68k Soft/Hard Power and ATX conversions . . .

This may be the dumbest question in the world, but i gotta ask it.

Why?

Well, for starters, if I wire up all the proper ATX lines to the proper voltage lines of a IIsi Power Plug, when I flip the switch to "on," nothing is going to happen. The IIsi isn't going to boot because the ATX PSU isn't going to do squat until it gets a Power On/Reset goose from a momentary switch or the circuit I'm proposing.

An old AT power supply is much easier to hack than ATX for a hard powered Mac. Just turn it on and . . . it turns on! The Mac boots and all is well until you shut down the Mac and it tells you to turn off the main Power Switch.

Things may be a bit more or less complicated for the entirely "Soft Powered" 6xx-6xxx series, dunno . . .

. . . but I figured getting a discussion going on how to do every permutation of ATX PSU hack and having the info readily available here was likely a good idea. :-/

. . . that or it's another ID10T error on my part.

oh wait, you mean if you wire up an ATX supply, nothing happens when pressing the keyboard power? If you want to wire up an ATX power supply so that the soft-start functions work normally as if the old original power supply was in place, its not hard at all to cook up a circuit to do this.

the ATX system power supply gives you a PS-ON signal, a PS-GOOD signal, and a 5VSB signal to do just this I have used an ATX power supply AND its soft-start circuitry in other electronic projects, being able to turn on the power supply and check the PS-GOOD signal to make sure the power supply started up correctly.

the PS-ON signal (green wire) gets grounded to enable the power supply to startup. Not exactly sure how the mac does it, have no II series macs at all. You would need to feed back the 5VSB signal into the logic board, when the power button is pressed, it trips a transistor that grounds out the PS-ON, and off she blows.

On the other hand, if its a non-soft start mac, that you want to convert into a soft-start mac, you need to build a custom circuit to do this. which is VERY easy with a couple transistors and an 8=pin microcontroller. Sense the powerkey/keyboard interrupt, fire up the power supply and itll boot.

Its the shut-down part that might be tricky. Need to figure out a place in the logic board to tie a signal into for the power-down command. I am not sure exactly how a real soft-start macintosh does this. probably a seperate signal line from the VIA and ROM code thats hooked/called during the shutdown proceedure.

Or... just use a timed interrupt inside the custom circuit micocontroller, that senses the power button pressed for 5 seconds, then shutdown. So when your at the safe to power off screen, press and hold keyboard power, and it shuts off

Oh and lastly, if you just want the thing to run hot all the time, just jump the ATX green PS_ON signal to ground. (the less fun way). Nice thing about ATX power supplies, it at least gives you the 5V standby signals, something that some of the mac IIs dont, so you gots to have a batt before it powers on.

8-o :O . . . which is VERY easy with a couple transistors and an 8=pin microcontroller. xx(

I'm not much into the microcontroller scene, but I think it would be overkill here. I'm already resarching about this, and in order to power up something from the ADB keyboard, I would use a flip-flop chip instead -- cheap, simple (even with a few more pins ) and no programming required.

So far I had no success on shutting down via software... I already located the relevant VIA port, but that won't change on the SE/30, like it does on the IIx :'( My circuit could, however, shut down a IIx with a hacked (non-soft-power) PSU, but there's little point on that.

Now, for something completely different : With the aforementioned grounding of the PS_ON line on the ATX supply, you can power your IIsi as long as the hard switch (on the PSU) is on; but since both the IIsi and the ATX are soft-power capable, you can keep that feature as intended. The Mac controls power via its PWF line on the PSU connector, which is the opposite of PS_ON -- an inverter is required. There's no need even for an IC, you can do that with a simple transistor. I used a BC547B but any cheap small-signal NPN transistor should do, like the 2N3904 (watch the pinout!): emitter to ground, collector to PS_ON, and the base, thru a 10K (or so) resistor, to PWF on the Mac side.

About the 630/6200 and the like... I experienced the same problem (tried with another board -- same results xx( ) but I think I know what's happening: from the Takky upgrade pages for the ColourClassic, it seems that the ADB power line (S3) is needed in order to power up -- the original PSU has a 6-pin header besides the power connector. But I haven't been able to check this, yet.

TTL and CMOS, NOW we're talking! :approve:

I'd really like to get things Ironed Out and posted in one thread with links to all past threads in the OP.

Edit: just read your linked post, could your problem be that you're not using a momentary switch? I could be reading it wrong because I'm not much of an electron pusher, but having the Backplane Power Toggle send the Power-ON blip to the ATX supply on my flat fat SE/30 is what I'm trying to do. Flipping the original switch off should shut down all but the standby Power on the ATX PSU.

TTL and CMOS is too complex for me. lol. i grew up in the microcontroller world, so thats what I use. I do use TTL and CMOS, but the way my mind thinks is different than most.

its just easy for me to write a 5 line assembler program and zip it on an MCU and perform the same functions as a flip-flop. thats just me. And it would solve the issue on macs like the SE/30 where the power-off line does nothing.

press-hold the power key, MCU recognizes it and kills the power supply.

on the TTL side of things, I guess what one could use is a D-Type latch mechanism with a de-bounce filter. when the D-Flop is toggled on, we could turn the power supply on. Then somehow once the system is on, we charge up an RC Time Constant circuit that must discharge when pressing power before it will allow the D-Flop to toggle off. Then we would work ok for hard-start based macs.

Your right, for an already-soft-start mac to ATX conversion, a simple transistor inverter would work fine.

Could we use all this to get a non softpower mac/IIGS to power on and off with the reset key using an softstart mac ps?

could we somehow get it to see the reset power switch in software to top it all of

(optional)?

rather just use the power key on the keyboard.

yes so would I,the reset/Power key on the KB is one and the same.

Oh, ok. I never referred to the power button as the reset button. Thats why I was confused for a second.

IIcx, IIci, IIsi, Quadra 700

PIN DEFINITION

1 +12V

2 +5V

3 +5V

4 +5V

5 ground

6 ground

7 ground

8 -12V

9 power good

10 +5V (or power on circuit supply voltage)

@Trash80toHP_Mini

so how would we wire up a hard power supply say mac classic to work with a adb keyboard reset button press and a soft power supply ? ie how do we adapt you idea from atx ps to soft mac ps ?

I don't really know enough to have an opinion, suggestion or token bit of information.

Zuiko is more up on this than I am and techknight is one of the real boffins around these here parts of the web.

OOPSIE! I just posted the table above in the wrong thread! :

Check the other thread for what I'm up to lately: Switching Mac PSU's @$$tBits + ATX = ?????

Soft power is an afterthought at this point in my hack . . . which explains why I'm designing a prototyping area into Adapter PCB Alpha.

I might create a new thread in the hopes ppl will brainstorm there, so i can unhijack ur thread Trash80toHP_Mini

here it is viewtopic.php?f=7&t=17937&start=0

please discuss this idea anyone interested/who has the skill! thx

Now back to ur regularly scheduled projecting...

I love the discussion you got started in the new thread. Great stuff goin' down! :approve:

Meanwhile, back at the banana ranch:

Here's the third rev. of my 10 Pin Soft-Power Switching adapter PCB . . .



It's still more of a schematic than the final PCB layout. Parts will need to be moved around some to get it all to fit in the real world.

I need to check it against the IIcx series PSUs for electrical and form factor suitability.

I've retained the fuses for the PSU

I've retained the IIsi's tuschie connector interface, but I may need to lose the header connector and hardwire it to the PCB to get it to fit.

The ten wire pads from the power connector and matching wires from the PSU look like stitching. I'll need to move them forward.

The relay will need to be moved for clearance when I get to the nitty gritty of mating the f2f PCBs IRL.

The power plug is backwards for clarity, the wires go to the adapter and the plug goes to the PSU.

I still need to get the soft power part straight in my head and onto the PCB/schematic and could use some help there.

It's getting somewhere! :b&w:

You dont use FreePCB? its free and its a PCB layout editor. compatible with just about all PCB prototype manufacturers. Hard to get used to, because you have to do netlists and route ratlines to complete the layout, but if you had a schematic, it could be imported and all the routing done automatically. (something I havent done yet.)

No need in my case, I roll my own!

If Compaq "started with a drawing on the back of the place mat . . . " I think this process will suffice. Besides I doodle this stuff at work when I get bored.

Final layout in Illustrator -> EPS -> Signmaking CAD/CAM Package -> Roland CAMMjet -> knife cut vinyl mask -> weeding table ->

two sided Copper Clad FRP PCB Blank from Radio Shack -> etchant from same -> drill press -> soldering station -> finished prototype = in house protoboard a/o short run production.

PCB Layout programs don't nest 3-D objects on PCBs mounted to the inside a tiny box facing one another . . .

. . . my brain does 3D with ease and I can print the revisions in Illustrator for fold-up visual aids and post .jpeg exports of the unit after fine tunin's finished! [] ]'>

PCB work in Illustrator is a far cry from routing a pair of NuBus Card sized double sided boards, a Short ISA card and an adapter board the same size by hand 23 years ago! Each trace started out as a rectangular pad that had to be turned into a trace by adding points and moving them around until they met the target pad which was the deleted. After a while I had a nice little library of trace shapes to modify, but the process was worthwhile, given the ability to cut silk screen stencils and print short runs of mask with enamel SS inks.Even he Edge card connector worked like a charm!

First sketch:



Second sketch:



Third rev is posted above.

Illustrator 3D visualization aid & documentary illustration:



The top view is mirrored as though you'd be looking through the PCB.

The flipped side, with the illustrated connector, would be as if viewed through the sheet metal on the opposite side of the box.

Same for the "side view at the top.

I like to print out boxes so that they can be folded and viewed for the outside as if looking with X-Ray Vision. )

Cooling will be accomplished by mounting a largish fan undernearth the FDD, blowing out the bottom of the case.

Standard galvanized ductwork will lead from the fan to the spot noted on the front of the PSU's sheet metal.

The PSU's PCB solder side will be facing up, with an airflow re-dirtect vane causing the air drawn in the top vents to flow toward the back of the IIsi's PSU sheet metal. It'll then be pulled down across my adapter board and back through the ATX PSU, then pulled out the front by the ducted fan. The entire assembly will be just about as stealth as possible for getting 250W of continuous power in there, with gonzo 5V to run the Rocket and as many NuBus/PDS cards as I can address and cram inside and under the lowly IIsi's case. [}] ]'>

I hope it works out. I would like to see pics when completed.

Thanks, it's working nicely so far, now I just need to find the wire to wire color coding matchups again . . .

. . . I seem to have misplaced them . . .

Anybody got some linkages to good info handy?

@trash80

any news yet mate?

Not yet, I'm trying to get the a boot drive sorted out to test the Rocket in the IIsi as the next step.

The PSU's been sorted out, I just need to build & test it. But, I've got another NIB HP-PSU that's a bit smaller that I want to give a good look see before barreling along with development based upon the original line of research as well.

Some day this spring or early summer the SuperIIsi™ PSU will work, catch fire or explode, depending upon the whims of the gremlins and my soldering skills. )

i have seen an old utube a few years ago, someone running sys 6.0.8 with a 030 mac with a 040 accellerator. it was fast.

IIRC, I Ran 6.0.8 on my IIx/Rocket33 until after the tuneup was released for 7.0.1. I'd have to check the date of release against the receipt to be sure, but the SE/Radius16 ran 6.0.8 until it was retired from plotter server duty by the IIx/Rocket33 when software compatibility forced me to make the jump to PPC. I bought a refurbished Duo 2300c which was eventually replaced by a refurbished 6360 that was slated for Sonnetization.

I don't remember much of a performance hit, because it was an Illustrator upgrade which forced me into System 7, making the comparison an apples & oranges thing

I had never run an unaccelerated NIB Mac until the day I got the G4/466 DA.

Windows 3.0 was the only OS I've ever run before a bugfix version release, but that was because I needed a Windows Box to run CorelDraw without messing around with Win286 . . .

. . . I'm a just a tad conservative when it comes to production machines and the OS under which they run. )

Some day this spring or early summer the SuperIIsi™ PSU will work, catch fire or explode, depending upon the whims of the gremlins and my soldering skills. )

Click to expand...

If I may make a small suggestion - perhaps it would be a good idea to make up a straight pin-pin extension cable between your board and the Mac, so that testing can proceed outside the Mac's case?

I've already got a Q700 PSU sticking out the top of the SuperIIsi ATM, as soon as I can get a Boot Drive set up (I just found all my old HDT versions and OS media) the Rocket will ride again.

After I test the setup with mine, IIfx's loaded Rocket will go in there and I'll hook up the VideoVision Studio Array box of SCSI II RAID Goodness! [}] ]'>

These may be overkill, but they may work, the price was right and they arrived a couple of days ago! :approve: