Martin's Acorn Atom (poorly)

discussion of games, software, hardware & emulators relating to the Acorn Atom
User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sat Aug 05, 2017 3:37 pm

The plot thickens ...

Still getting Acorn Atom and the CS is working if I do a 1 shot on it, falling edge 1.5v ish trigger and hit break. It's just got another signal superimposed on it when it's high.

IMG_5223.JPG

IMG_5221.JPG


It's looking like the CPU isn't driving it after that for some reason.

Martin

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sat Aug 05, 2017 4:00 pm

martinw wrote:The plot thickens ...

Still getting Acorn Atom and the CS is working if I do a 1 shot on it, falling edge 1.5v ish trigger and hit break. It's just got another signal superimposed on it when it's high.

OK, it looks like that is working. Don't worry about the other signal that appears to be superimposed, it's likely just crosstalk from an adjacent trace.
martinw wrote:The plot thickens ...
It's looking like the CPU isn't driving it after that for some reason.

So it looks like the CPU is possibly crashing, or hanging prior to calling the keyboard read code.

Maybe the next thing to do is to try capture the address bus (as outlined earlier) and upload this.

That should let us see where the 6502 gets to.

Dave

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sat Aug 05, 2017 4:11 pm

Martin,

One more thing.
oss003 wrote:Can you check pin 1,2,3,4 of the 8255 if there is any activity?

This is worth doing.

If there is activity, then the only thing I can think of which will cause exactly these symptoms is a key being stuck down.

In fact, I'm 99% sure this is what is happening.

This is where I guess it will be stuck:

Code: Select all

FE9F  20 71 FE  JSR #FE71       Scan keyboard
FEA2  90 F6     BCC #FE9A       ..wait for key to be released


Try holding down the REPT key, and that will give you a clue what key it is.

(Or I could be completely wrong :D )

Dave

User avatar
1024MAK
Posts: 6678
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Martin's Acorn Atom (poorly)

Postby 1024MAK » Sat Aug 05, 2017 4:55 pm

martinw wrote:Still getting Acorn Atom and the CS is working if I do a 1 shot on it, falling edge 1.5v ish trigger and hit break. It's just got another signal superimposed on it when it's high.

Welcome to the electrically noisy world of computers :P
It's not uncommon for a signal line to have bits of other signals present. It's caused by one, or a combination of capacitive coupling, inductive coupling, or variations in the supply rail voltages at the chip the signal comes from (caused by voltage drop along the power supply and ground PCB tracks).

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
oss003
Posts: 2516
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Martin's Acorn Atom (poorly)

Postby oss003 » Sat Aug 05, 2017 9:38 pm

hoglet wrote:If there is activity, then the only thing I can think of which will cause exactly these symptoms is a key being stuck down.

You can check if a key being stuck down by checking the key-matrix signals to the 8255 (pb0-7 + pc6)
All these signals should be high, if a low puls is detected, a key in the corresponding column is being stuck down.

Greetings
Kees

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sun Aug 06, 2017 6:37 am

oss003 wrote:
hoglet wrote:If there is activity, then the only thing I can think of which will cause exactly these symptoms is a key being stuck down.

You can check if a key being stuck down by checking the key-matrix signals to the 8255 (pb0-7 + pc6)
All these signals should be high, if a low puls is detected, a key in the corresponding column is being stuck down.

Or, as I discovered yesterday, just press REPT and the stuck key will show up on the screen.

I was thinking it would be nice to write an Atom troubleshooting guide. It's a simple machine, but there are a few subtleties.

Dave

User avatar
oss003
Posts: 2516
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Martin's Acorn Atom (poorly)

Postby oss003 » Sun Aug 06, 2017 8:10 am

hoglet wrote:I was thinking it would be nice to write an Atom troubleshooting guide. It's a simple machine, but there are a few subtleties.

Sounds like a nice idea, good to have a structured plan to solve Atomic problems.

I love it when a plan comes together ..... :lol:

Greetings
Kees

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 9:00 am

Hi,

Here are the WFM files from the scope.

I've done a few, sometimes it seems to get further and throw up some funny images on the screen.

Martins Atom WFM.zip
(9.66 KiB) Downloaded 10 times


Martin

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sun Aug 06, 2017 10:31 am

martinw wrote:Here are the WFM files from the scope.

I've done a few, sometimes it seems to get further and throw up some funny images on the screen.

I've just had a look at these, using this tool

Code: Select all

Channel CHLA
============
Enabled                  : 1
Time grid scale          : 2.000e-01 s/div
Samplerate               : 3.413e+03 Samples/s
Time delay               : 6.300e-01 s
No. of recorded samples  : 8192
Active channel           : 0
Enabled channels         : [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
]
Size group 1 (D0-D7)     : small
Size group 2 (D0-D7)     : small

The sample rate is far too coarse to be useful (1 sample every 293 us). And there are only 8192 samples.

I'm not familiar with Rigol scopes, so I can't really advise as to how to get a higher resolution / deeper capture. Maybe someone else can. I would expect the resultant file to be several mega bytes in size.

Also, CSV would be a better format to export to than the proprietary WFM.

Anyway, this may all be moot it if turns out just to be a stuck key.

Dave

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 2:20 pm

Ok so switching (no pun intended :)) attention to the 8255.

IMG_5234.JPG


D0 to D7 = pb0-pb7 D8 = pc6 rest not connected.

IMG_5235.JPG


As you can hopefully see they're all high.

If I press break a few times nothing moves.

If I press a key one of the lines does.

I'll look into the resolution of the captured startup file next week.

Thanks for your help everybody, I'm sure we'll get there in the end with this relatively simple but sometimes PAINFUL computer :)

Martin

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sun Aug 06, 2017 2:27 pm

So did nothing happen when you tried pressing REPT?

Dave

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 2:36 pm

Nadda.

Martin

User avatar
oss003
Posts: 2516
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Martin's Acorn Atom (poorly)

Postby oss003 » Sun Aug 06, 2017 2:53 pm

As you can hopefully see they're all high.
This means that no key is being stuck.
If I press break a few times nothing moves.
Looks ok, the BREAK key is outside the key matrix.
If I press a key one of the lines does.
Means that the 7445 is working which is controlled by the 8255 so the 8255 gets selected.

Greetings
Kees

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sun Aug 06, 2017 3:37 pm

martinw wrote:If I press a key one of the lines does.

Does what exactly?

It's not clear to me from what's been reported whether the atom, after reset, is:
- continuously scanning the keyboard
- not scanning the keyboard at all
- scanning the keyboard just once then stopping
- continuously scanning the keyboard until a key is pressed and then stopping
- something else

If you stick a scope probe on one of the keyboard row signals (e.g. IC26 (the 7445) pin 1), what do you see after reset is released?

This should be a very straightforward issue to debug :?

Dave

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 6:29 pm

I swapped out the two 74LS04 s (IC 45 was the one that made the most difference) in the top left of the PCB near the big crystal and the din connector and "hey presto" we have life =D>

IMG_5236.JPG

IMG_5237.JPG


Still not perfect but getting there :)

Martin
Last edited by martinw on Sun Aug 06, 2017 6:56 pm, edited 1 time in total.

User avatar
trixster
Posts: 512
Joined: Wed May 06, 2015 11:45 am
Location: York

Re: Martin's Acorn Atom (poorly)

Postby trixster » Sun Aug 06, 2017 6:47 pm

\:D/ :-D
A3020 | A3000 | BBC B + 128K RAM/ROM + 20K Shadow + Pi0 + VideoNuLA
BBC Master Turbo + DC | Atom | A1200 060 | A500 | Jaguar | A420/1
A4000/040 060 | Atari Falcon 060 | Saturn | PS1 | SNES | CPC6128 | C64 | 3DO | MD

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 6:59 pm

trixster wrote: \:D/ :-D

:D

User avatar
hoglet
Posts: 6369
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: Martin's Acorn Atom (poorly)

Postby hoglet » Sun Aug 06, 2017 7:02 pm

Well done for finding that.

Did you actually spot the fault on the scope?

The only explanation I can offer is that possibly the oscillator ran for enough cycles to boot the Atom, then just stopped.

What problems are remaining?

Dave

User avatar
1024MAK
Posts: 6678
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Martin's Acorn Atom (poorly)

Postby 1024MAK » Sun Aug 06, 2017 7:12 pm

Odd. Oscillators and astable multivibrators normally don't like starting, but once running for ten or more cycles, (as long as nothing changes) tend to keep going...

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
oss003
Posts: 2516
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Martin's Acorn Atom (poorly)

Postby oss003 » Sun Aug 06, 2017 7:17 pm

Nice catch Martin, maybe bad pin connections?

Greetings
Kees

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 7:27 pm

Cheers Dave, no it was divine inspiration :)

With so few ICs on there in the minimal configuration and all the ICs I bought off BeebMaster when he had an online shop next to it, I thought I'd start swapping them out. The 04s looked a bit iffy ...

IMG_5242.JPG


... so I started with them and it made a major difference.

Before it was just "unstable" graphics flying all over the screen whenever I pressed break but occasionally locking onto the right Acorn Atom and cursor etc, now it's nearly rock solid although it has been resetting itself sporadically so I might swap the other 04 out, although I've just ordered 5 more on EBay!

Thanks for all the help, just having a serious look at it, with everybody behind me made the difference I think :D

Martin
Last edited by martinw on Sun Aug 06, 2017 7:35 pm, edited 1 time in total.

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 7:28 pm

oss003 wrote:Nice catch Martin, maybe bad pin connections?

Greetings
Kees


Thanks Kees.
Last edited by martinw on Sun Aug 06, 2017 7:49 pm, edited 1 time in total.

User avatar
daveejhitchins
Posts: 3624
Joined: Wed Jun 13, 2012 5:23 pm
Location: Newton Aycliffe, County Durham

Re: Martin's Acorn Atom (poorly)

Postby daveejhitchins » Sun Aug 06, 2017 7:31 pm

Looks like just dirty pins. I've been considering trying Silver Dip as a cleaning agent - not sure how it'll work on none-silver, though.

Dave H :D

P.s. you may want to replace the sockets! If the IC pins are that bad, what must the socket contacts be like??
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

User avatar
oss003
Posts: 2516
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: Martin's Acorn Atom (poorly)

Postby oss003 » Sun Aug 06, 2017 7:32 pm

Well ...... this one was new for me.
Learning all the time ... 8)

Greetings
Kees

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Sun Aug 06, 2017 7:44 pm

daveejhitchins wrote:Looks like just dirty pins. I've been considering trying Silver Dip as a cleaning agent - not sure how it'll work on none-silver, though.

Dave H :D

P.s. you may want to replace the sockets! If the IC pins are that bad, what must the socket contacts be like??


They don't look too bad, although you never know with those stamped pins, turned pins especially the gold plated ones would be a lot better, maybe a job for another day ;)

User avatar
1024MAK
Posts: 6678
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Martin's Acorn Atom (poorly)

Postby 1024MAK » Sun Aug 06, 2017 9:38 pm

Well done Martin, good that you have got it working =D>

martinw wrote:I thought I'd start swapping them out. The 04s looked a bit iffy ...

Image

... so I started with them and it made a major difference.

Before it was just "unstable" graphics flying all over the screen whenever I pressed break but occasionally locking onto the right Acorn Atom and cursor etc, now it's nearly rock solid although it has been resetting itself sporadically so I might swap the other 04 out, although I've just ordered 5 more on EBay!

Thanks for all the help, just having a serious look at it, with everybody behind me made the difference I think :D

Martin

Tarnished chip pins are not uncommon. Although if undisturbed, if the pins were originally making good contact, they don't normal cause problems.
Anyway, a simple fix is to clean the pins with a glass fibre cleaning pencil. Like this one
Image
link to CPC site

They are also available from other electronic suppliers and also from eBay and similar sites...

Brings the pins up lovely :P

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Mon Aug 07, 2017 6:26 am

Thanks for that Mark, although that little bugger is going in the bin, the amount of hassle it's caused me :twisted:

Martin

User avatar
1024MAK
Posts: 6678
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...

Re: Martin's Acorn Atom (poorly)

Postby 1024MAK » Mon Aug 07, 2017 6:48 am

A different outlook to mine then :lol:

Once I have a suspect part (that can easily be reconnected and disconnected, like a plug in device) I like to swap it back in to prove that it was the cause of the fault. In the case of a chip with tarnished pins, I clean it first. So far cleaning the pins has allowed any 74 series logic chips with tarnished pins to continue to live, as they still work fine.

That is not to say that you don't get 74 series chips that go faulty. You do. It's just that most of the time they are very reliable and it is the higher integration chips like bipolar, or NMOS SRAM, DRAM, some types of ULA and CPUs that fail the most.

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
BigEd
Posts: 1393
Joined: Sun Jan 24, 2010 10:24 am
Location: West
Contact:

Re: Martin's Acorn Atom (poorly)

Postby BigEd » Mon Aug 07, 2017 6:59 am

I'm reminded of a recent post on Dave Curran's blog, where he finds a couple of logic TTL are failing to meet timing - but still pass a logic test. Some old chips are evidently past their useful life (whereas most are still fine.)

User avatar
martinw
Posts: 1190
Joined: Sat Nov 13, 2010 10:31 am
Location: Aberdeenshire, Scotland

Re: Martin's Acorn Atom (poorly)

Postby martinw » Mon Aug 07, 2017 8:20 am

BigEd wrote:I'm reminded of a recent post on Dave Curran's blog, where he finds a couple of logic TTL are failing to meet timing - but still pass a logic test. Some old chips are evidently past their useful life (whereas most are still fine.)


That's really interesting, thank you BigEd =D>

It's a TI component but the date code is 1979.

Martin


Return to “acorn atom”

Who is online

Users browsing this forum: No registered users and 2 guests