ATOM EEPROM

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
Post Reply
dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

ATOM EEPROM

Post by dominicbeesley » Thu Jul 11, 2019 3:14 pm

Hello,

I'm trying to get an Atom I bought from Roland back to working order after it died in transit. The symptoms are that it gets as far as the boot banner and the ">" prompt appears but the keyboard is unresponsive. It looks to be in a tight loop scanning the keyboard as the 8255's nCS line is being strobed every 20us or so (not checked scientifically just roughly on an old scope). However, the PA0-3 lines aren't changing - they stay at around 1.5V. Next I tried a new 82C55 and a NOS 8255A bought from ebay both these give the same symptoms but with random noise lines on the screen.

The machine is fitted with a new video RAM board which appears to be working though I'm not sure it is fitted correctly and a YARRB which I've "undone".

I've got hold of some 2114 chips to swap back to a 6502A and original main ram but now I'm stuck as I don't have a suitable rom. I thought I could use a 2764 but I find the pinouts are different so...some questions:

- any ideas what is likely to be the problem?
- are there modern/available EEPROMS that will fit the IC20 socket
- has anyone got a spare main BASIC/Kernel rom I could borrow/buy
- have I got the video ram board plugged in wrong - the flying lead "looks" to be in the right place but I've no idea what it is
- should I suspect the video ram board?

D
20190711_160140-s.jpg
20190711_160135-s.jpg

User avatar
hoglet
Posts: 8359
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: ATOM EEPROM

Post by hoglet » Thu Jul 11, 2019 3:25 pm

If you hold down the REPT key, do any characters appear on the screen?

If so, then you have a stuck down key, and the affected key is the one printed.

Space bars are particularly prone to sticking.
Last edited by hoglet on Thu Jul 11, 2019 3:32 pm, edited 1 time in total.

User avatar
oss003
Posts: 3043
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: ATOM EEPROM

Post by oss003 » Thu Jul 11, 2019 4:40 pm

Hi,

There is no zero page RAM in, IC's 51,52 .....

Greetings
Kees

User avatar
CMcDougall
Posts: 6690
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: ATOM EEPROM

Post by CMcDougall » Thu Jul 11, 2019 4:45 pm

doubt will be the Video Ram Board, as it a newer ram not the oldskool 2114s.

where is your basic/os rom? Did the postman eat it??! :lol:

I have about 20x old 2764 roms from old junk (ancient 1981 satelite dish firmware) but no way to program apart from a adaptor board to make it like a 27C128....

looks like you need the floating point rom too, I have both as now have a Prime's excellent rom/ram board, probably best way to go!
could sent my roms to try first? think mine just says 'Atom' as Roland needed some bytes of the 'Acorn' to do something else! (don't think he was hungry.... LOL)
ImageImageImage

User avatar
roland
Posts: 3269
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: ATOM EEPROM

Post by roland » Thu Jul 11, 2019 5:04 pm

I Have a spare rom and a floating punt eprom if you want them.
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
1024MAK
Posts: 9035
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: ATOM EEPROM

Post by 1024MAK » Thu Jul 11, 2019 6:51 pm

Acorn Atom basic hardware information

Zero Page RAM:
IC51 and IC52 both 2114 (1K byte).
Mapped to 0x0000 to 0x03FF.

Upper text space / Video RAM:
IC32 to IC43 (all 2114).
Mapped to 0x8000 to 0x97FF (6K bytes).
Minimum configuration IC42 to IC43 (1K byte).

Lower text space extension RAM:6
IC10 to IC19 (all 2114) (5K bytes).
Mapped to 0x2800 to 0x3BFF.
None required for a minimum configuration.

IC20 OS/BASIC ROM (shown as a MM52165) pin-out matches 2364 ROM (see https://ist.uwaterloo.ca/~schepers/ROMS ... S/2364.png).
No EPROM or EEPROM will directly fit in the socket unless an adapter board is used.

It’s worthwhile reading the technical manual. See http://electrickery.xs4all.nl/comp/atom/atm/atm.html

Also check for both dry solder joints and sockets where the contact is not making contact with the leg of the chip.

Mark
Last edited by 1024MAK on Thu Jul 11, 2019 8:02 pm, edited 1 time in total.

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Fri Jul 12, 2019 1:00 pm

Thanks for all the pointers everyone.

Good shout Dave but unfortunately that's not done it!

Thanks for the chip pointers - the picture I'd taken was my first attempt at installing 2114's but I'd not completed the job when I realised my EEPROM wouldn't fit!

I'm now more convinced that it is the 8255, it's not just the keyboard that has problems but the inputs to the 6847 are floating around too. (If I touch the Port lines from the 8255 to the 6847 it change mode and shows garbage). So I'll re-re-check the connections to the 6847 - I already did basic continuity checks but it looks like something isn't getting through. I scoped the 8255 pins and it looked like nCS, nRD, nWR are getting through ok.

The main reason I want to try and get the YARRB removed and a real 6502 working is so I can properly scope the pins of the 8255 when its running.

I've tried a lot of reflowing of the solder so I now suspect there's a problem with the IC holder though that looks to have been replaced recently...
roland wrote:
Thu Jul 11, 2019 5:04 pm
I Have a spare rom and a floating punt eprom if you want them.
Yes please! PM me if you'd like some postage money.

For now I'll try and fabricate a EEPROM adaptors - I have plenty of 27128's and 2764's to use but my programmer doesn't support the 23 series and I'll be doing rom hacking.

D

User avatar
hoglet
Posts: 8359
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: ATOM EEPROM

Post by hoglet » Fri Jul 12, 2019 1:25 pm

dominicbeesley wrote:
Fri Jul 12, 2019 1:00 pm
Good shout Dave but unfortunately that's not done it!
OK, then lets try again :D
dominicbeesley wrote:
Fri Jul 12, 2019 1:00 pm
I'm now more convinced that it is the 8255, it's not just the keyboard that has problems but the inputs to the 6847 are floating around too. (If I touch the Port lines from the 8255 to the 6847 it change mode and shows garbage). So I'll re-re-check the connections to the 6847 - I already did basic continuity checks but it looks like something isn't getting through. I scoped the 8255 pins and it looked like nCS, nRD, nWR are getting through ok.
You tried two other 8255's didn't you?

One thought....

If the 8255 Port A outputs are floating, then it's likely this instruction somehow failed:

Code: Select all

FF5F  A9 8A     LDA @#8A   // 1000 1010
FF61  8D 03 B0  STA #B003
Bit 7 = 1 of the write to B003 indicates this is a control word to setup the ports. Bit 4 = 0 indicates port A should be an output.

If either of these bits were written incorrectly, then port A would be it's power up default value, which is an input.

If you suspect the 8255 socket, and plan to change it eventually, then for the time being you could ram a 40-pin turned pin socket inbetween. That has fatter pins, so will make a better connection.

Dave
Last edited by hoglet on Fri Jul 12, 2019 1:25 pm, edited 1 time in total.

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Sat Jul 13, 2019 12:37 am

Thanks Dave, you were close - it was a knackered socket but not one of the 40 year old ones!

It was the turned pin socket soldered on the bottom of the yarrb, a single pin - pin39 was broken internally - when I reflowed the joint the pin part came out with the soldering iron tip. It must have either got fractured in the post or been cracked but just making contact.

It's been very hot in here this evening (and I'm still hung over from last night) so it's taken me longer than it should but here's the long story.

I made up an adaptor socket for the 2764 to mm5126 this afternoon and tried a real 6502 but the situation became much worse. Video was all over the shop and I just couldn't get anything...so I swapped back to the yarrb....and that no longer worked. So I tried swapping back to 2114's in the video ram...still no joy and starting to panic.

Had a break to make, and eat some of, a pair of rather nice bilberry pies with my daughter, thus replenished returned with a clear head.

At this point I had to abandon the move back to a real 6502 as I'd managed to snap off the bent up pin 10 on IC44

After much head scratching I strapped the A/G and A/S signals to low on the 6847 and got something that looked like text and then fiddled about with video ram until I found that IC30 socket was playing up with the 2114s and the plug in replacement had a bent pin....now back to working ok with the yarrb - a relief!

Next I started to scope the signals at the 8255 and I noticed that nCS and nWR looked fairly healthy but nRD didn't - it was a series of positive pulses instead of the expected comb with the odd missing tooth that I would have expected (not really knowing the Atom didn't help). I then traced back through IC 8 and 9 and found that pin 1 of IC9 wasn't toggling as expected. Then a bit of confusion and lots of testing the motherboard and sockets - I even tried a "princess and the pea" pile of 40pin turned pin sockets to lift the yarrb above the heatsink to let me get in. That's when I decided I'd reflow the joints and voila.

I'm too tired to get out the desolder station now but hopefully tomorrow I'll have a working machine! For now I've tacked a jump lead from the 6502's pin 39 to pin 1 of ic9.

D

Thanks for everyone's help
Attachments
20190713_012636-s.jpg

User avatar
hoglet
Posts: 8359
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: ATOM EEPROM

Post by hoglet » Sat Jul 13, 2019 6:29 am

dominicbeesley wrote:
Sat Jul 13, 2019 12:37 am
Thanks Dave, you were close - it was a knackered socket but not one of the 40 year old ones!

It was the turned pin socket soldered on the bottom of the yarrb, a single pin - pin39 was broken internally - when I reflowed the joint the pin part came out with the soldering iron tip. It must have either got fractured in the post or been cracked but just making contact.
That's a great bit of detective work.

I wonder if the pin had been bent and straigtened at some point in the past? That's the only time I've seen a turned pin socket pin break off.

Dave
Last edited by hoglet on Sat Jul 13, 2019 8:00 am, edited 1 time in total.

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Sat Jul 13, 2019 8:19 am

I've not had a chance to desolder it yet but thsts what i suspect. The turned pin socket has been cut in half before soldering but as you say they're usually tough. I did get some good coloured turned pin headers a couple of years ago which were incredibly fragile and easy to bend but didn't stand being bent back.

I can't say my detective work was all that great it took me a long whIle to suspect that nRD signal. In my defence I'd never used a 8255 before and when I first scoped it just assumed it was a positive enable.

D

User avatar
roland
Posts: 3269
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: ATOM EEPROM

Post by roland » Sat Jul 13, 2019 9:05 am

dominicbeesley wrote:
Sat Jul 13, 2019 8:19 am
The turned pin socket has been cut in half before soldering but as you say they're usually tough.
Can you post a picture of this please? I really don't recall that I ever cut in half a turned pin socket for a Yarrb board. And I still don't get it what could have caused this issue during transit because the Atom has always worked fine - I even tested it before shipping.

Anyway I am glad you got it working =D> Never forget: if the main board is in good condition we can get every Atom back to work !

Do you still want or need the ROM and FP Eprom? Can I send them to the address I send the Atom to?
256K + 6502 Inside
MAN WOMAN :shock:

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Sun Jul 14, 2019 12:01 pm

Hi Roland,

Here's a picture. Looks like a turned pin socket on the bottom with male to male header to make a stand-off. A perfectly good solution except for the cracked pin.

D
Attachments
20190713_131651-s.jpg

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Sun Jul 14, 2019 12:37 pm

I got the desolder station out and took the header off. As can be seen the top pin showed the same symptom. Unforunately my desolder gun has got the pin wedged some I'm having to strip that down!

I've soldered another turned pin socket on that side - I didn't bother with the other side as it seems ok and there's more risk in desoldering and lifting a track so I'll leave well enough alone!

D

User avatar
roland
Posts: 3269
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: ATOM EEPROM

Post by roland » Sun Jul 14, 2019 9:07 pm

This picture rings a bell. I used the two half sockets just in case a pin of the male-male strip breaks off. Then I only have to replace the broken strip. This was my first Yarrb board. In later boards I soldered the male-male strip directly on the board and I put a socket on these strips. This is for getting some space between the other component on the main board and the Yarrb board and it's also still a protection for the pins on the Yarrb board ( I only have to replace the socket if a pin breaks).

I'm still wondering how that pin got cracked in transit .... :-k
256K + 6502 Inside
MAN WOMAN :shock:

dominicbeesley
Posts: 799
Joined: Tue Apr 30, 2013 11:16 am
Contact:

Re: ATOM EEPROM

Post by dominicbeesley » Sun Jul 14, 2019 11:17 pm

It's a valid enough thing to do - it's what I do with some of my boards as the stand off headers are expensive and delicate!

I'd say these could have been cracked all along and making a press contact and a knock has opened them up. We'll never know what caused it but it had been shaken about a fair bit in transit

D

Post Reply