BGE's take on the Quadra 900/950 ATX PSU Mod

[BadGoldEagle]

Success! The switch fits the opening!



It almost looks like it's meant to be there from the factory.

There's only one small inconvenience: the tabs on the switch are slightly narrower compared to the ones Apple/Delta used. In order to keep things as simple as possible (and to prevent permanent modifications), wires will be soldered directly to the terminals. Spade connectors (4) will then be used to connect the switch to the AC board.

 

[cheesestraws]

That looks really, really nice.

 

[ktkm]

I forgot to say it looks terrific! I hope I have not missed anything? @BadGoldEagle, what is the status?

 

[joshc]

I somehow missed the last update. This is exactly what I want to do with mine (add a switch) - for now I left that hole empty after doing my conversion. Keen to hear how yours is going.

 

[WillJac]

I don't see any further posts on this great project from @BadGoldEagle. Is he still on the forum and what about the project? The GitHub has no updates for 2 years. June 26, 2022 was the last time he was on the forum.....I would like to build this but?????

 

[Joe_Goes]

Guys where can I buy these QuadrATX boards fully populated if possible? Desperately need to put my Quadra back in working order. My original psu made a loud pop sound and tripped by circuit breaker

 

[WillJac]

Guys where can I buy these QuadrATX boards fully populated if possible? Desperately need to put my Quadra back in working order. My original psu made a loud pop sound and tripped by circuit breaker

I have some boards but need to see if I have enough of the parts to make a second one. I made one for myself. But I do have to tell you, the parts not easy to find and are expensive compared normal parts. I am still waiting for 2 items for me to test mine. The parts I waiting for see not for the boards.

I don’t know what happen to the creator of the project but I hope to restart it if can.

 

[Joe_Goes]

I have some boards but need to see if I have enough of the parts to make a second one. I made one for myself. But I do have to tell you, the parts not easy to find and are expensive compared normal parts. I am still waiting for 2 items for me to test mine. The parts I waiting for see not for the boards.

I don’t know what happen to the creator of the project but I hope to restart it if can.

I myself am interested to buy two sets of boards just incase. This project is certainly worth it to be restarted not only because I need a psu for my machine but think of all those Quadra out there with a ticking bomb of a psu which can go off without any warning and take the machine with it in most cases. Please keep me updated on the assembly of these boards Will and thanks for the reply.

 

[WillJac]

I myself am interested to buy two sets of boards just incase. This project is certainly worth it to be restarted not only because I need a psu for my machine but think of all those Quadra out there with a ticking bomb of a psu which can go off without any warning and take the machine with it in most cases. Please keep me updated on the assembly of these boards Will and thanks for the reply.

This what I have already done. I am waiting on 1 final part to arrive to test and get back on it. I have loads other items going on so can be a few weeks before I get to it.

 

[Joe_Goes]

This what I have already done. I am waiting on 1 final part to arrive to test and get back on it. I have loads other items going on so can be a few weeks before I get to it.

Wow dude that looks really nice, do you mind if I can trouble you to build one for me? Let me know how much it’ll cost?

 

[WillJac]

Wow dude that looks really nice, do you mind if I can trouble you to build one for me? Let me know how much it’ll cost?

Let me test mine and be sure it works before I make it available for others. The parts are hard to source due to shortages and some I had to pay much higher reprice for. You seeing over $100 there just on parts. Just so you know!!

 

[Joe_Goes]

Let me test mine and be sure it works before I make it available for others. The parts are hard to source due to shortages and some I had to pay much higher reprice for. You seeing over $100 there just on parts. Just so you know!!

No worries Will, just let me know when you are ready to build as currently there aren’t any option for people like me with a blown original PSU. Pencil me down for two sets for sure.

 

[BadGoldEagle]

Notifications are on but I wasn't made aware of any activity.

Happy to see someone else working on this project!

One thing to keep in mind @WillJac
PLEASE double check the pinout on the Quadra side before you plug it in! The cables were meant to be soldered without a header and I am pretty sure this version of the board isn't compatible with ATX extension cables, something I changed in 2.0...

It might be fine as you have a female plug (what kind of extension cable did you find? Male <=> Male is quite tricky to find...). 2.0 has a male receptacle, for one of those:





******************************************************************************************************



Some Pics of the current boards (I still have to figure out the molex plugs that make reassembly tricky and the load resistors (which are a bit too hot for my liking). In other news, the AC Board was simplified a lot.

DC 2.0 Dev


AC 2.0 (not to scale obviously, almost done, J7 might get removed)


Currently thinking about using one of those secured to the original PSU's case. They're made for up to 12W in air and 25 W with a heatsink. The mounting solution is my main issue ATM as I don't want to drill any more holes in the case.

 

[BadGoldEagle]

DC Board Update

Wirewound resistor (Digikey Link, I just bought one from Farnell, also available on Mouser but not on Arrow at the moment)
+ Can be placed anywhere (not only on the PCB itself). Ideally close to the fan where it does not interfere with the PSU's heatsinks.
+ Should be way over dimensioned. Which means it should not overheat! Dissipated power for 12V => 7,2W, capable of up to 25W with a proper heatsink or 12W in the air...
+ Can be secured (needs to be verified!) with a thermal adhesive instead of screws and compound. I've found this on Amazon in Germany.

I have ordered the parts and will test them with a power brick. Should not heat up that much if I'm not mistaken.
Also added a couple of holes for "cable management" -- a zip tie should be able to keep that mess of molex cables "tidier"



Being sick I can't move much from the laptop... but I've found a few pics that confirm that J7 and J8 were indeed redundant on the AC Board.
New design:

NB: The new tab where the Steve Jobs face is located prevents the board from rotating once installed in the case (because there is only one mounting hole (Earth), I needed a bit more friction.

 

[WillJac]

DC Board Update

Wirewound resistor (Digikey Link, I just bought one from Farnell, also available on Mouser but not on Arrow at the moment)
+ Can be placed anywhere (not only on the PCB itself). Ideally close to the fan where it does not interfere with the PSU's heatsinks.
+ Should be way over dimensioned. Which means it should not overheat! Dissipated power for 12V => 7,2W, capable of up to 25W with a proper heatsink or 12W in the air...
+ Can be secured (needs to be verified!) with a thermal adhesive instead of screws and compound. I've found this on Amazon in Germany.

I have ordered the parts and will test them with a power brick. Should not heat up that much if I'm not mistaken.
Also added a couple of holes for "cable management" -- a zip tie should be able to keep that mess of molex cables "tidier"

View attachment 53551

Being sick I can't move much from the laptop... but I've found a few pics that confirm that J7 and J8 were indeed redundant on the AC Board.
New design:
View attachment 53553
NB: The new tab where the Steve Jobs face is located prevents the board from rotating once installed in the case (because there is only one mounting hole (Earth), I needed a bit more friction.

I like this very much and thanks for the replies to my messages. I will wait for you to have a final working version that you are satisfied with before I order boards. I have most of the parts and seems you have removed some items. This makes it great. My concern, in my PM to you, was about the other J connections as I could not see where they would go but now I see that they are all removed. I have a few of the previous version boards but I will dispose of those as they are not really valid anymore with the issues you have solved in this new version.

 

[ktkm]

@BadGoldEagle @WillJac I’m so glad this thread is moving again! I can’t wait to order a kit or a pre-assembled board.

 

[BadGoldEagle]

Okay, so the resistor is heating up quite a bit and since space (or rather lack thereof) is a problem, the only safe place to stick it in place would be at the back, where the exhaust is... But even a Noctua industrialPPC at full speed will not reach that far.

So we would need another one like so:


Front would be the intake, where the original fan is. The back one (push config) would work as an exhaust. This seems to be the only viable option... If you have a better one, let me know!

This new fan would be powered from the same source as the front one (ATtiny/MOSFET, using a splitter), its speed (in terms of PWM pulses) would be the same and controllable with the potentiometer.

But all of this is starting to look like a wild goose chase to me... Adding a fan just to cool down a load resistor? Really? Since I do not plan on drilling more holes in the case, the said resistor would have to be stuck to the casing using thermal adhesive. I am not sure to be able to cool it down efficiently without the fan (and rely solely on passive cooling).

Then I got an idea, a pretty dumb one but as Derek from Vice Grip Garage often says, if it works is it really a dumb idea?
The idea being reusing the original Sanyo fan, which should still work even if your PSU has already died. Its consumption is exactly what is needed to keep the new PSU happy: 500mA, 12V.

So here's the thing: It will not have an impact on the project as both options will still be possible, I am simply going to pick one for myself (I already have) but I'd like to have you guys feedback on this...

Can you have a look at this poll and give your opinion? Much appreciated!

Have your say: Quadra 9x0 Cooling

Voice your opinion on the poll: Quadra 9x0 Cooling

take.quiz-maker.com

 

[robin-fo]

Since most modern ATX PSUs regulate themselves based on what they see on the 12V rail, it is important to get it right.

- Without the load resistors, the 12V rail was too high (12.5V). The PSU was clearly having issues with that rail and the ATX fan was on at full blast.

- With some load resistors (2x 24Ohm 7W), the 12V rail sits at 12.25V (much better) and the fan is running at the lowest RPM setting BUT some smoke was coming out of the resistors… Not good!

Turns out that was caused by a miscalculation on my part. It’s only by doing mistakes that you learn!

Basically, I wanted to use two resistors in parallel and forgot to adjust the resistance.

Are you really sure a resistor is required? 12.5V lies within Apple's PSU specification (at least for the Mac II series). You could save a yourself a lot of trouble and energy The voltage will certainly drop when you put two 5.25" hard drives in

 

[BadGoldEagle]

Hi @robin-fo !

It will work without it but I noticed that the Bicker PSU was having to work harder (and therefore hotter) in order to keep things going. So for stability/longevity reasons, it's better to have a load resistor when using newer fans...

There are only four scenarios:
- Solid State Storage (SCSI2SD, BlueSCSI, PiSCSI etc) + Noctua fans => Load resistor necessary
- Solid State Storage + original fan => All Good. This is going to be my setup + DOS Card.
- SCSI HD from the 90s + Noctua fans => Should be good to go.
- SCSI HD from the 90s + original fan => More than fine!

 

[robin-fo]

I just wonder if your observation was correct or was really related to the lack of power consumption. I read in the description of the Bicker 630H:

Alle ATX-Ausgänge inkl. eines zusätzlichen -5V-Ausgangs benötigen keine Grundlast, so dass Mainboards der neuesten Generation sicher und zuverlässig versorgt werden können.

The datasheet also lists all outputs as 0A min output current and I couldn't find any hint that there might me issues with low output current except the efficiency curve, which is cut off for really low output power values but lets us assume the the efficiency will be very moderate.

I even quickly asked my workmate today (who is experienced electrical engineer) and he also didn't have an idea why there might be an issue, especially in higher quality PSUs.

Are there other reasons why a resistor might be necessary besides your observation?