Repairing a broken Archimedes A3000 ... Update 14.05.17 - mouse progress - the mouse fully works too! I think, it's done

discuss the archimedes & risc pc, peripherals and risc os/risc os on pi
YuT666

Repairing a broken Archimedes A3000 ... Update 14.05.17 - mouse progress - the mouse fully works too! I think, it's done

Post by YuT666 » Wed Mar 22, 2017 8:28 pm

I got an A3000 some days ago and i'll revive this one in the next weeks (cleaning, fitting with a AA battery holder and other things).

Here are some pics of the state, after i received the Archi (not cleaned or something else ... the typical battery problem) ...
Archi0.jpg
Archi1.jpg
Archi2.jpg
Archi3.jpg
Archi4.jpg
After the first cleaning (with 50%water & 50% vinegar) i've seen all the crappy mess. The condition of this area is absolutely bad (my opinion). Many scratches (maybe the old battery and the pieces of rust) and the two resistors L8 & L9 simply fell off. The crystal is broken anyway. Zero-OHM with 0,25??? Never heared of Zero-Ohm (i only know them from 1 to xxx).

It looks like, that i should replace ALL the resistors from L6 to L14 (are maybe the upper row L7-L9 and L11/L12. The diode D2 looks bad too.
ArchiMess.jpg
Garbage ...

Received the parts today (some crystals, zero-ohm resistors & D2 diodes). I'm a bit busy at work, so i hope i'll receive some results at the weekend and maybe ... this little beauty will run again ...
Last edited by YuT666 on Sun May 14, 2017 1:39 pm, edited 16 times in total.

User avatar
daveejhitchins
Posts: 5721
Joined: Wed Jun 13, 2012 6:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by daveejhitchins » Wed Mar 22, 2017 8:47 pm

Have you had the white vinegar on there yet? First job, if not - paint everything that looks blue and run it under as many components as you can. Then start the clean-up. I wouldn't fit your new parts until all the blue has gone. Note, also, that the solder will take some re-flowing. Get hold of some flux to make it a little easier.

Good luck =D>

Dave H :D
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ARA III, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Wed Mar 22, 2017 9:01 pm

Thanks for your answer ...

Well, no ... it was a brandy vinegar. I'll try some white vinegar (white wine).

User avatar
daveejhitchins
Posts: 5721
Joined: Wed Jun 13, 2012 6:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by daveejhitchins » Wed Mar 22, 2017 9:45 pm

No, no, no . . . You're not meant to drink it :lol:

Dave H :lol:
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ARA III, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

User avatar
1024MAK
Posts: 10080
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by 1024MAK » Thu Mar 23, 2017 5:31 pm

Picture here.

Mark

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Thu Mar 23, 2017 6:59 pm

Thanks ...
daveejhitchins wrote:No, no, no . . . You're not meant to drink it :lol:

Dave H :lol:
I don't drink any alcohol ... but maybe i should ... to fix the Archi ... :mrgreen:

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Fri Mar 24, 2017 4:55 pm

Well, there's a problem (or maybe not) ... the original zero ohm resitors have thicker wires as the ones i bought. I don't know, if this is important.

User avatar
1024MAK
Posts: 10080
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by 1024MAK » Fri Mar 24, 2017 5:40 pm

YuT666 wrote:Well, there's a problem (or maybe not) ... the original zero ohm resitors have thicker wires as the ones i bought. I don't know, if this is important.
Should not make any difference.

Mark

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Fri Mar 24, 2017 8:30 pm

:-k ... i'll try ...

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Sun Mar 26, 2017 12:16 pm

It is nearly impossible to get rid of all these battery remains ... i tried a thousand times. I'm a bit disappointed and I'll clean the board with a circuit board cleaner too (the Sonax Electro Cleaner is a good one)..

I replaced ALL the zero ohm resistors (they had all rusty wires and most of them fell off), the crystal and the D2 diode. After this i solder the battery wires with a diode.

Before:
ArchiMess.jpg
After:
Archi2.jpg

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Mon Mar 27, 2017 5:04 pm

I did a first test withouth the RAM expansion and the disc drive ...

Don't know, if this is a good sign.
A3000.jpg

Zarchos
Posts: 2355
Joined: Sun May 19, 2013 9:19 am
Location: FRANCE

Re: Repairing a broken Archimedes A3000 ...

Post by Zarchos » Mon Mar 27, 2017 6:01 pm

Yes it's rather a good sign.

Now : reset the CMOS values to their default values, by pressing DELETE while powering on the machine.

It should give you a working Archie, with default, factory CMOS values.
If not, enter STATUS and if you've got some fancy values, I think you'll have to change the CMOS chip (not difficult to find or do and it's cheap).

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Mon Mar 27, 2017 7:11 pm

Thanks for your help, but i encountered a problem ... some keys don't work (DEL too). I think, that there is still some corrosion within the keyboard connectors, 'cause the keyboard membrane looks fine. I try to clean the connectors with a electronic cleaner, which will arrive tomorrow. I'm busy at work again, so i hope, that the whole thing will run fine the next weekend.

Zarchos
Posts: 2355
Joined: Sun May 19, 2013 9:19 am
Location: FRANCE

Re: Repairing a broken Archimedes A3000 ...

Post by Zarchos » Mon Mar 27, 2017 7:28 pm

YuT666 wrote:Thanks for your help, but i encountered a problem ... some keys don't work (DEL too). I think, that there is still some corrosion within the keyboard connectors, 'cause the keyboard membrane looks fine. I try to clean the connectors with a electronic cleaner, which will arrive tomorrow. I'm busy at work again, so i hope, that the whole thing will run fine the next weekend.
I hope so.
I've got some dead machines because of corrosion in the keyboard connectors area.
Fortunately some spare connectors are available.

In case you decide to remove the keyboard, pull the 2 strips very gently otherwise you can break some pins inside the connectors if they are corroded ...

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Mon Mar 27, 2017 7:40 pm

Ok, we'll see ...

but another question ... which ICs are responsible for the keyboard and are they get faulty sometimes?

User avatar
cliffh
Posts: 61
Joined: Wed Jan 02, 2013 9:45 am
Location: UK
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by cliffh » Wed Mar 29, 2017 1:34 pm

YuT666 wrote:which ICs are responsible for the keyboard and are they get faulty sometimes?
The 74LS145 decoder/driver chips IC4 and IC5 generate pulses for scanning the keyboard. They are just above the battery location in your photographs and they are vulnerable to battery leakage.

Failure of these chips can disable clusters of keys associated with whole rows of the keyboard wiring matrix. Bad connections and broken tracks can do the same thing, so finding the cause might be a process of elimination.

If you list the failed keys, we might be able to associate the fault with particular rows.

It is fairly easy to snip out and replace the 74LS145s. This cured the keyboard faults on one of my machines. One bonus is that it allows you to clean up and inspect the tracks under the chips.

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Wed Mar 29, 2017 3:26 pm

A big thanks for your help too ...

Well, when i press the keys in the first row (for example), only 5 keys are responding (the ones with a green mark).
Unbenannt 1.jpg
I cleaned the board and the keyboard connectors a 1000 times, but there is no difference. I heared, that some people are cutting 2-3 mm of the keyboard connector foil (the tail) ... but anyway ... i bought some SN74LS145N.

http://www.ebay.de/itm/5x-SN74LS145N-De ... Sw0rdTxVnA

Maybe one of the keyboard connectors (on the board) has still some corrosion ... i don't know, where to get a replacement.

User avatar
cliffh
Posts: 61
Joined: Wed Jan 02, 2013 9:45 am
Location: UK
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by cliffh » Wed Mar 29, 2017 5:56 pm

That pattern of failed keys is consistent as far as it goes. There are two very useful diagrams by jms2 which colour code the keys according to their positions in the keyboard matrix. See this thread:

http://www.stardot.org.uk/forums/viewto ... =16&t=6002
Find the post dated May 04, 2013 and open the diagram called A3000 keyboard rows.png.

You say that ' and 1 work. Can you confirm whether all the keys in red on the diagram (associated with pin 3 on SK7) also work? If so, that row is OK (note that the term row here refers to the wiring matrix, not the physical layout of keys).

Similarly, 2 works, so if all the keys in orange work, that indicates another good row.

4 and 5 don't work, so if all the other keys in light green also don't work, that indicates a bad row.

Working along the keys you have marked, can you confirm a consistent pattern of good and bad rows?

Zarchos
Posts: 2355
Joined: Sun May 19, 2013 9:19 am
Location: FRANCE

Re: Repairing a broken Archimedes A3000 ...

Post by Zarchos » Wed Mar 29, 2017 6:08 pm

YuT666 wrote:A big thanks for your help too ...

Well, when i press the keys in the first row (for example), only 5 keys are responding (the ones with a green mark).

Unbenannt 1.jpg

I cleaned the board and the keyboard connectors a 1000 times, but there is no difference. I heared, that some people are cutting 2-3 mm of the keyboard connector foil (the tail) ... but anyway ... i bought some SN74LS145N.

http://www.ebay.de/itm/5x-SN74LS145N-De ... Sw0rdTxVnA

Maybe one of the keyboard connectors (on the board) has still some corrosion ... i don't know, where to get a replacement.
Yes cutting the tail can be a good idea if you see there's no more the black conductive substance.
Also, try to spray some 'special electronics contacts spray' on the connector and in the area, it could solve the problem.
You could also try to reflow the solders (from the other side of the motherboard).

If you need some keyboard connectors, they are available from Farnell IIRC, just use the search engine to know where to get them from exactly.
If no more are available, I've got a stock somewhere (they are so hard to find that I thought stockpiling them would always be useful one day to me or other people like you needing help to resurrect an Archie).

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Wed Mar 29, 2017 6:15 pm

I thought, Farnell is only for dealers ...

So i could do the thing you mentioned ... reflowing the solders of the connector. I already used an electronic cleaner for the connectors.

@cliffh:

Yes, i can confirm it. What does it mean?

User avatar
cliffh
Posts: 61
Joined: Wed Jan 02, 2013 9:45 am
Location: UK
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by cliffh » Wed Mar 29, 2017 7:51 pm

YuT666 wrote:Yes, i can confirm it. What does it mean?
It means there are failed rows on the input side of the keyboard, and you should probably focus on the 74LS145s, the integrity of the tracks leading to and from these chips and the keyboard socket SK7. It doesn't pin down the location of the faults precisely, but it rules a lot of things out - because it is highly unlikely that anything else would produce this particular pattern of key failures.

The supervisor prompt confirms that most of the machine is working fine. I hope you manage to fix the remaining problems - prospects look good to me.

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Wed Mar 29, 2017 8:27 pm

Maybe one or more of the ic legs are damaged, so i should replace both 74LS145 anyway. I think, i'll solder two sockets.

RobC
Posts: 2921
Joined: Sat Sep 01, 2007 10:41 pm
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by RobC » Wed Mar 29, 2017 8:43 pm

I'd definitely buzz out the tracks from the keyboard connectors to the ICs first. The connectors and the tracks are prone to corrode - I fixed mine with some kynar wire.

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Thu Mar 30, 2017 8:29 am

The connector is still corroded and many tracks leading to the connector are rusty. Looks like a bigger problem. Maybe i should use a fibreglass pencil to clean the tracks, but i don't want to destroy more.
A3000.jpg

User avatar
DutchAcorn
Posts: 2250
Joined: Fri Mar 21, 2014 9:56 am
Location: Maarn, Netherlands
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by DutchAcorn » Thu Mar 30, 2017 9:04 am

YuT666 wrote:The connector is still corroded and many tracks leading to the connector are rusty. Looks like a bigger problem. Maybe i should use a fibreglass pencil to clean the tracks, but i don't want to destroy more.

A3000.jpg
Best to check continuity of the tracks and bypass tracks that are faulty using a bit of wire. Not much point in "cleaning" them imo, as you say it will not fix anything and you may make things worse.
Paul

Image

User avatar
jms2
Posts: 2616
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by jms2 » Thu Mar 30, 2017 1:00 pm

I had the same problem with mine. I don't think it is possible to buy a new keyboard socket, but I would be delighted if someone was able to source an equivalent.

The best thing to do is use a multimeter to check continuity of all the tracks in the area, and then repair the bad ones with Kynar wire. You will probably find that there are not too many broken tracks.

User avatar
paulv
Posts: 3854
Joined: Tue Jan 25, 2011 6:37 pm
Location: Leicestershire
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by paulv » Thu Mar 30, 2017 1:47 pm

jms2 wrote:I had the same problem with mine. I don't think it is possible to buy a new keyboard socket, but I would be delighted if someone was able to source an equivalent.

The best thing to do is use a multimeter to check continuity of all the tracks in the area, and then repair the bad ones with Kynar wire. You will probably find that there are not too many broken tracks.
Your wish is my command ;)

http://www.stardot.org.uk/forums/viewto ... rd#p136785

Paul

User avatar
jms2
Posts: 2616
Joined: Mon Jan 08, 2007 6:38 am
Location: Derby, UK
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by jms2 » Thu Mar 30, 2017 5:15 pm

Ok, thanks, so it looks like Zarchos has them. I'd missed that. :D

Now, my A3000 is working ok despite having some corrosion in the sockets, but I'd say that if you are fault finding then renewing the sockets would be a smart move.

YuT666

Re: Repairing a broken Archimedes A3000 ...

Post by YuT666 » Wed Apr 05, 2017 6:28 pm

I cleaned the whole keyboard (i disassembled and cleaned it, so it looks like new) and the board again ... some more keys are working now. But there are still many not working keys. Delete works too, but i can't reset the Archie with powering on while pressing Delete. I only get to the supervisor prompt.

So should i cut 2-3mm of the keyboard connector foil anyway? Or resolder the keyboard connector joints? I can't find any cutted tracks (from the connectors to the 74LS145 ICs) and replacing the 74LS145s would be my very last try.

User avatar
DutchAcorn
Posts: 2250
Joined: Fri Mar 21, 2014 9:56 am
Location: Maarn, Netherlands
Contact:

Re: Repairing a broken Archimedes A3000 ...

Post by DutchAcorn » Wed Apr 05, 2017 9:40 pm

YuT666 wrote:... Delete works too, but i can't reset the Archie with powering on while pressing Delete. I only get to the supervisor prompt.
This can be a sign of a problem with the CMOS chip. Do you have enough keys to type *status? Or *st.?

You can try to set a configuration option and check if it is retained. Even if there is no battery connected it should keep the settings as long as you don't switch it off. E.g. you can set the default mode to 1 using "*co. mo. 1" (abbreviated). Resetting should then get you to mode 1.

If it does not, there is probably an issue with either the CMOS chip itself or the connections to the chip.

YuT666 wrote:So should i cut 2-3mm of the keyboard connector foil anyway? Or resolder the keyboard connector joints? I can't find any cutted tracks (from the connectors to the 74LS145 ICs) and replacing the 74LS145s would be my very last try.
Do you have a way to measure the connectivity of the keyboard to the PCB before trying to fix those?
Paul

Image

Post Reply

Return to “32-bit acorn hardware”