Caps! -- LC I / II / III

[uniserver]

ok * UPDATE *

Per this thread:

viewtopic.php?f=29&t=20936

The LC-III Has an ERROR on the silkscreen.

What is labeled for C22 is wrong, It's backwards.



So especially with tantalums, because they can catch fire.

Install your cap the opposite polarity of what is listed on the silkscreen for C22 (only).

or just use ceramics for the whole thing, that way polarity is not a concern any more.

viewtopic.php?f=29&t=21044

 

[uniserver]

Also one more tidbit with the LC-I

I added ram to a LC-I from another bad LC-I, Populated the empty pads. (used my new heat gun)

Afterwards, all i got when powered on, was the chime of death and no video. Removed the ram chips I added, and it started to work just fine.

I am wondering if maybe trash80 is right and there is something with Addressable space and Firmware in ROM.

So………. Being the relentless guy that i am …

I popped those LC-I rom chips out, installed LC-II rom's and it works!, So then i used the heat gun again and re-installed those other ram chips. and the chime of death, Again… So took those ram chips back off, popped the LC-I rom's back in … made it like how it was, and put it back on the shelf.

Enough fun for one day

So to recap, adding more ram chips to populate the empty pads with my LC-I didn't seem to work.

I even inspected the bottom of the main board.

To see if maybe there were some missing resisters or caps? (to maybe save on parts, because of the un pouplated ram pads)?

It looks like apple still placed caps there for all those empty pads anyways.

It's strange, because I really thought it would work, You know, just show 4mb ram instead of only 2. but no dice.

 

[olePigeon]

Awe shucks. I was wondering about those unused pads.

 

[James1095]

They may have put those pads there to enable the use of 2x as many chips of half the capacity. It's also possible that they were in anticipation of future production upgrades, quite possibly requiring a different memory controller chip or other changes. You might try tracing out the circuit. I suspect that the address and data lines will all parallel with the original RAM so it may not be all that difficult to see where the enable pins go.

 

[MinerAl]

You said the RAM came from a bad LC. Could it be that the RAM was bad?

 

[Trash80toHP_Mini]

Have you compared Memory Controllers/ASICs between the the LCs I, II & III? There will be a block diagram in the DevNotes for each model that lists the major components along with the memory mapping of each. Because the LC will only address 10MB of RAM on its measly 16MB bus, I'm suspecting that the memory controller in the LC is well and truly hobbled or that the RAS lines etc. were left out of the PCB circuitry.

To add insult to injury, Apple programmed the ASICs to support no more than 10 MB of RAM even if more was installed.

Yep, I figured as much, but I'd check out patch wiring the addressing signals first. Swapping the ROMs was a great first test, this leaves room for hope, but you're probably going to need to try the LCIII's ROM to get it over 10MB, as the LCII is limited to 10 MB as well.

< WAG mode >

If LEM is correct, the lines may already be there and the ASICs are the limitation and cause of Apple's carpet bombing campaign against you, not the ROMs and not necessarily the PCB layout.

Just checked the LC II specs and it's not lookin' good. Are the LC and LCII block diagrams and board layouts as similar as I'm suspecting?

If they're the same board, Apple just changed the function of the ASICs to address the RAM on those empty pads, but not more than 10MB in total for the model bump, bat rastards! >

If the ASIC of the LCIII containing the Memory Controller is pin compatible with the ASIC of LC and LCII, swapping that and the ROM (if possible, check the sizes/pin counts of the ICs first) may give you the remotest possibility of hacking a fully addressable 16bits of memory, but because of the hobbled memory bus, you're likely not going to see all that much improvement in speed.

When Apple designs a slug, most often it remains a slug. :-/

< /WAG mode >

I hope this makes some kind of sense, apparently I'm bleary enough that I just ran the coffee maker with the water still in the pot again. :

________________________________________________________________________________________________

edit: Yep, coffee helps a bit. :approve:

Similar discussion: There're 8 RAM chips soldered to my Quadra 605's LB...

Definitely do the DevNote R&D first. A two MB bump to the LC just gives you the LC II's 2MB of wasted memory when you put an 8MB SIMM in the slot. It looks to me like the only bang worth a buck for the LC & LC II would be an LCIII ROM/ASIC swap and I'm guessing the ROM packages are different as well as the pin counts of the Memory controller ASIC.

RoadApple hacks ain't something Apple made easy. :disapprove: