• Hello MLAers! We've re-enabled auto-approval for accounts. If you are still waiting on account approval, please check this thread for more information.

LC 475 Project - Upgrades - Mods

I used to have an LC 475 with 132MB of RAM. This was the solution to get rid of the annoying boot delay. It also works in a beige PowerMac G3 that hits the RAM ceiling.

A button to disable memory test does not appear when opening Memory control panel with Option and Control key pressed under System 7.5.5.

Are you running 8.1?
Is change persistent, when starting with another OS version?
 
Designed my own solution today.

Nice work. A blower would probably work better in that setup than an axial fan.

In my 475, I used a 60x25mm Noctua fan with the frame cut out. At the time, Noctua didn't have 60x15, but the NF-A6x15 is available now.
 
I think the change should be persistent, as most likely it has to be stored in the PRAM. I made the screenshot recently in SheepShaver with Mac OS 9.1. I have used the same feature on an LC 475, for sure not with 9.1. It might have been 8.1, but I still am a big fan of System 7.1. The latter runs very snappy on any 68040 Mac. You might try other key combinations to invoke the hidden feature from the »memory« control panel (like just an depressing »option« on opening the control panel, or similar). Sometimes it also works to transplant a control panel from an older or newer version of Mac OS into your System. On my computers with System 7.1 I always install the newest available version of the LaserWriter 8 software. I have tried several instances of Franken-Systems with several Apple system software components mixed from different generations. The only issue I ever ran into, apart from the software not running at all, is a little quirk in System 7.1, where the file dialogue does not show folder icons properly in remotely added folders when accessed via AppleTalk.
 
I think the change should be persistent, as most likely it has to be stored in the PRAM. I made the screenshot recently in SheepShaver with Mac OS 9.1. I have used the same feature on an LC 475, for sure not with 9.1. It might have been 8.1, but I still am a big fan of System 7.1. The latter runs very snappy on any 68040 Mac. You might try other key combinations to invoke the hidden feature from the »memory« control panel (like just an depressing »option« on opening the control panel, or similar). Sometimes it also works to transplant a control panel from an older or newer version of Mac OS into your System. On my computers with System 7.1 I always install the newest available version of the LaserWriter 8 software. I have tried several instances of Franken-Systems with several Apple system software components mixed from different generations. The only issue I ever ran into, apart from the software not running at all, is a little quirk in System 7.1, where the file dialogue does not show folder icons properly in remotely added folders when accessed via AppleTalk.
I tried all combinations of Control, Shift, Option and Command while opening Memory control panel of System 7.5.5: No additional Choice for Startup appeared. I had a look into resource fork as well, no text or Icon found that would fit.

I will install System 8.1 on another partition later to test.
 
Now there is System 8.1 installed. Monitor control panel that came with System 8.1 is Version v7.5.9.
I tried all combinations of Control, Shift, Option and Command while opening Memory control panel: No additional Choice for Startup appeared.

It is quite probable that the screenshoted Version D1-v8.1.1 is from System 9 or newer.
System 9 does not run on 68k Macs. And I assume the components of System 9 does neither.

So it is most probable that Power Macs (or emulated Power Macs) are able to bypass memory test but 475 is not.

@register Do you have any way to reproduce the behaviour on a 68k Mac or emulated 68k Mac?
 
Now there is System 8.1 installed. Monitor control panel that came with System 8.1 is Version v7.5.9.
I tried all combinations of Control, Shift, Option and Command while opening Memory control panel: No additional Choice for Startup appeared.

It is quite probable that the screenshoted Version D1-v8.1.1 is from System 9 or newer.
System 9 does not run on 68k Macs. And I assume the components of System 9 does neither.

So it is most probable that Power Macs (or emulated Power Macs) are able to bypass memory test but 475 is not.

@register Do you have any way to reproduce the behaviour on a 68k Mac or emulated 68k Mac?
My understanding is, without an actual upgraded ROM (solder in the slot if you need to) you CAN NOT bypass the memory test.
 
In my 475, I used a 60x25mm Noctua fan with the frame cut out. At the time, Noctua didn't have 60x15, but the NF-A6x15 is available now.

Just wanted to drop-in to confirm that the NF-A6x15 FLX is almost a drop-in replacement for the original, and a good choice. It's got around 25% more airflow and less objectionable noise. Looks like Apple used the lowest speed of the range of fans available in my Q605 (HDF6015L-12LB). So they weren't expecting this case to move a lot of air.

You do need to swap the pos/neg pins in the plug, but some flush cutters to trim the heat shrink back and a SIM eject tool (or other similarly small screwdriver bit) is all you need.

Nice work. A blower would probably work better in that setup than an axial fan.

Considering these CPUs relied on case airflow to pull heat from the heatsink fins, a blower seems like overkill unless you are going to really push the machine.

That said, I'd be very curious about a before/after with these ducts with measurements. With the case being designed around negative pressure, you are going to get a decent chunk of flow over the heatsink so long as the lid is installed. My assumption would be that with the lid on, the benefit would be much smaller than testing with the lid off. Might be a good reason for me to finally grab a thermal imaging camera perhaps.
 
I just got 2 XC68040HRC25M, Mask E31F from Vesalia on eBay for 25€ each. They easily overclock to 40MHz and i could run both at 46MHz without fan (55°C with heatsink). They feel cooler than the RC40 of the same Mask i have used previously. That RC40 does barely 45MHz and crashes at 46MHz after a while with heatsink and fan. I think these 68040 are not too expensive compared to the other offers found on eBay.
 
Be aware though that they have a mixed inventory with mostly RC25 chips… I think I only got one or two HRC25 out of all the chips I bought from them so far.
 
I would not attribute anything special to HRC chips as there's no definitive source (ie. Motorola) stating what it means. At best I have only ever found apocryphal statements that they run cooler, and those claims are best regarded the same as the myths surrounding L88M mask chips.

Motorola did use HRC suffix on some microcontrollers, on those it simply indicated a variation allowing for the use of a RC oscillator. Doubt that applies here, but I'd think it's a good indication that HRC could mean anything. HRC chips seem to have been used interchangeably with standard RC chips on 25mhz 68040 machines - I received a Centris 650 with one installed. At least in my experience, the HRC I have doesn't differ from a standard chip @ 40mhz.

Motorola doesn't seem to have updated their datasheets after a certain point; I can't even find any official documentation acknowledging that full 68040s were made in QFP package or defining the pinout. PCNs acknowledge that HRC existed, and also "BRC", and also note some some interesting suffixes like D, E, M. I've also received some late PLCC Freescale 68882s with an EI suffix which I have *no idea* what it actually means. Safe to say it all means something, but unless you can find someone who worked on the 68K teams at Motorola in the 90s who could reliably define it, I would treat them all as standard members of their respective families.
 
That said, I'd be very curious about a before/after with these ducts with measurements. With the case being designed around negative pressure, you are going to get a decent chunk of flow over the heatsink so long as the lid is installed. My assumption would be that with the lid on, the benefit would be much smaller than testing with the lid off. Might be a good reason for me to finally grab a thermal imaging camera perhaps.

I somewhat answered this on my own by printing out one of the ducts and trying it. At least with my Noctua, I notice a definite drop in airflow with the duct installed. So while it is more directed, now I'm wondering if it might be a bit of a placebo.
 
Back
Top