68kMLA Classic Interface

This is a version of the 68kMLA forums for viewing on your favorite old mac. Visitors on modern platforms may prefer the main site.

Click here to select a new forum.
jmacz journey
Posted by: cgp on 2025-06-08 21:10:48
The dark spots are the carbon residue after evaporust reacts with the rust. It comes off with scotchbrite.
You know, there’s quite a variety of scotchbrite products, which type are you referring to here?
Posted by: jmacz on 2025-06-08 21:25:05
You know, there’s quite a variety of scotchbrite products, which type are you referring to here?

Not sure the name exactly, but it's their green heavy duty pad. Normally most of the black spots come off with some slight rubbing with the pad.

But for this LC III PSU, not sure it's going to work as it's pretty bad. I might have to resort to sanding it.
Posted by: obsolete on 2025-06-09 07:47:44
On a second look, I think it's fine... it's still the usual yellow color, it's just there's rust particles which is making things cloudy. I usually pour the stuff back into the bottle through a filter. But in this case, since I did use it on a battery bombed chassis, I'm going to take it to the hazardous disposal center and buy a new 1 gallon bottle.
I would hang onto it; it might be fine. I was pretty lazy and didn't scrape off much of the battery residue. Just don't mix it back into a larger container. I'm actually interested to hear whether a similar reaction happens in your case.
Posted by: jmacz on 2025-06-09 09:27:08
I would hang onto it; it might be fine. I was pretty lazy and didn't scrape off much of the battery residue. Just don't mix it back into a larger container. I'm actually interested to hear whether a similar reaction happens in your case.

We'll see. I only had a single 1 gallon of it so this is essentially all that I have. Maybe I'll mark the bottle and keep it for heavy stuff and ensure I have a more pristine new 1 gallon for lighter things. Still don't smell anything so maybe it's fine.
Posted by: jmacz on 2025-06-09 15:49:16
Top part of PSU and bottom shield out of the bath after a few days. Not bad.

IMG_0464.JPG

This is what the shield looked like.

IMG_0432.JPG

And the PSU top:

IMG_0431.jpg

The bottom part of that PSU was a disaster... needs more time in the evaporust bath. It's getting there, about 2/3rds of it is gone but still have some work. The bottom shield is also almost ready. The big chunk in the picture immediately above on the shield is gone. Just need to get it off the sides.
Posted by: CC_333 on 2025-06-10 10:00:03
Wow, that evaporust really does wonders!

I wanted to get some a couple years ago, but I couldn't find it anywhere.

c
Posted by: jmacz on 2025-07-09 17:56:52
I need to get back to that LC III and the 840AV, but I have been distracted.

@cheesestraws pointed me at a writeup that @Phipli did on upgrading his Radius Rocket 25i to a Rocket 33:


Decided to give it a try.

I didn't have any spare socketable full 68040s at 33MHz or higher lying around so I decided to use a QFP to PGA adapter designed by @zigzagjoe that I received from @croissantking in exchange for something recently. The adapter was already populated with a Freescale L88M mask 33MHz 68040 (QFP) that @croissantking had saved from death (it had lost some pins but he was able to save it by shaving some of the packaging in the corners - I think he mentioned it on one of his profile posts a few months back). I picked up a 66.666MHz crystal and burned a 27C256 EPROM with the ROM image for a Rocket 33 which @Bolle had provided in another thread (link here).

IMG_1082.JPG

Here's the Rocket with new ROM, the adapter which amazingly fits (I think it will probably clear the add on cards too), and the new crystal (I added a socket so I can easily reverse the change). You can see the two corners that were modified by @croissantking to save the chip. And here with a heatsink added:

IMG_1083.JPG

With the heat sink it will actually prevent the next Nubus socket from being usable. The Freescale chips run a bit cooler but I might add a fan as @zigzagjoe 's adapter conveniently has pads for providing 5V to a fan. I have a 40mm Noctua lying around which I can add once I design a clip that will secure it to the adapter board (I had a clip designed previously for socketed cpus, but since this one is not socketed, won't work).

The modification seems to work. The card does identify itself as a Rocket 33 and I was able to run System 7.1 on it via RocketShare. Seemed stable as I was messing with it. I did not benchmark to prove it's faster. But I assume with the new clock crystal and ROM, it's running at 33MHz.

I haven't decided whether to keep the adapter on there permanently or pick up a new 68040 chip. I kinda don't want to lose a Nubus slot. But then again, I have nothing else I want to put into my IIfx so maybe I'll just leave it for now.

Thanks to @cheesestraws, @Phipli, @zigzagjoe, @croissantking, and @Bolle for the pointers (and ROM image).
Posted by: zigzagjoe on 2025-07-09 19:56:36
You probably could get away with just a sheet of aluminum to act as a heatspreader in order to get the low profile. 33mhz isn't asking much, around 3ish watts to dissipate on a L88M.
Posted by: jmacz on 2025-07-09 20:56:10
You probably could get away with just a sheet of aluminum to act as a heatspreader in order to get the low profile. 33mhz isn't asking much, around 3ish watts to dissipate on a L88M.

Ok, will consider that, thanks!
Posted by: croissantking on 2025-07-10 05:35:38
I need to get back to that LC III and the 840AV, but I have been distracted.

@cheesestraws pointed me at a writeup that @Phipli did on upgrading his Radius Rocket 25i to a Rocket 33:


Decided to give it a try.

I didn't have any spare socketable full 68040s at 33MHz or higher lying around so I decided to use a QFP to PGA adapter designed by @zigzagjoe that I received from @croissantking in exchange for something recently. The adapter was already populated with a Freescale L88M mask 33MHz 68040 (QFP) that @croissantking had saved from death (it had lost some pins but he was able to save it by shaving some of the packaging in the corners - I think he mentioned it on one of his profile posts a few months back). I picked up a 66.666MHz crystal and burned a 27C256 EPROM with the ROM image for a Rocket 33 which @Bolle had provided in another thread (link here).

View attachment 88680

Here's the Rocket with new ROM, the adapter which amazingly fits (I think it will probably clear the add on cards too), and the new crystal (I added a socket so I can easily reverse the change). You can see the two corners that were modified by @croissantking to save the chip. And here with a heatsink added:

View attachment 88681

With the heat sink it will actually prevent the next Nubus socket from being usable. The Freescale chips run a bit cooler but I might add a fan as @zigzagjoe 's adapter conveniently has pads for providing 5V to a fan. I have a 40mm Noctua lying around which I can add once I design a clip that will secure it to the adapter board (I had a clip designed previously for socketed cpus, but since this one is not socketed, won't work).

The modification seems to work. The card does identify itself as a Rocket 33 and I was able to run System 7.1 on it via RocketShare. Seemed stable as I was messing with it. I did not benchmark to prove it's faster. But I assume with the new clock crystal and ROM, it's running at 33MHz.

I haven't decided whether to keep the adapter on there permanently or pick up a new 68040 chip. I kinda don't want to lose a Nubus slot. But then again, I have nothing else I want to put into my IIfx so maybe I'll just leave it for now.

Thanks to @cheesestraws, @Phipli, @zigzagjoe, @croissantking, and @Bolle for the pointers (and ROM image).

Very nice to see that chip being so proudly resurrected. And that is rather serendipitous regarding fitment!

BTW, I see that the decoupling cap at C7 is missing - I would suggest replacing that. I fitted 100nf passives on the backside so you’ll need something like a 100uf cap on the top.
Posted by: zigzagjoe on 2025-07-10 06:23:01
Very nice to see that chip being so proudly resurrected. And that is rather serendipitous regarding fitment!

BTW, I see that the decoupling cap at C7 is missing - I would suggest replacing that. I fitted 100nf passives on the backside so you’ll need something like a 100uf cap on the top.
That was at my direction, C7 is for decoupling when the jumper above is cut for current measurements/3.3v operation/other shenanigans. As long as the bottom caps are populated it's not essential as there should also be plenty of capacitance near the 040 socket on the logic board.
Posted by: croissantking on 2025-07-10 06:24:51
That was at my direction, C7 is for decoupling when the jumper above is cut for current measurements/3.3v operation/other shenanigans. As long as the bottom caps are populated it's not essential as there should also be plenty of capacitance near the 040 socket on the logic board.
✌️ 👍
Posted by: croissantking on 2025-07-10 06:36:46
I need to get back to that LC III and the 840AV

Looking forward to seeing how you go with the 840AV. When I get back from holiday I’ll pick up on my PSU rebuild, quite excited to see how that turns out.
Posted by: jmacz on 2025-07-10 08:35:38
Looking forward to seeing how you go with the 840AV. When I get back from holiday I’ll pick up on my PSU rebuild, quite excited to see how that turns out.

Thanks! Need to find some time. I'm looking forward to your PSU rebuild 🙂
Posted by: jmacz on 2025-07-10 08:37:29
Very nice to see that chip being so proudly resurrected. And that is rather serendipitous regarding fitment!

What's crazy is that the edge of the adapter PCB that is over the Nubus connector actually perfectly rests on top of the connector plastic (no potential of short underneath either). It's almost like @zigzagjoe saw into the future and designed it perfectly for this application. 🙂
Posted by: jmacz on 2025-07-10 12:36:56
Complete overkill based on what @zigzagjoe said, but decided to just finish what I started since it was already printed.

IMG_1088.JPG

With fan and retaining clip installed. Note that the angle of the picture makes it look like the clip extends beyond the nubus connector but that is not the case. No clearance problem.

Will leave this for now until I want the slot back at which point I will go with the low profile heat spreader as suggested or go with a socketed CPU.
Posted by: joshc on 2025-07-10 18:42:13
That’s turned out looking very neat, nicely done!
Posted by: olePigeon on 2025-07-15 19:57:30
@jmacz Nice! That is slick looking. 🙂

Any plans to attempt the first 50Mhz Rocket?
Posted by: jmacz on 2025-07-15 20:57:01
@jmacz Nice! That is slick looking. 🙂

Any plans to attempt the first 50Mhz Rocket?

I guess I could try 🙂 Next stop is 40MHz as I have an 80MHz crystal around here somewhere. I guess we'll see.
Posted by: jmacz on 2025-07-19 17:34:15
Looking forward to seeing how you go with the 840AV. When I get back from holiday I’ll pick up on my PSU rebuild, quite excited to see how that turns out.

The 840AV on further testing wasn't working... it was failing to boot fully and sometimes wouldn't power on. I have since recapped the logic board, recapped the PSU, cleaned the logic board, removed the heatsink, cleaned off the old thermal paste, applied some new paste, put the heatsink back on, and then tried it today. Consistently works now and boots! I also removed all the rust from the metal shielding using evaporust. Now I just have to fix the broken plastic inside and print a new bezel for the tray loading CD... oh that reminds me, I need to recap the CD-ROM drive. I have extra 1.4MB auto inject floppy drives so I will probably replace the manual inject one that came with this machine.
< 19 >