Old poorly Atom

emulators, hardware and classic software for atom + system machines
Post Reply
Boydie
Posts: 529
Joined: Sat Oct 24, 2015 9:25 am
Location: Sunny Wigan
Contact:

Old poorly Atom

Post by Boydie »

Having just got an AtoMMC, I've been having a play with my old atoms.
The Issue 5 still lives, but the Issue 4 is a bit more poorly. At first it displayed nothing, but swapping the 6847 brought the display back, but unfortunately that display is garbage. I believe this means the machine is not completing its reset routine, as CTRL-G does nothing, and Break appears not to either.
So far, I've removed all ram chips except for IC51 and 52, including video ram except for IC42 and 43. Changing ICs 51 and 52 has had no effect, and all changing ICs 42 and 43 does is change the overall pattern of garbage that appears on the display.
I've removed ICs 1-5, 21, 24, and 50 (I believe these are all optional) with no effect.
I've swapped the 6502, 8255 and IC9, also with no effect.
The board has had a few damaged tracks repaired at some time, but these all test good, with no continuity to neighbouring tracks.

Tomorrow looks like it's going to be a day of testing as per other post, and I'll post again with pics of ther board. scope readings etc. However, I've noticed that there is a faint, high-pitched noise coming from the speaker, probably RF interference (like the BBC's 1MHz noise but quiter and higher pitched). Is this normal, or does it point suspicion in a potentially useful direction?
User avatar
1024MAK
Posts: 11067
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Old poorly Atom

Post by 1024MAK »

Not sure about the noise, the Atom does not have a long PCB track leading to an audio amplifier like the Beeb.

If you have a ‘scope, testing the control pins on the 6502 would be useful as the next step. Then if the PH0 clock in (pin 37), PH2 (pin 39), reset (pin 40) are okay, test R/W (pin 34) and sync (pin 7). Also have a gander at the data bus and address bus. With these, at first, it is simply a case of seeing if there is any activity, is the pin stuck low, stuck high, if changing, is it going between valid logic levels?

Oh, have you removed the ROM and checked the socket contacts and the ROM pins?

Mark
Last edited by 1024MAK on Sat Sep 28, 2019 6:59 pm, edited 1 time in total.
Boydie
Posts: 529
Joined: Sat Oct 24, 2015 9:25 am
Location: Sunny Wigan
Contact:

Re: Old poorly Atom

Post by Boydie »

Not been able to test yet. However, there has been (hopefully) significant progress. I've found that if the board is lifted slightly in one corner then I'm sometimes rewarded with a prompt. Tapping around the 8255 also sometimes causes the garbage on the screen to change (admittedly to something that's even worse, but it still changes).
So the next plan of action is to replace all the lousy-quality sockets and see what that does, along with a careful inspection for bad joints/tracks.

At least it means it's fixable...
Boydie
Posts: 529
Joined: Sat Oct 24, 2015 9:25 am
Location: Sunny Wigan
Contact:

Re: Old poorly Atom

Post by Boydie »

It lives!
It seems to have been a combination of:

1) Bad 6847 (no display)
2) Dodgy 8255 socket
3) Very dodgy keyboard contacts - over half the keys not working. When this includes the Break key, it takes a while before even a working machine can be persuaded to reset.

All keys have now been cleaned and the whole thing works.

The Break key issue also didn’t help when I’d just been using an Issue 5, which started immediately to the Atom prompt, without needing to press Break.
Is this difference recognised between Issues 4 and 5, in that Issue 4 starts to garbage and requires Break to reset it, but the Issue 5 starts completely, or is one of mine behaving atypically?
User avatar
roland
Posts: 4481
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Old poorly Atom

Post by roland »

I think there are only a few (issue 5) Atoms that start directly after powering on. Most Atoms that I ever had needed a break to start. So this is quite normal.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:
User avatar
JudgeBeeb
Posts: 1118
Joined: Thu Sep 10, 2015 9:56 pm
Contact:

Re: Old poorly Atom

Post by JudgeBeeb »

Boydie wrote:
Mon Sep 30, 2019 1:28 pm
When this includes the Break key, it takes a while before even a working machine can be persuaded to reset...

...The Break key issue also didn’t help when I’d just been using an Issue 5, which started immediately to the Atom prompt, without needing to press Break.
Is this difference recognised between Issues 4 and 5, in that Issue 4 starts to garbage and requires Break to reset it, but the Issue 5 starts completely, or is one of mine behaving atypically?
Whilst I have been playing around with my Atom, I removed C11 and R38 and replaced them with a DS1813 reset controller. This resulted in a very clean reset, both from power on and from pressing the BREAK key. if you bend the leads carefully, it fits very nicely in the holes left by C11 and R38.

Ultimately, I ended up removing the DS1813 because I want to experience this machine, as it would have been, back in the day.
There is so much wonder in the universe; why should you want to imagine that there is more?
User avatar
1024MAK
Posts: 11067
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Old poorly Atom

Post by 1024MAK »

Although the values used for R38 (1kΩ) and C11 (10uF or 22uF) appear to meet the minimum time needed for a reset pulse (to meet the 6502s requirements), it is possible that they don’t. Power on is the worst case scenario, as it takes time for the main oscillator to get going (the 6502 CPU needs a stable clock in order to reset correctly).

And these values will not prevent multiple reset pulses if there is too much switch bounce from the BREAK key. Hence there may be an advantage in fitting a higher value component in either R38 (I suggest 2.2kΩ if C11 is 22uF, or 4.72kΩ if C11 is 10uF) or increasing the value of C11 (say up to 47uF) or both. Doing this will increase the length of the reset pulse.

Pressing the BREAK key, especially if the human holds it down for a period of time, increases the length of the resulting reset pulse, and of course, pressing BREAK does not affect the main oscillator which will just continue running.

The other problem is that the gate that this signal feeds, a 74LS04 (IC9) has a normal logic input and some type/makes will produce an output that jitters rapidly between logic low and logic high if it experiences a slowly (to it) changing DC level at it’s input pin while in the undefined area between valid logic low and logic high levels This can be made worse if it is in an electrically noisy environment like a computer.

A specialist reset IC can be used (like Robert describes). See also this post on the subject from 2012.

Or you could just do like most Atom owners do, and rest you finger on the BREAK key... :P

Mark
User avatar
oss003
Posts: 3556
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Old poorly Atom

Post by oss003 »

Most people did enlarge C11 to 47 uF to get a decent reset after powerup.

Greetings
Kees
Post Reply

Return to “acorn atom and acorn system series”