Yet Another Acorn Atom Repair

emulators, hardware and classic software for atom + system machines
User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Yet Another Acorn Atom Repair

Post by CommanderCoder » Sat Aug 29, 2020 7:42 pm

I recently got an old Atom from trixster and the necessary chips from roland : more info here. viewtopic.php?f=44&t=20152

I've been away from home for a week or so, but this afternoon I managed to get them into the Atom. I couldn't find a 5V PSU (I'll buy one later) so I hooked up at 12V supply through DC-DC Buck Converter. I got the necessary 5V going through it but the 6502 seemed to get warm quickly - not burning hot, but noticeably warm.
IMG_8055.jpg
IMG_8054.jpg



I plugged a TV into the UHF modulator and couldn't get a picture. I had to leave it there, other things to do, but just now while tracing around the board I discovered why I didn't get a picture! I'll try again with a composite video TV tomorrow.
IMG_8050.jpg



If it is working - hurray - but I can't believe it will be. I've been reading some of the other repair stories on this forum, and no doubt they will be useful. I've got an oscilloscope so I'll work through the board bit by bit. I'll post my progress as I go.

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Sat Aug 29, 2020 7:54 pm

I am looking forward to your story :D
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
oss003
Posts: 3364
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yet Another Acorn Atom Repair

Post by oss003 » Sun Aug 30, 2020 7:09 pm

Great to see another working Atom . ..
Is it working ok does it need some work?
A normal Atom can take up to 3A with expansions .... can your powersupply do that?

The easiest way is to use a 5V USB power supply.

Greetings
Kees
Attachments
Screenshot_20200830-201538_Chrome.jpg

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Mon Aug 31, 2020 1:23 pm

It didn't work when I plug the composite video in.

I've started working from the CPU outward and immediately I discovered something missing.. Can you (NOT) see it?
IMG_8056.jpg
What's a good source of 4Mhz crystals for the Acorn Atom?

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Mon Aug 31, 2020 1:28 pm

oss003 wrote:
Sun Aug 30, 2020 7:09 pm
Great to see another working Atom . ..
Is it working ok does it need some work?
A normal Atom can take up to 3A with expansions .... can your powersupply do that?

The easiest way is to use a 5V USB power supply.

Greetings
Kees


I'll get a good 5V PSU - but in the meantime the Buck Supply Module should be able to supply up to 3A. https://hobbycomponents.com/power/215-l ... ply-module

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Mon Aug 31, 2020 1:41 pm

CommanderCoder wrote:
Mon Aug 31, 2020 1:23 pm
I've started working from the CPU outward and immediately I discovered something missing.. Can you (NOT) see it?
X2 has left the building ;)
CommanderCoder wrote:
Mon Aug 31, 2020 1:23 pm
What's a good source of 4Mhz crystals for the Acorn Atom?
RS components, Farnell, Mouser.....
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
trixster
Posts: 1062
Joined: Wed May 06, 2015 12:45 pm
Location: York
Contact:

Re: Yet Another Acorn Atom Repair

Post by trixster » Mon Aug 31, 2020 2:55 pm

All fingers are crossed that you can get the Atom up and running, CommanderCoder! [-o<

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Mon Aug 31, 2020 2:57 pm

He will :lol:
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
1024MAK
Posts: 10292
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Yet Another Acorn Atom Repair

Post by 1024MAK » Mon Aug 31, 2020 3:08 pm

CommanderCoder wrote:
Mon Aug 31, 2020 1:28 pm
I'll get a good 5V PSU - but in the meantime the Buck Supply Module should be able to supply up to 3A. https://hobbycomponents.com/power/215-l ... ply-module
Image

These cheap DC-DC converter modules are okay, but having been made down to a price in China, are not exactly top quality items. So I suggest you divide any current or power ratings by a factor of at least 2 if used continuously. It would also be worthwhile adding an extra ‘low ESR’ electrolytic capacitor to both the input and the output near as possible to the DC-DC converter module as possible.

4MHz crystals should be easy to find.

Good luck :wink:

Mark

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Mon Aug 31, 2020 6:07 pm

4Mhz Oscillator ordered from Ebay - £1.80 inc postage.

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Thu Sep 03, 2020 7:49 pm

It appears I ordered 10x 4Mhz oscillators. So I've got 9 spare now.

After fitting the crystal, I have a solid 1Mhz clock into the CPU. However, still nothing. The CPU is getting very hot within about 5 minutes so I think something is wrong with the chip. I've probed some of the address and data lines and they appear to be fixed on #0002. Pressing BREAK on the keyboard brings the RST low but nothing changes on the CPU address or data lines.

I have a 6502 in my BBC but I'm not keen to try moving them around.

In the meantime I'll probe some more. Any thoughts on things to check first? I tried the keyboard into the PPIA and it is bringing lines high which looks promising. The VDG is difficult to test without a CPU I think. The video out isn't generating any signal that my oscilloscope or TV can pick up.

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Thu Sep 03, 2020 10:00 pm

I don't expect that your beeb will break if you put the Atom 6502 in it. Did I send you a 6502A? If so, you can place it in your beeb to test it.

Can you also post some clear high res picture of your Atom in the current state?

The best way to troubleshoot this is starting to check the power at each IC. When the cpu get quite hot, you can leave it out until we know that the power is fine.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Thu Sep 03, 2020 10:06 pm

If you want to test the 6502 in the Atom you could build the nop generator: http://6502.org/mini-projects/nop-gen/nop-gen.htm

Remove all memory from the Atom and place this in an eprom socket. The cpu will see the opcode #EA and do nothing. The address bus will behave like 16 square wave generators. This will test if the cpu starts and runs.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Thu Sep 03, 2020 10:48 pm

Cheers. I'll give it a go and see what happens. I need to remove some chips from the Atom to get a really basic setup and double check the connections.

User avatar
oss003
Posts: 3364
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yet Another Acorn Atom Repair

Post by oss003 » Fri Sep 04, 2020 7:14 am

Maybe this will help: viewtopic.php?f=44&t=6318&p=60505&hilit ... ion#p60505

PS I also noticed that C11 is missing ..... this is for delaying the RST puls at power up to start normal instead of first have to press BREAK to remove a screen full of rubbish. If adding the capacitor doesn't help, you can always install a bigger one (20u or bigger).

Greetings
Kees
Attachments
c11.png

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Fri Sep 04, 2020 8:14 am

Good spotting Kees! It also looks like resistors R40 and R39 hit each other on the blank leads. That will also cause troubles. I had that once in an Atom.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Sat Sep 05, 2020 11:34 am

Wow. Great eyes fellas.

I’ve tested the 6502 chip in my BBC and it works fine. Not getting hot.

I’ll sit with the atom, a cup of tea, the circuit diagram and a continuity tester and figure out what else is missing..

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Sat Sep 05, 2020 12:01 pm

That's nice to know. IHMO it's time to add the missing capacitor, build a minimal Atom configuration and check the voltages. And don't forget to check the resistors R39, 40 and 41.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
jgharston
Posts: 4121
Joined: Thu Sep 24, 2009 12:22 pm
Location: Whitby/Sheffield
Contact:

Re: Yet Another Acorn Atom Repair

Post by jgharston » Sat Sep 05, 2020 2:48 pm

I've also got an Atom with no PSU, the cruicial information missing above is: what's the polarisation of the PSU plug? Is the tip positive, or is the tip ground? I don't want to risk popping my Atom by plugging a PSU the wrong way around. ;)

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.32
(C) Copyright J.G.Harston 1989,2005-2020
>_

User avatar
IanS
Posts: 1411
Joined: Mon Aug 31, 2009 7:02 pm
Contact:

Re: Yet Another Acorn Atom Repair

Post by IanS » Sat Sep 05, 2020 3:17 pm

jgharston wrote:
Sat Sep 05, 2020 2:48 pm
I've also got an Atom with no PSU, the cruicial information missing above is: what's the polarisation of the PSU plug? Is the tip positive, or is the tip ground? I don't want to risk popping my Atom by plugging a PSU the wrong way around. ;)
Centre-positive. Check if the regulators have been bypassed. Start at 5V, only try a higher voltage if sure the regulators are still in use.

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Sun Sep 06, 2020 8:44 pm

I stripped down my Atom to something minimal (just 4 ram chips i.e. 2K), replaced the capacitors and checked those resistors (they're not touching). I put my oscilloscope probes on the address lines of the CPU and after pressing BREAK a few times it started to show some signals that look like data moving around. Kind of overlapping square waves with different spacing. It needs a few presses of BREAK to get this going so something else might still be wrong but it's a start.

The video output isn't showing anything and my TV doesn't even try to catch a signal, so the next step is checking the video output. Is there any way to get the video chip to just send out a 'test card' signal?

Cheers,

User avatar
oss003
Posts: 3364
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yet Another Acorn Atom Repair

Post by oss003 » Sun Sep 06, 2020 9:21 pm

Are you using the composite video-in of your tv because the RF modulator is removed?
The composite video signal and ground should be connected to pin 9 + 10 of PL4 or use these signals directly from PL4.

There is no testsignal available.
IC's 42 and 43 should be in for minimal RAM configuration.

Greetings
Kees
Last edited by oss003 on Mon Sep 07, 2020 9:08 am, edited 2 times in total.

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Sun Sep 06, 2020 9:23 pm

There should be a video signal regardless whether the processor works or not. The 6847 will read from its memory and display that. If the memory is not cleared by the reset routine then it holds the well known garbage and you should see that.

With a scope you can start at pin 28 of the 6847 and follow the signal through C4, Q1 and Q2. The video circuit is not very complicated :lol:

And it would help us if you post a hi-res picture of your Atom.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
hoglet
Posts: 9445
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yet Another Acorn Atom Repair

Post by hoglet » Sun Sep 06, 2020 9:33 pm

This photo was posted in the other thread:
IMG_7634.jpg
There seems to have been damage to the video amplifier that someone has tried to repair.

I would start checking the connectivity here against the schematic.

Specifically, on transistor Q2 it looks like all the pads on the underside are missing (just below the blue wire). The connection here to R6 looks very dodgy. That would defintely cause you to have no video output.

Dave

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Mon Sep 07, 2020 11:38 am

Okay - more news.. I was checking PIN 1, PIN 9 and PIN 17 of my 6847 to see if power was good and, no it wasn't! (thanks viewtopic.php?f=44&t=7308&start=60#p79104)

I should have checked there first. After more testing it seems that the whole left side (pins 1-20) are not connecting through the IC Socket. The pins of the socket are connected to the board okay but the legs of the chip aren't making contact inside the socket. I think they've all become bent or oxidised.

Other than replacing the socket, is there a quicker fix for this? I'm thinking of shoving 40 little solid wires down the side of each leg to force connections!

User avatar
hoglet
Posts: 9445
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yet Another Acorn Atom Repair

Post by hoglet » Mon Sep 07, 2020 11:41 am

CommanderCoder wrote:
Mon Sep 07, 2020 11:38 am
Other than replacing the socket, is there a quicker fix for this? I'm thinking of shoving 40 little solid wires down the side of each leg to force connections!
Sometimes putting a 40-pin turned-pin socket in between can help, because the pins are a bit fatter than the IC pins.

Dave

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Mon Sep 07, 2020 12:05 pm

great idea!

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Mon Sep 07, 2020 12:49 pm

Perhaps there once was a colour board in that Atom; that might have worn out the 6847 socket.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

User avatar
CommanderCoder
Posts: 87
Joined: Wed Nov 06, 2019 5:50 pm
Location: Royal Leamington Spa
Contact:

Re: Yet Another Acorn Atom Repair

Post by CommanderCoder » Thu Sep 10, 2020 7:59 pm

I think you're right. The turned pin socket slotted straight in and I put the 6847 into it (broke/bent a couple of pins but easy to fix onto the new socket).

This is the current output from the Atom. Much better than the blank screen I had been getting.
IMG_8077.jpg
Pressing BREAK caused the bottom half of the picture to go fuzzy, releasing it gave the same image.

So I guess this might be because the CPU isn't resetting memory or memory is bad or just the usual dodgy wires.

Oh.. and the picture got clearer when I fiddled with the Q1/Q2 transistors. I haven't check on the soldering on those yet. Q1 and Q2 are both BC239 - so they should be fine.

User avatar
roland
Posts: 4050
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yet Another Acorn Atom Repair

Post by roland » Thu Sep 10, 2020 10:41 pm

Can you check if there's a 60Hz square wave on pin 7 of the 8255? This is the FS signal from the 6847. The Atom waits for this signal when it wants to write to the screen. And since it starts with clearing the screen it needs this signal.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

Post Reply

Return to “acorn atom and acorn system series”