Jump to content
IGNORED

Yet Another Ram Upgrade question


Cheshire Noir

Recommended Posts

Hey Folks,

Sorry if you are sick to death of these already but...

So my 800xl came with a RAM upgrade. Using eXtended RAM Test, I see 192k of RAM in 3 banks.

image.thumb.png.1843eaa238d51ff3bc184cab7ebe442d.png 

Looking at the motherboard I see 8 x 41256 RAM modules. 

image.thumb.png.46e6f344d3fb68bfe963a98f8d7ea30c.png

To me this suggests 64K of RAM is "missing".

 

This machine is horribly bodged together, and I have already had to remove a ROM tower that seemed to be causing more problems than it's worth.

image.thumb.png.edb292fc166d4edc7489d524b5ded0e1.pngimage.thumb.png.759b1c1057db80d873e0da87098dfa0c.pngimage.thumb.png.683e4f0c51383b7b57cf2f4d88028c5b.png

So my questions are:

  1. Am I off the beaten track? Is my RAM upgrade only 192k?
  2. If it should be 256k, where on the board should I be looking for faults?

 

Happy to pop the hood and photograph additional bits if it'll help.

 

Long term I figure I'll source a modern RAM replacement and "undo" as much of this mod as I can, even if only for my own sanity.

 

(BTW, the ROM tower only had MC Monitors, things that are of no interest to me. Are there alternative ROMs that are of use to have in an Atari 800xl, such as Diagnostics or additional functionality?)

 

Cheers!

 

Chesh

Link to comment
Share on other sites

28 minutes ago, Cheshire Noir said:

So my 800xl came with a RAM upgrade. Using eXtended RAM Test, I see 192k of RAM in 3 banks.

You have a 256KB upgrade which replaces the original 64KB, so the increase is only 192KB or 3 banks of extended memory.

The "missing" 64KB is doing what the original RAM did before the upgrade.

  • Thanks 1
Link to comment
Share on other sites

49 minutes ago, bbking67 said:

RAMBO XL, Wizztronics,

It's neither of those.

 

I've never seen this design before. Between the '74 and '173 chips there are 6 flip flops. Why? Not bank registers, since it uses the standard PIA bits. Not enough for an 8-bit refresh counter, but there are 2 other unidentified chips.

Link to comment
Share on other sites

5 hours ago, ClausB said:

Actually 16 banks of 16K each, 4 of which mirror the base 64K

Would the extra chips potentially add the "feature" to prevent 4 of the banks from mirroring the base 64K?

 

I think running the previous version of xRAM test (0.21) which didn't check for that condition might reveal that if so...

XRAM v0.21.xex

Link to comment
Share on other sites

5 hours ago, Nezgar said:

Would the extra chips potentially add the "feature" to prevent 4 of the banks from mirroring the base 64K?

I don't know. Maybe the OP could post a better photo of that 6-chip board?

 

But why disable the mirror? It's useful for loading up the upper 16K without disabling the OS ROM. Also you gain access to that 2K RAM hidden behind the I/O area.

  • Like 2
Link to comment
Share on other sites

2 hours ago, ClausB said:

I don't know. Maybe the OP could post a better photo of that 6-chip board?

 

But why disable the mirror? It's useful for loading up the upper 16K without disabling the OS ROM. Also you gain access to that 2K RAM hidden behind the I/O area.

I fully agree regarding the benefits of the main memory mirroring...

 

Just refreshing my memory here... (there's a pun...) But reviewing the A8FAQ the "original" Peterson type upgrade description notes that main memory was protected from banking, and something about bit 5 being "inverted" compared to RAMBO type upgrades (what does that mean?), and later mods supplied after "Rambo" type upgrades gained popularity to allow banked access to main memory and ANTIC access to banked memory switched to CPU calling the "256KXLE"

 

Methinks there may be a connection here....

  • Like 2
Link to comment
Share on other sites

Whoah. OK, the board is out for a protracted period anyway at the moment. (It keeps crashing when the system gets hot. My first step is to replace the ROM socket as it's been stretched to all sh*t by the old ROM tower)

I'll trace all the pins and identify all the chips and write it up for you all. May only be of interest to a handful of people but what the hey ?

 

Chesh

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

Starting with that board, the 6 chips on it are (from top left, clockwise) a 74174 (pin1 top right - Hex D FlipFlip), a 74173 (pin 1 bottom left - Quad D FlipFlop), two 74158s (pin1 top right - Quad 2 input multiplexers, inverted), a 7427 (pin1 top right - Tri NOR gate), and a 7432 (pin1 top right - Quad OR gate).

It's a double sided board, which is adding a layer of complexity to my investigation, and it has components soldered to the underside ?

I've started tracing it but it's gonna take a couple of days. VERY tempted to desolder it and clean it up.

IMG_20210222_195058.thumb.jpg.b29a06606bd50960d6c70102cd7e7046.jpg

(Numbering guide for the rest of what I've found)

R32, R104, R105 removed

The hole for R32 east is connected to board pin 11 (Black wire)

The hole for R104 north is connected to board pin 10 (blue)

The hole for R105 north is connected to board pin 12 (orange)

U27 pin 1 connected to board pin 7 (brown)

U27 pin 8 connected to board pin 21 (white, labelled "-") <- this would make sense. Just tapping ground and +5v

U27 pin 16 connected to board pin 22 (Yellow, labelled "+")

U28 pin 2 connected to board pin 2 (Purple)

U28 pin 5 connected to the switch (black wire)

U28 pin 1 connected to U3 pin 4 (underside mod)

U28 pin 4 connected to U28 pin 12 (underside mod)

U28 pin 7 connected to U3 pin 5 (underside mod)

U2 pin 9 connected to board pin 14 (yellow)

U30 pin 13 connected to board pin 13 (grey)

U19 pin 2 connected to board pin 17 (red)

U19 pin 12 connected to board pin 20 (grey)

U7 pin 19 connected to board pin 4 (yellow)

U7 pin 20 connected to board pin 6 (white)

U7 pin 25 connected to board pin 9 (purple)

U7 pin 26 connected to board pin 8 (red)

U7 pin 30 connected to board pin 16 (green)

U7 pin 31connected to board pin 15 (white)

U8 pin 35 connected to board pin 19 (blue)

U8 pin 40 connected to board pin 18 (black)

U23 pin 12 connected to board pin 3 (green)

U23 pin 13 connected to board pin 5 (orange)

U23 pin 14 connected to board pin 24 (white)

U23 pin 15 connected to board pin 27 (red)

U23 pin 16 connected to board pin 26 (yellow)

Board pin 1 goes to the switch (black wire)

Board pin 25 goes to board pin 23 (it's a bodge)

 

(I initially thought there was a hack on u29 before I discovered it's supposed to look like that :-D )

I haven't started investigating the chips on the Atari itself.

 

I'll post more as more info comes to hand ?

 

Chesh

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

10 hours ago, Cheshire Noir said:

... the 6 chips on it are (from top left, clockwise) a 74174 (pin1 top right - Hex D FlipFlip)...

In your original photos it's clear that chip is a 74HCT74, a dual flip-flop.

 

Thanks for tracing all that! I'll try to deduce all its functions.

Edited by ClausB
  • Like 1
Link to comment
Share on other sites

Quick look:

 

There are between 1 and 4 configuration registers of up to 2 bits each at addresses $D6xx. Purpose unknown.

 

CPU HALT is wired, so it may be capable of separate ANTIC / CPU banking.

 

Sound familiar to anyone yet?

Edited by ClausB
  • Like 1
Link to comment
Share on other sites

ClausB you are quite correct on it being a PAL system, and also on that chip being a 7474 series. I pulled the board out to clean it up and it's quite visible now.

I tried desoldering the chips off the board but they're soldered from the top, not the bottom, except for ground and power rails. I couldn't budge them.

I've cleaned up the soldering, and all the horrible double sided tape.

I'll trace the traces if I get a chance, clean off the pad I accidently lifted, and then reinstall it, while cleaning up as much as I can. Hopefully it'll work when it all goes back together.

If it doesn't I may try and return it to "stock" for now, until I can afford something more modern.

 

On the bottom of the board is "© DM 1986". On the top is "1.2", if that helps anyone trying to guess the origin of the board. The board itself has two cut traces, so it's a modded mod I guess?

 

Chesh

  • Like 1
Link to comment
Share on other sites

Below is my best guess for the bank switch circuit made up of the two '158 ICs and the '27. This design is equivalent to mine and, in some ways, better. It has 3 added chips rather than my 2, but it doesn't require removal of U27, the '158 on the motherboard. That's better for unsocketed boards. Just remove 3 resistors and patch this circuit in. Also, the RA8 signal timing is better in this design.

 

Maybe the build is sloppy, but the engineering appears to be competent, at least this portion of it. I still don't know what those other 3 chips do.

 

256K.PNG

Edited by ClausB
  • Like 4
Link to comment
Share on other sites

So I have traced it out. There are undoubtedly errors. I've checked it to the best of my ability. My next step is to refit it, and clean up as I go.

Hopefully it'll all work at the end :-D

 

If anyone needs me to check anything, let me know by Friday night US time (My Saturday morning)

 

Chesh

AMEMEX.pdf

  • Like 1
Link to comment
Share on other sites

On 2/23/2021 at 6:23 PM, Cheshire Noir said:

On the bottom of the board is "© DM 1986". On the top is "1.2", if that helps anyone

"Dreadful Mess" committed in 1986? 1.2? Not just once, but twice?

 

One thing clear here is that yours and Claus forensics are already x10 more valuable than that ghetto of fried noodles you found.... 

 

Hope you guys share a beer, once the crime scene is cleaned up!

Link to comment
Share on other sites

3 hours ago, ClausB said:

Possible mistakes:

 

U6 pin 3 is a floating input.

 

U5 pin 8 is an output connected to wire 6 which is A15, an output from the motherboard.

Now I'm actually awake, rather than trying to dump a couple of hours tracing into the intarwubs in a hurry, I can go through this ?

 

U5 pin 8 you are absolutely right. I drew a wire over another wire. I have corrected that.

 

The wire from U6 pin 3 goes to U4 pin 3 but has been cut. (Cuts are marked as open jumpers)

Wire 26 and 27 are soldered directly to U6 (now noted as JPatch).

 

ClausB I hope this means something to you. I never got that far in logic. 

 

Also apologies for the very messy KiCAD drawings. I'll upload the original files when I get back from work.

 

Chesh

 

 

AMEMEX1.01.pdf

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...