Hi All,
Apologies in advance for yet another Apple IIc RAM related post! Bit of a story for background and what I've already tried doing.
Background:
About a year ago, I bought myself an Apple IIc on TradeMe, a local auction site here in New Zealand. While I've got a rather large decent sized Mac collection, this was my first venture in to the Apple II lineup.
It worked perfectly on arrival although I only had two 5.25" disks at the time so I couldn't do much with it that I didn't want to have to type in again! Managed to get myself a couple boxes of disks, and boostrapped via ADTPro successfully after making my own serial cable while I waited for my FloppyEmu to arrive. Had great fun playing games with some friends with the matching IIc monitor and connected to the TV using the PAL colour modulator that came with it. Eventually my IIc to VGA adapter arrived so was able to connect it to a more modern monitor as well.
Initial Problem and Fix:
Unfortunately about a month later when I went to play LodeRunner I noticed the disk just kept spinning, and nothing came up on the screen. Without a disk I struggled even to get into BASIC or to Ctrl+OpenApple+Reset. Sometimes I'd get a screen full of blinking characters, sometimes strange colour blocks. From what I was reading it looked like a RAM fault. Having the original ROM I ordered a ROM 4X so I could use the on-board RAM diagnostics as I couldn't boot from disks. Installed that, did the required cut/solder etc. After booting the machine without the keyboard I eventually managed to get a * RAM ZP 11000000 style error code. Installed sockets and replaced a few chips as indicated, and eventually I managed to get it to work again. Was pretty proud of myself! Took it to work one day and a few of us played some games like LodeRunner and PacMan after work one evening. As accountants, we had to have a go with Visicalc too!
Current Problem and Steps Undertaken:
Sadly a few months later I ran into the same problem, disks just kept spinning, the FloppyEmu seemed to get stuck on a track and not move further. Repeated the process I used before and swapped the indicated chip. However, I've swapped this same chip a few times now and it's still producing the same error - *RAM ZP 0000100
I grabbed my oscilloscope out and began probing around. Didn't know what I was expecting to see, but was hoping I might see something that looked obviosuly different compared to other chips. And I did, pins #2 & #14 on the indicated RAM chip (UF8/ADR 3 RAM) just had a flat ~1.8v signal. On the other chips, the same pin gave a regular 5v square wave like pattern.
I've downloaded the IIc schematics, and found that I'm getting the same flat signal from Pin #5 on chip UE11 (74LS245), which is ALT3 on the schematic. This leads to Pins 2 and 14 on chip UF8. The other ALT pins on chip UE11 are giving me the 5v square-wave style I noted on the other RAM pins.
That's got me wondering if I need to replace the 74LS245 chip; I suppose I'm asking if anyone else had similar experiences or suggestions of what to do before I order more a replacement and grab the soldering iron out!
Thanks in advance!
I think you're on the right track, and your troubleshooting looks good to me. This is one of the biggest ways that a scope is a useful troubleshooting aid: it reveals invalid signal levels which can be traced back to defective chips.
The 74LS245 bus transceiver is "exposed" to ESD coming from the I/O slots and is somewhat of a sacrificial part as a result. There are many other examples here on Applefritter demonstrating this, for example.
There is a fundamental problem with RAM tests on Apple II of any generation, although the IIc of the OP did not have that problem.
The problem is that in the Apple II, the viceo buffer resides in the same RAM which is tested.
So if the RAM is badly broken, you can't see the diagnostic output of the RAM test (even when built into the firmware ROM).
Yet another, but related problem is how to load a RAM test into the machine when no such test is built into the firmware.
You would try to load the diagnostics from floppy disk but more often than not you'd get a "disk boot error" or similar (I have too many different computers to remember the exact words the Apple spits out).
This symptom is most prevalent with the IIc ... when people see "disk boot error" with every disk they try to boot, they assume it's a problem with the floppy disk drive but in most cases it's a bad RAM which corrupts the check sum. The consequence of this futile tinkering with the IIc disk drive and electronic surroundings are IIc with missing disk drive latching mechanics or blown up IWMs. The sad remains then may be sold at Ebay. Here is one such case:
https://www.applefritter.com/content/aiic-floppy-drive-doorlatchlever-wanted
When I ran into said problem with some IIc I had, neither would it boot anything nor would the built-in RAM test give a readable message, I coded a RAM test which resides in an EPROM that plug into the upper ROM socket of any Apple II (you would program it into any 2716, 2732, 2764, 27128 or 27256 using a EPROM programmer). It runs a powerful, field proven RAM test (same algorithm the DRAM manufacturers use) and if a bad DRAM is found, it tries to report it via text mode on the screen. If the DRAM is so bad that this does not work, it will tell the bad bit via a side channel observable by an oscilloscope.
Here is a short video (courtesy of Jeff D who sent it to me after he got the DRAM test binary from me) which shows the "optics" of a running test:
IMG_5312.mov
Run it in a loop mode to get the full impression. Note: seems the embedding of videos on Applefritter does not work well enough. This one sticks out to the right (it is not automatically scaled as uploaded still photos are) and when I tried to watch it the video would not run smoothly. So you may need to download it before you can watch it.
If there is enough interest by onlookers, I'd upload the binary (which is easy) but I can't find the source code anymore, so it's not clear how the side channel actually works. This came about the the hurry I was in when I needed that code. All my other projects which span weeks, months, years or decades have a better documentation / archiving policy. But makeshift quickies don't fall under that policy.
Anyone interested ?
Any need for such a DRAM test EPROM seen ?
Comments invited !
- Uncle Bernie
Just a few comments. You can't use a 2716 in an Apple II without and adapter socket. And a better way to indicate a bad RAM chip is to beep the speaker a number of times to indicate its position. All of the ROMX products have such a test and can identify bad RAM as long as the CPU is operational.
In post #4, 'jeffmazur' wrote:
" You can't use a 2716 in an Apple II without and adapter socket."
" And a better way to indicate a bad RAM chip is to beep the speaker a number of times to indicate its position."
" All of the ROMX products have such a test and can identify bad RAM as long as the CPU is operational."
Uncle Bernie comments:
- Use of adapter sockets for earlier Apple II for 2716 is not necessary if it's a Taiwanes clone (they used 2716 or 2732, no mask ROMs, for the firmware they stole from Apple), or, to put the 2716 into a 2k ROM socket of an original Apple II, a few pins need to be lifted / rewired, and a /CS wire added. This is trivial to do. No special adapter socket needed. Even the 2716 can be salvaged afterwards.
- Agree that beeping the speaker is a better way to tell the location of a bad RAM. Would be more programming effort, though.
- ROMX is a 'product' and costs a lot of money. My stuff is free, as long as you have an EPROM burner and an empty EPROM around.
The latter is basically the same issue as with any other repair project needing a tool you don't have: if it's once and done, it's unwise to buy an expensive tool which afterwards falls into disuse. It's a waste of money. This is why I never sold my "Toolkit" PROMs for the Apple-1. They would be used once and bipolar PROM blanks are too expensive and rare for being used once and then tossed into a drawer, never to be used again.
I dimly remember I once saw a diagnostics card for the Apple II which would plug into a slot, and it even had a LED display to show error codes. What a nice, convenient tool ! But for anyone not in the business of repairing Apple II computers, it was a waste of money.
- Uncle Bernie
https://github.com/misterblack1/appleII_deadtest
DeadTest is a great tool, I use it myself and am very happy with it.