A5000 intermittent fault

Arc/RPCs, peripherals, RISCOS operating system & ARM kit eg GP2x, BeagleBoard
Post Reply
steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

A5000 intermittent fault

Post by steve3000 » Sat Mar 28, 2015 10:04 am

I'm after some suggestions here as my A5000 has developed an intermittent fault.

I replaced the HDD last week and when I switched back on afterwards, the HDD spun up correctly along with the fan, but the computer itself was totally dead, with no video output and only a brief flicker of the keyboard lights when power applied.

So I double checked everything I'd moved (to replace HDD, you need to remove and refit the backplane from the PCB, I also replaced the IDE cable at the time, so checked this), and all looked ok, but still no response from the computer.

Next step I removed the PCB and continuity-tested the usual suspect tracks around the CMOS chip (which I'd repaired two years ago) but everything checked out fine. Then I used my logic probe to test various ICs for signs of life. The Arm3 is active after switch on, with very brief activity on the data bus (less than 1 second), then nothing. I confirmed this at the ROM chips and found the address lines active briefly after power-up, then they freeze.

Further investigation and holding the RESET button on the keyboard brings back life to the data and address buses, but releasing the button and everything freezes.

So, having given up on a quick fix, last night I popped the PCB back into the computer case, closed everything up and tried one last switch-on, and as expected, saw nothing. Then I pressed RESET and surprisingly the computer started - booted RISC OS and fully working! :)

The problem is clearly intermittent, after switching off and on again last night it booted fine, but leaving it overnight and it's dead again. Pressed RESET a few times and still nothing.

Any thoughts on what could cause this? I suspect a dry solder joint or fractured track, but where to start...

User avatar
1024MAK
Posts: 8106
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: A5000 intermittent fault

Post by 1024MAK » Sat Mar 28, 2015 10:27 am

Do you see a difference between a power up and a restart after pressing reset?

Maybe confirm that the reset circuit is working correctly by using the logic probe?

An intermittent dry joint can be very frustrating to find. Try gently tapping each corner of the PCB with the insulated handle of a screwdriver to see if the vibration causes any symptoms...

Mark

User avatar
flaxcottage
Posts: 3328
Joined: Thu Dec 13, 2012 8:46 pm
Location: Derbyshire
Contact:

Re: A5000 intermittent fault

Post by flaxcottage » Sat Mar 28, 2015 11:11 am

This might sound silly but is the CMOS circuit charging correctly? Mine had similar symptoms to yours until I found an intermittent open-circuit track and repaired it.
- John
Image

sirbod
Posts: 904
Joined: Mon Apr 09, 2012 8:44 am
Location: Essex
Contact:

Re: A5000 intermittent fault

Post by sirbod » Sat Mar 28, 2015 10:36 pm

Unrelated, but my Iyonix does this - it damned annoying. Have to keep hitting the RESET button on the front until it decides to boot.

Definitely sounds like an issue with the reset circuit (IIRC), like it's not pulling the reset line low and getting stuck in an infinite reset loop.

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 12:37 am

Thanks all for the suggestions - I've been out all day, but back now and spent some time running through these, with interesting results...

Firstly, tapping the corners of the PCB while running a BASIC program in a graphics intensive screen mode (high memory bandwidth). The tapping had absolutely no effect, even with heavy 'tapping' (so the PCB flexed). I also 'tapped' a few ICs for good measure (all the ARM ones and a few others, RAM, CMOS chip, latches, ROM, etc) - still no effect, and my BASIC program continued to run happily.

So perhaps not a dry joint?

Second I checked the CMOS circuit continuity. This could easily be suspect, because it required major repair surgery when I fixed the battery leakage. But everything checked out, no open circuits (including the C0/C1 tracks to IOC), battery was charged and the CMOS chip was keeping time nicely when switched off.

So not the CMOS circuit?

Next up - powering on and RESET differences. After several on/offs while testing the above, I noticed some possible logic to this 'intermittent' fault. When powering up after a gap of a few minutes (or longer), the computer fails to boot and needs a Ctrl-RESET in order to boot. However if I switch off then immediately (10 seconds later) switch back on, it starts ok... This seems fairly reproducible.

Also (and this may be random but...) on one occasion I pressed RESET (not Ctrl-RESET) and got a partial boot - frozen cursor on screen (all previous failed boots had no video output), pressing Ctrl-RESET then got it to boot.

So, as suggested by Jon and Mark, this is looking like a reset-circuit/related issue!

I'd like to continue testing, but losing an hour tonight and need to be up early, so no more for now. I'll dig out the A5000 schematics tomorrow. :)

sirbod
Posts: 904
Joined: Mon Apr 09, 2012 8:44 am
Location: Essex
Contact:

Re: A5000 intermittent fault

Post by sirbod » Sun Mar 29, 2015 6:06 am

steve3000 wrote:this is looking like a reset-circuit/related issue!

I'd like to continue testing, but losing an hour tonight and need to be up early, so no more for now. I'll dig out the A5000 schematics tomorrow. :)
I believe the circuit is in the Keyboard, do you have a spare you can try?

A5000 keyboard schematic:
A5000_keyboard.png

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 12:59 pm

sirbod wrote:I believe the circuit is in the Keyboard, do you have a spare you can try?
Hi Jon, interesting suggestion - I hadn't thought about the keyboard.

I had assumed the problem would be on the main PCB, as the keyboard switch is generally successful in getting the computer to boot, when switch-on fails... But it's a simple test, so I connected up my trusty work-horse A310 keyboard (which is known-good as it usually moves between everything apart form the A5000). Alas, it gave the same response.

I've now switched the A5000 on/off more than 20 times while observing this very closely and majority of those times it refuses to boot until I press the reset button. On just a couple of occasions pressing the reset button doesn't work first time but a few more/longer presses always get it to boot in the end, and once booted, it remains stable.

So I'm now even more convinced it's something on the reset circuit, specifically the power-on-reset, which isn't behaving.

I'm going to delve further into the circuit diagrams now...

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 1:14 pm

Ok, decoding function of circuits really isn't my expertise, so maybe someone can check the following pics?

I've found what appears to be the trigger for the IOCs Power-on interrupt. I don't think there is a problem with this circuit, as this is just sets an IOC interrupt register via pin 28 of the IOC (you can then read the Por register by software to determine if the computer has just switched on). But this circuit is useful to compare to the 2nd pic below...
por_ioc.png
The 2nd pic (section in red box) shows a similar set-up to the IOC pin 28 trigger, however this time it acts directly on the Reset* line. Does this cause the system to reset at power on? And could this be the source of my A5000's problem? (eg. if that C28 capacitor was leaky?)
reset_at_power_up.png

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 1:48 pm

Ok, ignore questions in the last post above - I don't think this is the problem. I've checked with logic probe and Reset* is low for a fraction of a second at start-up. And comparing to my A4000, it gives a similar fraction of a second low on Reset*...

So is it more likely to be something further along? I'm thinking the circuit (below) which generates Rst* from Reset* could be suspect.

Rst* is the reset feed to all the ICs - so if the fraction-of-a-second that Reset* is low, isn't enough to trigger the following circuit to generate Rst*, then no start-up reset takes place... But manually holding the keyboard reset switch down for a few seconds could give longer for the circuit below to generate Rst*???
rst_from_reset.png
Does any of this sound reasonable?

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 3:05 pm

Ok, forget the last post - my A5000 is a later model fitted with the IOEB, so has completely different reset circuitry to that shown in the TRM (non-IOEB model). :(

And of course, it looks like Reset* to Rst* is handled by the IOEB on my computer... But thankfully the A4 TRM drawings show the IOEB :)

Below is the A4 power on reset circuit feeding to the IOEB. This could well be what the A5000 uses and corresponds to the box in red in the diagram a couple of posts up. The internal reset switch is fitted after the capacitor/resistor, so a manual reset would bypass a fault here (and note the external keyboard reset switch is ignored by the A4, so not wired in here).
A4_reset_to_rst.png

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 3:30 pm

After a bit of track hunting, I've figured out the A5000 IOEB power on/reset circuit and (after cutting/pasting a bit of the A4 circuit diagram) this is how it works on my A5000...
a5k_ioeb_reset_at_power_up.png
So the question is...based on this diagram, if either of C28, R35 or D11 are faulty, could this cause my power-up problem?

User avatar
1024MAK
Posts: 8106
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: A5000 intermittent fault

Post by 1024MAK » Sun Mar 29, 2015 4:04 pm

Maybe, the first thing to try is to wire a 10uF (or similar value) capacitor in parallel with the 4.7uF capacitor and see if that improves things.

In most power up reset circuits, it is the electrolytic capacitor that is the cause of the problem.

Mark

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sun Mar 29, 2015 10:12 pm

Ok, just tried a 10uF electrolytic cap in parallel with 4.7uF C28, but this didn't seem to make a difference, certainly it didn't solve the problem...however I am now wondering if my 10uF cap is any good (it was purchased new, but has been sat at the bottom of my spares box for a few (10+) years...!).

I have known-good 47uF and 220uFs, otherwise I'll pop to maplin at lunch tomorrow and grab a new selection.

User avatar
1024MAK
Posts: 8106
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: A5000 intermittent fault

Post by 1024MAK » Sun Mar 29, 2015 10:43 pm

If you use a 47uF or a 220uF, you should be able to see the voltage change, assuming you are using a good quality digital meter.

If the voltage rises after switch on and decays after switch off, the diode and resistor are likely okay, but the 33k ohm resistor can be tested on the meters resistance range.

Note that using a 220uF capacitor will cause a delay of 5 to 10 seconds... You can't miss that!

Mark

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Mon Mar 30, 2015 4:34 pm

Ok, update...

The voltage rises and falls nicely with a 47uF cap in parallel, but still the computer doesn't start without a manual reset.

So I tested the resistance of the two resistors. R34 correctly gets 100 ohms, but R35 only measures 2.5 kohms (expecting 33 kohms).

Could this cause the problem?

Annoyingly it's a SMT resistor, which is a pain to remove with a standard soldering iron, but I'll give it a go if this is likely to be the cause of the issue?

User avatar
1024MAK
Posts: 8106
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: A5000 intermittent fault

Post by 1024MAK » Mon Mar 30, 2015 6:18 pm

If the voltage rises nicely to +5V with the 47uF capacitor, the value of the resistor is less of an issue if the computer does not start up. Also keep in mind that if you cannot lift the resistor (to isolate it), the meter will also measure normal parallel leakage paths as well as the resistance of the resistor which will result in the meter showing a lower value. You can also try the leads with the other polarity.

One question: if you leave it off for more than one hour, does it start up the first time you switch on, or do you have to press the reset switch to get it going?

From what you say, the resistor / capacitor part of the reset circuit appears to be working correctly. Time to look at the next section...

Mark

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Mon Mar 30, 2015 7:39 pm

So I pulled the resistor, was a little easier than expected, just tricky not to lose it...

It wasn't at fault, 33k exactly.

Couldn't think of anything else to try, so popped it back on. Then thought about the psu...figured this could be at fault, so pulled a PC psu, and tested the A5000 board. But this didn't solve the problem.
1024MAK wrote:One question: if you leave it off for more than one hour, does it start up the first time you switch on, or do you have to press the reset switch to get it going?
Still have to press reset.

The only time I don't always need to press reset is if I switch off for less than 10 seconds, then it usually switches back on and boots.
1024MAK wrote:From what you say, the resistor / capacitor part of the reset circuit appears to be working correctly. Time to look at the next section...
Yep I agree.

However there is no 'next section '. The feed goes straight into the IOEB...

I'm reasonably sure the IOEB is working because a manual reset works through the same pin.

So I'm running out of ideas. Any thoughts?

sirbod
Posts: 904
Joined: Mon Apr 09, 2012 8:44 am
Location: Essex
Contact:

Re: A5000 intermittent fault

Post by sirbod » Tue Mar 31, 2015 6:51 am

steve3000 wrote:So I'm running out of ideas. Any thoughts?
Suspect ROM's or bus interface to them?

Sounds like you've ruled out the reset circuit and PSU, so probably need to "think outside the box" !!

Is there any indication that POST has started? Could it be hanging on a check?

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Fri Apr 03, 2015 8:57 am

sirbod wrote:Suspect ROM's or bus interface to them?

Sounds like you've ruled out the reset circuit and PSU, so probably need to "think outside the box" !!

Is there any indication that POST has started? Could it be hanging on a check?
Good thoughts - I swapped out the ROMs with a set from my A310, and got exactly the same problem, so it's not the ROMs.

As for the bus interface, I decided to get the computer running and see how long it would run. 8 hours later (overnight) and it was still happily running. Switched it off for 1 minute, then back on and it was dead again, until Reset pressed...

So, as it's happy running for hours without any issue, I must assume all the Arm chips and I/O chips are fine.

One interesting development which I hope may help identify a cause... Most of the time no start up happens at all without pressing Reset (as discussed above), but just a couple of times now I've seen a partial startup after pressing Reset, then a freeze. What I see on those occasions, is usually a flashing cursor which then freezes (before it gets round to displaying 'RISC OS') but once I saw an error displayed "Error: AddressException: Address exception at &%0 (Error number &80000003)" followed by a '*' prompt, but the computer had frozen, so no input possible.

Thoughts?

User avatar
BeebMaster
Posts: 2632
Joined: Sun Aug 02, 2009 4:59 pm
Location: Lost in the BeebVault!
Contact:

Re: A5000 intermittent fault

Post by BeebMaster » Fri Apr 03, 2015 11:17 am

I'm pretty sure I had this with my A5000, it's had two little paddies on me over the years, deciding it doesn't want to work.

The most recent fault which I've only just resolved in the last few days seemed to be mainly due to corrupted CMOS settings and not having the correct monitor type connected so I could see what was happening.

The fault a few years ago was similar, but I did have a monitor connected and I think most of the time I was getting the red border or red screen and nothing else and sometimes some incomprehensible RISC OS error and a supervisor prompt.

I did eventually get it going again with repeated use of power on CMOS reset, I'm sure you've tried that, but maybe change the battery or take it out and manually configure everything with * commands to see if that helps.
Image

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Sat Apr 04, 2015 4:56 pm

Well, after some further testing today, and a bit of reflowing of the solder on all the ICs within walking distance of the CMOS battery (most likely to be dry/corroded), still no success.

Interestingly it now takes several manual resets, over several minutes before the computer boots, and then it often throws up an internal error (strangely coinciding with the initiation of the 'Messages' module) before getting going.
BeebMaster wrote:The fault a few years ago was similar, but I did have a monitor connected and I think most of the time I was getting the red border or red screen and nothing else and sometimes some incomprehensible RISC OS error and a supervisor prompt.

I did eventually get it going again with repeated use of power on CMOS reset, I'm sure you've tried that, but maybe change the battery or take it out and manually configure everything with * commands to see if that helps.
CMOS reset with Delete+switch on doesn't work, because switching on doesn't work :( (and Delete+reset doesn't reset the CMOS) However, I have tried manual reconfiguration by writing a quick BASIC prog to wipe the CMOS, including the 'unused' and 'reserved' bytes, then manually reconfiguring using *configure, but no help...

But I am starting to think the CMOS chip may be the culprit here. I can't think of anything else I've not checked/tested which could lead to the effects I'm seeing - but a gradually dying CMOS chip might just be the cause... Nothing to loose by changing the CMOS chip, so I'll get one ordered and see where that takes us.

User avatar
hoglet
Posts: 7694
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: A5000 intermittent fault

Post by hoglet » Sun Jul 01, 2018 6:20 pm

steve3000 wrote:
Sat Apr 04, 2015 4:56 pm
But I am starting to think the CMOS chip may be the culprit here. I can't think of anything else I've not checked/tested which could lead to the effects I'm seeing - but a gradually dying CMOS chip might just be the cause... Nothing to loose by changing the CMOS chip, so I'll get one ordered and see where that takes us.
I know this thread is three years old....

Did you try replacing the CMOS RTC/RAM chip?

Anything other developments to report?

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Mon Jul 02, 2018 12:20 am

Well there's a blast from the past!

Not much to report I'm afraid. I ordered the replacement CMOS RAM chip, but it took a few weeks to arrive, then my wife gave birth and this computer got rapidly packed away into the loft. I later picked up a fully functional A5000 from a friend so had no urgent need to repair the original...so it's still up in the loft!

As my daughter is now 3, and life has got a bit more back to normal, I guess I should really dig that A5000 out and fit the chip... I'll add this to my list for the summer :)

Frank990
Posts: 1
Joined: Fri Aug 31, 2018 7:52 pm
Contact:

Re: A5000 intermittent fault

Post by Frank990 » Wed Sep 19, 2018 5:27 pm

Hello.

I was avidly reading this thread, as I had a very similar problem with my A5000.
For me, the battery leaked, so I fixed the tracks, and replaced battery. This didn't improve things much, so I replaced the CMOS chip too.
Now it was remembering the CMOS settings, but still took multiple resets on boot, and gave 'address exception &%0' a lot.

What would happen is that each time I used the keyboard reset button, it would get a little further in the boot sequence, but would freeze in odd places, like even while playing the power on 'boop' sound, so the sound would continue forever. A Crtl+delete reset generally got to the desktop much faster.

So, inspired by this thread, I started checking the power on reset circuitry.
For me, C68 was saying 'no capacitance' when I tested with my capacitor tester.

I replaced that, and now it works fine!

I think that many of those little silver can surface mount caps might be going bad in mine, as I think I caught a little whiff of that fishy bad capacitor smell a few times while working on the board. I can't be totally sure though.

Anyway, hope that's some help to someone.

User avatar
danielj
Posts: 6692
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A5000 intermittent fault

Post by danielj » Wed Sep 19, 2018 5:57 pm

That's a timely warning. It might be time to give them a check up... Certainly the Amiga is plagued by SMD cap issues, we've no reports in Acorns but that's not to say it won't happen...

d.

steve3000
Posts: 1917
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A5000 intermittent fault

Post by steve3000 » Thu Sep 20, 2018 12:20 am

Frank990 wrote:
Wed Sep 19, 2018 5:27 pm
So, inspired by this thread, I started checking the power on reset circuitry.
For me, C68 was saying 'no capacitance' when I tested with my capacitor tester.
Ahhh, thanks for sharing!

I've still not got back to my dead A5000, but it sounds like you had exactly the same symptoms! I had even looked at the reset circuitry but had not tested capacitors... I'll do this and report back as soon as I find an evening to unpack that A5000... :)

Post Reply