4116 memory For Apple II

[bibilit]

Is there a way to test memory chips in an Apple II Europlus ?

I have probably several bad, where i can get replacement units ?

 

[Gorgonops]

Is there a way to test memory chips in an Apple II Europlus ?

Not really, short of pulling them and putting them in a known working machine. If the computer is working well enough to load some diagnostic software you could try that, there are several memory testers out there for Apple II, but they all suffer from the problem that the computer has to *mostly* work to get the diagnostic loaded in the first place.

4116s are pretty easily located online through the usual suspects. In the US I'd suggest this outfit; I haven't done business with them but they have a good reputation:

https://www.unicornelectronics.com/

In Europe I have no idea off the top of my head, but I'm guessing good sources. And there's always eBay.

It's also pretty trivial to replace 4116s with 4164s (which are still available from mainstream outfits like Jameco) if push really comes to shove:

https://atariage.com/forums/topic/257923-replacing-4116-ram-with-4164-for-reliability/

Obviously it won't look stock, but 4164s are actually significantly kinder to your power supply. One of my Commodore PETs has a bank of 4164s replacing the really oddball 4108 chips (basically 4116s with a bad bank relabled as 8k chips) it shipped with.

 

[bibilit]

Hi, thanks.

ordered a set in the US, cannot find easily in Europe.

The Apple II mostly works, moving chips from one bank to another improved thinks, but not a 100 %.

Any link to a memory tester i can use ?

 

[bibilit]

I have made a copy of MECC Computer Inspector and running the program detected faulty chips.

The message is probably Hexadecimal, but how can i translate the Hexadecimal to a phisical location in the Logic Board (i can translate, but result is not clear to me)

 

[Cory5412]

Keep an eye out during Kansasfest this year, because just the other day, someone on the kfest mailing list asked if anybody has any bad 4116 RAM chips on a motherboard or on a language card or other card that might have those chips.

I think they're working on a software tester or perhaps work-around tool.

 

[olePigeon]

You might have better luck asking on vcfed.org.

 

[Gorgonops]

but how can i translate the Hexadecimal to a phisical location in the Logic Board

What does the output look like? According to the Apple II Reference manual the arrangement of the memory is as so:

• The row of RAM chips closest to the front of the machine has RAM starting at address 0. Next row is the RAM after that, etc. So if, for instance, you have output saying you have a bad RAM cell at location 0x434A, which is 17,226 in decimal, and your machine has the full 48k with all rows equipped with 16k DRAMS (which you'd expect on a Plus; any surviving Apple ][ with 4k DRAMs would be a heck of a find) the bad chip will be in the middle tow.
  
• The leftmost column of chips is the least significant bit. So if the program's output tells you it's the 5th bit that's stuck then it's the fifth from the left.
  

 

[bibilit]

The leftmost column of chips is the least significant bit. So if the program's output tells you it's the 5th bit that's stuck then it's the fifth from the left.

The first part of your message is pretty clear, but this one isn't.

Anyway i managed to find the bad chips biggybacking new on top and running MECC until no more error message was displayed.

 

[Gorgonops]

The first part of your message is pretty clear, but this one isn't.

Do you understand how the RAM in the Apple II is arranged? I thought I was being reasonably clear. In computer-speak the "least significant bit" is the storage element in a byte/word of memory that holds the smallest quantity. IE, a standard "byte" (which conveniently is also the Apple ]['s memory words size) has 8 binary bits capable of storing unsigned integers from 0 to 255. The quantities stored in each bit are, starting from the least significant:

bit 1: 1

bit 2: 2

bit 3: 4

bit 4: 8

bit 5: 16

bit 6: 32

bit 7: 64

bit 8: 128

IE, if you store the number "197" (128+64+4+1) in memory at a given location it's going to be represented by "1"s in bit locations 8,7,3,1 and zeros in 2,4,5,6. The logical arrangement of the 4116 chip is it's a 16k long array of *single bit* memory locations. (IE, each chip is only 1/8th of a full 16kiloBYTEs of ram.) That's why the Apple ][ has three rows of 8 chips.

So... again, I don't know what the output from the MECC program looks like. (If you'd included a screenshot/photo I could possibly interpret it.) I assume it spits out something like: "Error at bit location 0x4567: pattern: 55, read: 45". What this would tell you is the program wrote this out to memory in binary: (I'm putting the least significant bit on the left to match the Apple's physical layout; humans speaking European languages would of course usually write numbers with the LSB on the right)

10101010

But when it read it back it got this:

10100010

From this you'd know, per my example, that the fifth chip over has a bit that's stuck on zero.

 

[bibilit]

I assume it spits out something like: "Error at bit location 0x4567

Yes, that's the kind of message MECC gave to me, to be honest didn't have any picture of it, but you are correct.

Thanks for your reply, pretty clear now.