• Please review our updated Terms and Rules here

Weird characters in kaypro II boot

J

jmellidg

Experienced Member
Joined
Nov 6, 2015
Messages
93
Location
La Coruña, Spain
Hi, I have some when I boot my kaypro II. As you can see in the picture, there are odd characters in the boot screen. I can't boot the kaypro by the way. I've double checked the character rom 81-146 and the monitor rom 81-149c with the ones from Internet sources and there are identical.

Do you know what's going on??

Thanks in advance!!
 

Attachments

  • IMG_20231111_124138.jpg
    IMG_20231111_124138.jpg
    1.6 MB · Views: 45
The display says

* KAYPRO II *
Phaasa phaca ykqr 'iskappa ijpk Drira A

I'm guessing that ought to be "Please place your diskette into Drive A".
In which case, the cipher includes these substitutions:

d -> `
e -> a
l -> h
n -> j
o -> k
t -> p
u -> q
v -> r

Next step is to examine the bit differences between those letters...
 
Code: 
d -> ` is 0110.0100 -> 0110.0000
e -> a is 0110.0101 -> 0110.0001
l -> h is 0110.1100 -> 0110.1000
n -> j is 0110.1110 -> 0110.1010
o -> k is 0110.1111 -> 0110.1011
t -> p is 0111.0100 -> 0111.0000
u -> q is 0111.0101 -> 0111.0001
v -> r is 0111.0110 -> 0111.0010

It certainly looks like bit 3 is stuck low!

Capital letters are AIDKOPRY. This is a little bit mysterious since the byte for D is 0100.0100 and for O is 0100.1111; bit 3 is OK there! (Other capital letters have bit 3 low.)
Unaffected lower-case letters are acikrsy, and there is also *, but all of those characters have bit 3 low.
 
Last edited:
A reasonable chance it is a DRAM chip that provides Bit 2 as was mentioned, but until you check with some diagnostics tools, you can't be certain. Well, I think Bit 3 was said, but I think they meant the third bit, which is probably listed as Bit 2 in your pinouts.
 
When a RAM chip goes bad, it is not too uncommon for just SOME of the memory cells to stop working right. Print the same text to somewhere else on the screen and it might turn to gibberish. Or the bits for those cells could just happen to randomly be stuck in a way that makes the text happen to look right.
 
It seems that all the times show the same text. The video ram is not socketed so I want to be sure about it before change it..
 
My first thought is bad video RAM, but more might be revealed by using the monitor to try some patterns. As SomeGuy mentioned, being able to plant various data in different locations might reveal a pattern to the errors. There is "fill" command that allows you to put the same character in all locations of the video RAM, and also a "substitute" command that would allow you to enter characters (by hex value) in specific locations.
 
jmellidg said:
Hi durgadas thank you very much, I'll try. Do you think that it could be a memory issue or in a bus multiplexer??
Click to expand...

It most certainly "could" be. there's not enough information provided by the screen alone, though it has given some clues.

You can begin by examining the schematics. https://oldcrap.org/wp-content/uploads/2023/04/kaypro-ii-schematics-v2.pdf

On the third page, it suggests the video ram is 4 x 2114 memory chips. This would suggest the top of the screen and the bottom of the screen are on different chips as each is 1k x 4 bits in size and would explain why the top half of the screen is fine, while the bottom is not working.

This suggests U31 Pin 13 is your fault... You will need to progress to testing further here to find out what is wrong. Old static memories have all kind of things that can go wrong, but I'm going to guess that an oscilloscope will show that error, though you'll need to sync it with other signals to be sure, since it should be written from the shared bus from the other 2114s which means it's not likely to be any other chips there, or a short between address and data lines.

The bad news is that 2114s have gone the way of the dodo... You might need to make up a jig to place a 4116 into it's place or similar to repair it, which is going to be very messy.
 
Just noticed those chips are now on Ali Express...

They look old, and the legs are all corroded and clearly retinned, which is probably a good thing so long as they didn't cook them. Some other sellers have ones that look like they are relabeled too and are who-knows-what.

Anyway, they weren't there a few months back when I looked so someone clearly scrapped some old computers... I wonder where they are getting their equipment from?
 
I've seen Here that it can be a safe way to test the video ram and all the circuits related. Just put an empty monitor eprom (2716) and look at the screen. It would be nice, I'm just waiting for the chips
 
I've already tested, and I'm afraid that the video stuff works correctly, do you think that I could discard the video memory??
 

Attachments

  • IMG_20231112_124549.jpg
    IMG_20231112_124549.jpg
    1.6 MB · Views: 23
According to previous analysis, the characters produced by the "empty EPROM" test would not fall into the category of ones that get "corrupted". So, while this test confirms that basic video functions are working, it does not prove video RAM (or whatever is corrupting those characters) is working.

If you are able to use the debug monitor to check out the video RAM, keep in mind that the video RAM addresses have a "scrambler" on them, in order to pack the data into 80-character lines. I can produce a table of RAM addresses that correspond to lines, if that is needed.
 
FYI, I was able to get 28C16 EEPROMs, which might work better in modern programmers (doesn't require the odd voltages). Of course, legacy programmers can't handle EEPROMs. But the 28C16 should work in the Kaypro II just fine.
 
You must log in or register to reply here.
Back
Top