• 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.

PowerBook 540c Restoration Project

3lectr1cPPC

Well-known member
Excellent work once again... I'm blown away at how good a job you've done with all this.
This stuff is all getting linked on the resource pages on my site for the 500 series laptops for sure!
Do you mind if I host the stl files directly on the site? Credit where due of course will be given.
 

jmacz

Well-known member
Excellent work once again... I'm blown away at how good a job you've done with all this.
This stuff is all getting linked on the resource pages on my site for the 500 series laptops for sure!
Do you mind if I host the stl files directly on the site? Credit where due of course will be given.

Nope, don't mind, by all means go ahead. And if anyone finds any issues with the models, let me know and I can make changes.
 

Dogmander

Active member
One last piece on this project... the front bezel was cracked and will not go back on properly as the little teeth were broken. You can see the missing front bezel here:

View attachment 61959

It's where you can see the mic and ribbon cables. :) So of course, needed to make a new replacement.

View attachment 61960

Just like the battery slot cover, the pins were too detailed here and my 3D filament printer just wasn't good enough to print it out. Since I needed to use my 3D resin printer, I needed to split the model into two parts because the printing volume for my 3D resin printer is too small. Here are the parts printed:

View attachment 61961

And here's the two pieces glue together with epoxy and installed.

View attachment 61962

One of these days I will get someone with a larger printer to print a one piece one for me so I can avoid the seam in the middle. I could have filled/sanded but left it for now.

Attached are the STL models for this bezel cover. I have included both a left/right two piece version as well as a one piece version.
How do you model stuff like that? I want to learn how to do complex parts like that for my own Powerbook
 

jmacz

Well-known member
How do you model stuff like that? I want to learn how to do complex parts like that for my own Powerbook

I've been using Fusion 360 and it's just time/practice. Fusion's not as great if you have more freeform arbitrarily curvy/blobby objects (I think something like Blender might be better for that) but for more engineering/technical models, it's very intuitive and easy, because Fusion 360 is a parametric modeler (you design things with dimensional input parameters and if you get something wrong, you go back to change that one dimension and the entire model -- if you designed it right -- updates automatically so you don't have to redo the entire design). As long as you can think in terms of 2D layout in X/Y/Z, and have a good set of calipers to measure with sub-millimeter precision, it's very straightforward. Lots of tutorials on YouTube if you want to get started. Many years ago, I found that some of the early Fusion 360 tutorials from NYCCNC were easy to follow and covered a lot of the basics.

There are also a lot of assumptions you can make to get the model close to where you need it to be, after which you can further dial it in. By assumptions what I mean is that the original engineers also relied on circles, ellipses, rectangles, etc so rarely do you find arbitrary curves for these types of parts. They also usually aligned against solid metric or imperial values (ie. millimeters or half of millimeters... ie. 3.0mm, or 3.5mm, rather than 3.38mm) so that makes getting things close pretty easy. For example, if you measure and you get 3.03mm, it's probably 3mm. Within a few measurements, you can figure out if a part was designed in metric or in imperial (american standard). So if imperial, and you get like 0.241 inches, it's probably 0.25in or 1/4 inch. Same with any holes. They are almost always standard sizes. So you're usually looking at holes designed for M2, M2.5, M3, M4 sized screws. You don't see random sizes. This all helps.

After that it's really spending the time to measure against the right dimensions, the right anchor points for parametric design so that if you get something wrong and need to update that one dimension later after you test fit, that every other dimension automatically adjusts based on that change later. And that comes with experience/time/practice.

Anyhow hope this helps. CAD is really fun and to be honest, I get a lot of enjoyment modeling these parts. It's interesting, it's sometimes challenging, it's like a puzzle.

But it's also a slippery slope... because once you get into CAD, you start itching to get into CAM where you think about manufacturing the part ie CNC machining. Putting together the sequence of cuts with speeds/feeds/tooling via CAM is tons of fun and interesting. I spend time manufacturing parts in aluminum as well. LOTS of fun. I'm by no means an expert or even very advanced. But I think I know enough to at least design parts like this. :)
 

alexGS

Well-known member
Ohhh really? Well no wonder they break like they do. Wow, that's a genuinely terrible design. I thought they did have them, but I guess not.

I'm surprised any are still intact at this point if that's the case.
I was surprised to find my 520’s are intact! Can see them in the photo below near the display connectors (behind the upside-down bezel). I was very careful taking it apart and reassembling. My first task was to clean/restore the battery contacts, and I’ve re-celled a battery which seemed to be working, charged to 100%, ran the machine for more than an hour, but has stopped responding to EMMpathy while it was charging,

This 520 is a parts machine I’ve just acquired for my own 540c restoration project. It’s really like-new inside, which seems often the case for those base-spec machines that no-one wanted ;)

My 540c is on the way to me, I think it has the usual damage so I’ll be making a good unit from two. This passive-matrix greyscale display is not very nice, and someone has scribbled on it with ballpoint pen :( Also has a dark patch that inverts at low contrast, and much ghosting.

Meanwhile I bought a power adapter separately which had no voltage at all - I replaced all seven electrolytic caps and it has come back to life, full strength - so I’d imagine that will need doing to most 500-series power adapters by now
 

Attachments

  • IMG_6047.jpeg
    IMG_6047.jpeg
    2.6 MB · Views: 27
Last edited:

alexGS

Well-known member
Nope, don't mind, by all means go ahead. And if anyone finds any issues with the models, let me know and I can make changes.
Your work is amazing - my reply was shown at the bottom of page 2 in my browser (Safari on iOS) and so I didn’t even notice the other pages! :eek:

The power plug is brilliant, I almost regret that I was able to repair my power adapter.

I especially like your battery bay cover and the display access cover, I’m going to try filament-printing those myself (might have to see if a ‘fuzzy skin’ works). Thank you for sharing these with us.

I use DesignSpark Mechanical (Spaceclaim), I couldn’t get my head around Fusion360 at all :)
 

3lectr1cPPC

Well-known member
My 540's were intact as well. I added a huge mass of epoxy around then to stop them from breaking, and they've been fine so far.

Can you post a photo of your 520's screen, showing the problems? Some may be fixable.
 

alexGS

Well-known member
My 540's were intact as well. I added a huge mass of epoxy around then to stop them from breaking, and they've been fine so far.

Can you post a photo of your 520's screen, showing the problems? Some may be fixable.

Thanks, umm… I don’t have a great photo but this first one shows the area where the contrast is uneven, noticeable on the upper part of the tartan background that someone had chosen. I’m pretty certain it’s the rear polariser film starting to break down. It still looks ok in a pure black-on-white window, though some of the black text can appear a little grey.

I feel bad that we’re now hijacking the thread :) This is only the ‘parts’ machine remember :D
 

Attachments

  • IMG_6031.jpeg
    IMG_6031.jpeg
    4.8 MB · Views: 22
  • IMG_6069.jpeg
    IMG_6069.jpeg
    6 MB · Views: 28

alexGS

Well-known member
Yeah, could be. Hard to tell. A recap will probably help that display out a bit.
Cheers. Yes, one of the faults is a shimmering that affects the greyscale backgrounds (i.e. doesn’t affect black-on-white) which is perhaps capacitor-related.

I think my bigger worry right now is my rebuilt battery (which I’m hoping to use in my 540c) - the readings were odd, especially current (either quite high or very low), and then it just stopped reading at all, with the menu bar displaying the battery with an X, and EMMpathy or Lind utilities no longer recognise that a battery is installed. I’m going to try rebuilding the other battery that I have…
 

3lectr1cPPC

Well-known member
You might want to take it apart again and check that there's nothing off about the wiring in there. Honestly though, I don't have much of an idea of what could be causing that.
 

alexGS

Well-known member
You might want to take it apart again and check that there's nothing off about the wiring in there. Honestly though, I don't have much of an idea of what could be causing that.
As it turns out, that first controller board is faulty. I took apart the second battery and put its board onto the same new cells, and it seems to work perfectly. The utilities still say the memory is faulty and can’t be repaired, but charging and discharging seems to have calibrated it well.

The first board had been reached by the white and blue crystals, so I suspect it has a corroded via. It wasn’t measuring the current correctly (but the shunt was in place)
 

jmacz

Well-known member
Your work is amazing - my reply was shown at the bottom of page 2 in my browser (Safari on iOS) and so I didn’t even notice the other pages! :eek:

Thanks!

As it turns out, that first controller board is faulty. I took apart the second battery and put its board onto the same new cells, and it seems to work perfectly. The utilities still say the memory is faulty and can’t be repaired, but charging and discharging seems to have calibrated it well.

My battery wasn't recognized either. After probing most of the components on the board, ended up finding a mosfet that seemed to be shorted. I replaced that and now the battery is recognized when I insert it. But Lind, EMMpathy, etc all report the memory is corrupted no matter how many times I try to recondition/repair it via those tools. The laptop does work with just the battery, it's just those tools report errors.

My EMM board has four main chips on it... there's a proprietary Apple one (which I assume has the intelligent battery logic in it). Then there's an HC4060A which seems to be a binary ripple counter. Then there's a TI chip (2201AC 52T) - not quite sure what that one is. And then an ATMEL 93C66 which is a memory chip. Not sure if that ATMEL chip holds the memory which the tools are complaining about... but these are readily available (93C66A, 93C66B, 93C66C) so going to try replacing that. There's another thread where someone provided a link to this site:


where you can make a change to the Lind control strip module to force it to overwrite all of the EMM memory with factory values. I'm hoping that means I can replace the entire memory chip and use the tool to rewrite the whole thing. Going to give that a try in a few weeks.
 

alexGS

Well-known member
Meanwhile I fitted appropriate-capacity tabbed AA cells (2000mAh) instead of the 4/3AA 1500mAh cells I was using - the actual measured capacity had been only 488mAh. Then I saw -13mA as the measured battery current, with the measured voltage being correct (9.4V) but jumping to 16.4V when charging began - and measured current was stuck on -13mA.

I’m happy to say the problem was the corroded ends of the ribbon where they attach to the battery cells. The connection must have been failing under load but working just enough to report voltage when not charging. I bypassed these with small pieces of wire (not shown in the photo) and now my charging is correct. Happy days :)
 

Attachments

  • IMG_6100.jpeg
    IMG_6100.jpeg
    1.8 MB · Views: 19
  • IMG_6104.jpeg
    IMG_6104.jpeg
    4.2 MB · Views: 19

jmacz

Well-known member
Got some more time today to tinker with the PB540c's battery. I bought a Microchip 93C66C from Mouser and replaced the ATMEL 93C66 with it. Previously, I was getting uncorrectable errors from all the various tools including EMMpathy and LIND. I connected up the battery again and this time, LIND said the battery memory was totally corrupted (instead of the usual 1-4 errors) and that it would rewrite to factory. Clicked OK and let it do its thing. This time, it wrote and verified successfully! Just to be sure, I powered everything down, started it back up, launched LIND and it found zero errors. Nice.

Still needs some calibration though as the system seems to power down after 30 minutes even though the battery is charged.

Also, I've got some weird issue where after going to sleep, the system won't come back, this is even without any battery and just the power supply connected. After wake up from sleep, the display comes on, but everything just freezes. Happens even with extensions disabled.
 

alexGS

Well-known member
That’s amazing, well done! I’d never have thought of trying a memory transplant for the battery :)
The calibration always seems to take a complete discharge/recharge cycle, or two, for EMMPathy to show the measured capacity in mAh. The default after a reset is 1752, mine seems to have come down to 1680. What’s yours?

Meanwhile I’m having a search to find whether anyone with a working battery has connected VBatt and VMain together (to use a replacement power supply). We’ve had people running the machine off VBatt (instead of a dead VMain) and we have people with USB-C PD-supplied VMain, but I don’t think anyone has run the machine and charged the battery with VMain and VBatt both from a single source?

As for the Sleep problem, have you tried that ‘sleep fix’ extension that I think came with EMMPathy? I‘ll check my machine soon. I also wonder if that might be hard-drive related, since spinning up the drive and waiting for it to become available seems part of waking up.
 
Last edited:
Top