy. Each of my caddies has a keyswitch, which either locks the drive into the bay and turns it on, or the reverse. This also turns the 5V and 12V on and off.

It is now fairly certain that the little white plug SIMILAR TO that called "power-sensing connector" on the Addonics bay PCB provides 12V, and that it is also switched on or off with the Keyswitch in the caddy. I must still determine if all these likelihoods apply to my own bay and caddy.

I had two identical IDE-to-SATA Startech adapters and did the stupid thing of assuming the white plug carried 5V, so the first of those popped and blew out as soon as I powered the system with the caddy locked in. Since bay was providing the power, there was no damage to either the bay or the drive. Only the adapter was destroyed. Similarly, no damage to the SATA signal cable or the controller at its other end.

so I have the spare adapter and some $2 similar units to play with.

This time, I'll be thorough and put my multi-tester on the PCB bay's white-plug to assure that it's 12V, but I can't imagine what else it would be. [There's always the possibility that it WAS a 5V connection and I had a defective adapter, but the empirical details suggest otherwise.] The UK techie says that his bay PCB white-plug powers a fan in the rear of the bay and is 12V; my hot-swap bay/caddies appear to be of SIMILAR manufacture, and the rear of the bay seems to have a place for a small fan. The UK tech verified that his 12V white plug is powered down by turning the caddy keyswitch, and I can similarly prove that mine either does or doesn't.

If all that proves true -- (a) the white plug provides 12V power and (b) it powers up or down with the keyswitch, then I only want to convert it to 5V to power the IDE-to-SATA adapter. If the adapter is powered up or down with the keyswitch simultaneous to the caddy/drive, then the problem of reboot and post without a drive connection to a powered adapter goes away.

So far, the consensus is that the IDE-to-SATA adapter only draws at most 500 mA -- possibly as little as 50mA. It is simply a matter of providing a "detour" of PSU power through the bay/caddy assembly which is controlled by the caddy keyswitch together with the power to the drive.

Now I've written all this stuff, and I know what folks are thinking: "Why bother!?" DOING IT should be quick and easy. NOT DOING IT leaves me only a minor inconvenience. TALKING ABOUT IT, RESEARCHING IT, DOING STUPID S*** like I did with the first of the two spare adapters -- that takes time.

And for their speed, even at SATA 1.5 through an SATA-III controller, these Hitachi 500GB IDE drives are only good for certain types of backups.

 

[Cerb]

You rely on the adapter chip to manage the drive and your data. If it is bridging grounds, who knows what would result if it were damaged over time from a bad reg, or if it silently corrupted along the way.

Meanwhile, you have a PC PSU with 5V lines, no?

 

[BonzaiDuck]

You rely on the adapter chip to manage the drive and your data. If it is bridging grounds, who knows what would result if it were damaged over time from a bad reg, or if it silently corrupted along the way.

Meanwhile, you have a PC PSU with 5V lines, no?

Yes. And I explained this, though not impatient with your question.

The IDE bay-caddy assembly had been sold with software that managed true hot-swapping. The software only works for 32-bit OS's. So the question arose "What do you do NOW!?"

It appears that the IDE drive IS hot-swappable DURING an OS session IF it is connected through the IDE-to-SATA adapter. The CAUTION promoted by the adapter instructions: "not advisable to remove the powered adapter from its IDE port on the back of a drive/drive bay."

With the software, the "safe-removal" was still managed by the Windows hardware layer with a direct IDE-to-IDE connection. Windows STILL manages it, but the adapter remains powered on. "Safely Remove . . " actions at the system tray icon prompts Windows to say "The drive has been prepared for removal." At that icon, "you may safely remove the drive." In the device manager, the drive remains a transparent device (through the adapter). That is, it remains on the DM tree identifying the drive, with the information "has been prepared for removal; power down the device and remove it." "Uninstalling" this node in the DM tree makes it possible to replace the drive/caddy with another, and the system recognizes it with "installing driver . . . device is now ready to use."

IF the adapter is powered down with the drive through the keyswitch, this additional step would be unnecessary -- it would be managed by Windows automatically.

But the simple fact that IDE-to-IDE was managed through software and Windows, the other fact that the drive is totally "Smart" transparent through the adapter -- suggest that turning the keyswitch to unpower the drive after it has been "prepared for removal" is safe. IT is the powered adapter which is recognized again as "something" without a drive attached that presents a problem at reboot as the drives on the storage controllers are "re-enumerated." The adapter was meant to use as a permanently connected device, but hot-swap through SATA is possible, as was hot-swap IDE-to-IDE with the SOFTWARE through Windows was possible.

And the manufacturer -- the installation guide -- to the adapter assures "hot-swappable" as an SATA device. The drive is transparent in every SMART detail. So if Windows says "prepared for removal," then it's also prepared for removal at the bay-to-caddy connection. It just leaves the adapter powered on, showing the "drive" in the system tray until it is "uninstalled."

 

[Cerb]

It was my understanding from the prior post(s) that the key switch mechanically turns a switch that removes or applies power w/ 12V. Is that the case? If so, aside from LED indicators, is it actually doing anything (IE, has its own hardware logic--it looks like a pass-through)? If not, wiring it for 5V from the PC's PSU connector to the switch back to the converter should make it shut the converter off.

Once Windows unmounts the partition(s), the actual method you power the drive off with shouldn't matter. So, if +12V stays hot until you pull, it shouldn't hurt anything unless you have a very old drive (the kind resting heads on the center of the platters). If the switch is 2-pole, you could even give it 12V and 5V, just sacrificing whatever else is connected to (the photos I could find make it look like it's going straight to one of the LEDs, FI).

 

[BonzaiDuck]

Just thought I'd update this stale thread about making an IDE "mobile bay/caddy" work like your typical SATA mobile bay/caddy.

The little white two-pin port on the bay PCB is indeed a 12V fan port. Further, we verify over at least a couple makes of bay/caddy assemblies (and that would also include SATA) that the current shuts on and off with the bay/caddy key-switch.

So the ingredients include:

Matching female plug to the PCB bay 12V port

red and black wires to connect the female plug to the 5V pins of a floppy female power plug

A 3-leg "7805" 5V regulator for input voltages 7 to 25V

Your typical IDE-to-SATA3 adapter (the old ones only work at SATA-150. Big difference, and worth the $15 for the IDE-to-SATA3)

Connect the ground wire from both the 12V end and the 5V adapter side to the ground on the regulator

Connect the 12V red wire to the regulator "input"

Connect the 5V red wire to the regulator "output"

Hook it all up. Your IDE hot-swap is now an SATA hot-swap that is "totally hot-swappable."

If the regulator fails, it will simply stop providing power to the adapter. This might of course cause data-loss. But it is not likely to happen. Not likely -- period.

Generally, the regulator model with least amperage is the 1.0A model. The adapter draws between 50mA and 500mA, so . . . no problem there.

 

[Elixer]

You got any pics of this frankenstein creation of yours ?

 

[BonzaiDuck]

You got any pics of this frankenstein creation of yours ?

As soon as I pick up the blister pack of regulators, mod my wiring and receive the IDE-to-SATA3 adapter I'm going to use -- I'll post.

But basically, just a six-inch red/black wire pair from one female plug to another, with the regulator added to it. Again, the whole idea is to unpower the IDE-to-SATA adapter with the same keyswitch that unpowers the drive caddy.

It wouldn't be too different than what this Addonics kit provides for their own mobile hot-swap bay/caddy (eyeball the pics on the right side of the page):

http://www.addonics.com/products/saturn_hot-swap-kit.php

The 6-inch wire they use to connect the bay PCB to either the USB or SATA adapter only differs from mine in that I have to add the 5V regulator to that wire. Apparently, either the two-pin (white) port on THEIR bay-PCB is providing only 5V, or their adapters convert 12V to 5V.

If for instance someone had a similar IDE hotswap bay that had a rear fan, it would likely be connected to the bay PCB "white" port. One could probably just tap into that fan wire (in parallel) with a patch that included the 5V regulator, and hook it up to the IDE-SATA adapter.

 

[BonzaiDuck]

So . . . Radio Shack is not conducting a great fire-sale. The employees aren't worried. They only had the "7805" 3-leg regulator in stock, with ample current spec of 1.0A.

Probably three among the best solder joints I ever made.

But . . . I should have trimmed about a 1/2" off the first and third legs, so the length of ground-wire would equal the total length of the input 12V and output 5V wires. I found myself wrapping the red wires around the black one so I could then wrap them up in self-adhesive rubber bandage.

Now as soon as the pit-crew adhesive cures to a clear color where all the "legs" enter the regulator, I'll hook it up and give it (the hot-swap drive) a spin.

 

[denis280]

Nice Job:thumbsup:

 

[BonzaiDuck]

Nice Job:thumbsup:

Well, I want to be modest about this. First, the hard-wire solder-mod of UK's Dave Curran was the only perfect solution available searching the web high and low. So it seemed. He was the only person who knew how this worked.

But he had a 35mm fan in the rear of his IDE mobile bay, plugged in to the similar two-pin port of the bay's rear PCB. He just forgot to test whether the fan went on or off with the keyswitch -- until I posed the question to him. Otherwise, he could've tapped into the fan connection in a parallel circuit, perhaps by doubling the wire ends in the pin-retainers of the female plug.

We STILL haven't tested the 12to5V mod. My original plan was to test it with the current six-year-old IDE-to-SATA(150) one-directional adapter I'm using at this moment. I had RMA'd the PATA2SATA3 Startech plug -- thought I didn't need it -- which had worked fine two weeks ago, and I decided to order another for a few dollars less.

So I'm going to be careless and sloppy and test my switched-power connector with the brand-new, untested SATA3 adapter. I rationalize that ST had always been reliable parts, that I'd used the same plug before the RMA, and "what have I got to lose, except a little extra time?"

Meanwhile, while I'm STILL waiting for the Pit-Crew acrylic to cure into a translucent stiff rubber, I decided to post this picture with the wiring plugged into the PATA2SATA3 together with the SATA cable. I rubber-banded the cable to the side of the IDE port, to show there was just enough slack in the wire so that everything fits snugly with no stress on solder joints or wires. The bay PCB port for the white plug is located right next to the IDE plug on the PCB:

The other common-ground of the regulator has greater purpose as a mounting-point or heatsink. Nothing to do with it. It's going to be in a case-zone of reasonable air-turbulence.

It may look like a klooge -- it may BE a klooge, but all the solder-points are neat, small and solid. Everything is properly insulated. I've got the right plugs for the right ports, with the right wires assigned to the right pins.

Maybe a few more hours -- Let's LIGHT this CANDLE!

 

[BonzaiDuck]

So . . . . put it all together . . . Verified keyswitch in the bay/caddy is turned off . . . Power on . . . Drive enumeration at post doesn't see the new PATA2SATA3 adapter . . . boot into WHS . . . Insert the key into the switch and turn it on . . . "Installing driver . . . Device Hitachi [blah-blah . . 80] ready for use."

Stablebit Scanner shows 47 to 53 MB/s after scan on the drive has been initiated . . .

Go to the "Safely Remove" icon . . . Click on "Hitachi [yada-yada] Drive L:" "It is now safe to remove the device . . . " turn the keyswitch . . . No node for "Hitachi . . . " in the Device Manager under "Drives/Disks" . . .

$2 for a 5V 1A regulator . . . old plugs and wires from the parts locker . . . $17 for the PATA2SATA3 adapater . . . $350-worth of old hardware still useable as fully "hot-swap-capable."

Mission Accomplished . . .