This means the CMOS RAM has lost it's content.
Most common cause is a the battery having died.

A couple things to check.

f you have a voltmeter, measure the battery and see if it's the right voltage
(if a common CR2023 - that would be 3V)

If it is, then there may be a problem/corrosion in the battery holder
or leads going it the RAM. If you can check the voltage at the RAM chip
itself that could give a good clue.

If it's not, change it again and make sure you use a good/new one.
If you have a low-current meter, see how much current is being drawn
when everything is OFF (should be nanoamps)

If you don't have a low-current meter, leave it as long as it took to
fail after the last time you changed it and see if it's still 3v.

If not, there is probably "excessive draw" somewhere, could be crap
sticking on the board or it's components, a failing cap. etc.

If it still fails and the battery consistently shows high enough voltage, the
circuitry looks good - either the RAM chip is failing (unlikely as it keeps it's
mine while power is on) - or something is selecting/activating it while the
system is powered down.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

DaveDDS wrote on 2025-01-30, 01:17:

This means the CMOS RAM has lost it's content. Most common cause is a the battery having died. […]
Show full quote

This means the CMOS RAM has lost it's content.
Most common cause is a the battery having died.

A couple things to check.

f you have a voltmeter, measure the battery and see if it's the right voltage
(if a common CR2023 - that would be 3V)

If it is, then there may be a problem/corrosion in the battery holder
or leads going it the RAM. If you can check the voltage at the RAM chip
itself that could give a good clue.

If it's not, change it again and make sure you use a good/new one.
If you have a low-current meter, see how much current is being drawn
when everything is OFF (should be nanoamps)

If you don't have a low-current meter, leave it as long as it took to
fail after the last time you changed it and see if it's still 3v.

If not, there is probably "excessive draw" somewhere, could be crap
sticking on the board or it's components, a failing cap. etc.

If it still fails and the battery consistently shows high enough voltage, the
circuitry looks good - either the RAM chip is failing (unlikely as it keeps it's
mine while power is on) - or something is selecting/activating it while the
system is powered down.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Hi Dave. Thanks for the suggestions. Unfortunately, I don’t have a voltmeter. Yet. It’s something I’ll certainly have to look into for these situations. One thing I didn’t do this time, that I’ve always done before and was never an issue, was get the system via the BIOS to detect/configure each IDE drive. It seems to do that itself now on startup anyway. I don’t see how that would be the cause, but so far, it seems to be fine. Of course, it could all start up again tomorrow.
I believe the battery is a standard 3V one, just like the one I took out. The RAM is different though, from what I was using before, but it’s 133 MHz instead of 100. The system has set it to 100 as that’s all it supports. Though could that be an issue…? Maybe not the core memory clock, but the RAM itself. I might have to run some memory test.

It's not the RAM that will be at fault, it's the memory in the CMOS chip. What DaveDDS describes as CMOS RAM (I had forgotten what they call that). This is the small amount of static ram that stores BIOS settings, and the CMOS battery that you replaced provides the power to enable it to store the settings long term. The settings are lost after a period with no power, but not straight away (hence how you can remove the battery to replace it, but leave it out overnight without losing settings). This is different from the main RAM that you you are describing as being 133 MHz, which only holds data for short bursts while the main power is switched on.

I hope I haven't just made it sound more complicated than it is, I'm not as good a "teacher" as I once was 😀 I just don't want you wasting hours running memtest when that test the wrong component 😀

DustyShinigami wrote on 2025-01-30, 01:38:

Hi Dave. Thanks for the suggestions. Unfortunately, I don’t have a voltmeter. Yet. It’s something I’ll certainly have to look into for these situations.

I believe the battery is a standard 3V one, just like the one I took out. The RAM is different though, from what I was using before, but it’s 133 MHz instead of 100. The system has set it to 100 as that’s all it supports. Though could that be an issue…? Maybe not the core memory clock, but the RAM itself. I might have to run some memory test.

A decent "multimeter" is definitely worth having when you are trying to figure out pretty much anything electrical.
You can pay a lot for a high-end one, but a simple basic one is quite cheap and all you need most of the time
(and infinitely better than none)

A lot of (most?) BIOS will auto-detect IDE/SATA drives ... the on-board firmware of those has an ID command
where it will report the type and configuration of the drive) - rarely in the last 20+ years have I had to manually
set up a drive type.

The CMOS RAM is not the system RAM - it's a little tiny chip often inside one of the "system chips", mounted
directly to the main board, or sometimes in an IC socket. Never in a "slot" socket like system RAM. It's capacity
is only a few bytes - enough to hold the system settings, but nowhere near enough to run anything out of.

The most common reasons I've seen when CMOS batteries are replaced and immediately fail are:
- Took an old battery from junkbox and it was dead or near dead.
- "Clips" in battery holder have some corrosion or other "junk" on them. Removing/replacing the
battery moves it just enough for this to interfere with conduction. Always a good idea to give these
clips (and the battery itself) a good cleaning when re-inserting.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

debs3759 wrote on 2025-01-30, 02:30:

It's not the RAM that will be at fault, it's the memory in the CMOS chip. What DaveDDS describes as CMOS RAM (I had forgotten what they call that). This is the small amount of static ram that stores BIOS settings, and the CMOS battery that you replaced provides the power to enable it to store the settings long term. The settings are lost after a period with no power, but not straight away (hence how you can remove the battery to replace it, but leave it out overnight without losing settings). This is different from the main RAM that you you are describing as being 133 MHz, which only holds data for short bursts while the main power is switched on.

I hope I haven't just made it sound more complicated than it is, I'm not as good a "teacher" as I once was 😀 I just don't want you wasting hours running memtest when that test the wrong component 😀

Ohhhh, I see. Makes perfect sense. Thanks for clarifying. 😄

DaveDDS wrote on 2025-01-30, 03:06:

A decent "multimeter" is definitely worth having when you are trying to figure out pretty much anything electrical. You can pay […]
Show full quote

DustyShinigami wrote on 2025-01-30, 01:38:

Hi Dave. Thanks for the suggestions. Unfortunately, I don’t have a voltmeter. Yet. It’s something I’ll certainly have to look into for these situations.

I believe the battery is a standard 3V one, just like the one I took out. The RAM is different though, from what I was using before, but it’s 133 MHz instead of 100. The system has set it to 100 as that’s all it supports. Though could that be an issue…? Maybe not the core memory clock, but the RAM itself. I might have to run some memory test.

A decent "multimeter" is definitely worth having when you are trying to figure out pretty much anything electrical.
You can pay a lot for a high-end one, but a simple basic one is quite cheap and all you need most of the time
(and infinitely better than none)

A lot of (most?) BIOS will auto-detect IDE/SATA drives ... the on-board firmware of those has an ID command
where it will report the type and configuration of the drive) - rarely in the last 20+ years have I had to manually
set up a drive type.

The CMOS RAM is not the system RAM - it's a little tiny chip often inside one of the "system chips", mounted
directly to the main board, or sometimes in an IC socket. Never in a "slot" socket like system RAM. It's capacity
is only a few bytes - enough to hold the system settings, but nowhere near enough to run anything out of.

The most common reasons I've seen when CMOS batteries are replaced and immediately fail are:
- Took an old battery from junkbox and it was dead or near dead.
- "Clips" in battery holder have some corrosion or other "junk" on them. Removing/replacing the
battery moves it just enough for this to interfere with conduction. Always a good idea to give these
clips (and the battery itself) a good cleaning when re-inserting.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

I see. If it does it again (and each day for the past 2-3 days now, something seems to go wrong), I’ll look at giving it a clean. I should be fine using a q-tip and some cleaning alcohol, right? The stuff I have is 99.9% so it’ll evaporate quickly. 😀

I don’t suppose those multimeters check a component’s health as well?

DustyShinigami wrote on 2025-01-30, 03:16:

I see. If it does it again (and each day for the past 2-3 days now, something seems to go wrong), I’ll look at giving it a clean. I should be fine using a q-tip and some cleaning alcohol, right? The stuff I have is 99.9% so it’ll evaporate quickly. 😀

I don’t suppose those multimeters check a component’s health as well?

Qtip/alc should be OK - I tend to be a bit more brutal and use some fine emery paper to "sand"
surfaces clean....

A basic multimeter will usually let you measure:Voltage, Current and Resistance.

If you know how a circuit works, you can often figure out which components are likely failed with these readings,
but basic meters don't have detailed component diagnostics (other that resistors which you can easily check/confirm
with the resistance measurement function - you can do a pretty good job on diodes with most
meters as well).

As you go higher-end, you can get frequency, capacitance, inductance, transistor testing and other features, but these
last few are almost always out-of circuit... You can usually do a much better job in-circuit if you know how the circuit works
and what you "should" be seeing.

One of my favorite of my own multimeters is a "scopemeter" which has a oscilloscope built in. This lets
you see how a signal is changing in real-time. This one is just a single-channel and doesn't hold a candle
to my 4-channel bench scope ... but it's very portable and battery powered .. so still very handy at times!

(You won't be ready for a scope yet, but if you get more into fixing/building electronic stuff you may find
yourself wanting one - at that point, contact me - I found a very decent and cheap USB scope - most USB
ones are total crap, relying on your PC to be fast (and consistent) enough to take real-time measurements
- but I found one that's a "real" scope, using the PC only as a display - night and day differance - OK
enough of me rambling on about test equipment).

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

DaveDDS wrote on 2025-01-30, 04:15:

Qtip/alc should be OK - I tend to be a bit more brutal and use some fine emery paper to "sand" surfaces clean.... […]
Show full quote

DustyShinigami wrote on 2025-01-30, 03:16:

I see. If it does it again (and each day for the past 2-3 days now, something seems to go wrong), I’ll look at giving it a clean. I should be fine using a q-tip and some cleaning alcohol, right? The stuff I have is 99.9% so it’ll evaporate quickly. 😀

I don’t suppose those multimeters check a component’s health as well?

Qtip/alc should be OK - I tend to be a bit more brutal and use some fine emery paper to "sand"
surfaces clean....

A basic multimeter will usually let you measure:Voltage, Current and Resistance.

If you know how a circuit works, you can often figure out which components are likely failed with these readings,
but basic meters don't have detailed component diagnostics (other that resistors which you can easily check/confirm
with the resistance measurement function - you can do a pretty good job on diodes with most
meters as well).

As you go higher-end, you can get frequency, capacitance, inductance, transistor testing and other features, but these
last few are almost always out-of circuit... You can usually do a much better job in-circuit if you know how the circuit works
and what you "should" be seeing.

One of my favorite of my own multimeters is a "scopemeter" which has a oscilloscope built in. This lets
you see how a signal is changing in real-time. This one is just a single-channel and doesn't hold a candle
to my 4-channel bench scope ... but it's very portable and battery powered .. so still very handy at times!

(You won't be ready for a scope yet, but if you get more into fixing/building electronic stuff you may find
yourself wanting one - at that point, contact me - I found a very decent and cheap USB scope - most USB
ones are total crap, relying on your PC to be fast (and consistent) enough to take real-time measurements
- but I found one that's a "real" scope, using the PC only as a display - night and day differance - OK
enough of me rambling on about test equipment).

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

I did see this one on Amazon - https://www.amazon.co.uk/AstroAI-Digital-Mult … tag=snxgb313-21
Very basic. I don't think I'll be going down the rabbit-hole with any fancy ones, to be honest. ^^; I've still yet to get my head around the basic functions and readings 🤣.

I'll report back later if the issue resurfaces. And if it doesn't, I'll continue to monitor it over the next few days. Hopefully the issue is resolved. Maybe.

DustyShinigami wrote on 2025-01-30, 13:28:

I did see this one on Amazon - https://www.amazon.co.uk/AstroAI-Digital-Mult … tag=snxgb313-21
Very basic. I don't think I'll be going down the rabbit-hole with any fancy ones, to be honest. ^^; I've still yet to get my head around the basic functions and readings 🤣.

Looks reasonable for the price. I don't know that brand, but most of the cheap one are good enough for basic
testing...

I did see on the specs. that it uses a "Lithium Metal" battery - depending on the exact type it could be hard to find/
pricey to replace - but you shouldn't need to replace very often.

I didn't see what the LCD is, but from the photos it looks to be a 3-1/2 digit (1/2 means the extra lead digit can only
show '1') - the most common type, you'll notice that most of the ranges that can be selected are 2...

The meter will be able to display -1999 to 1999 (insert the decimal point for the scale position)
for example, the 20v range will measure from -19.99v to +19.99v

It has limited AC capability ... just 200v and 600v (good for us and european AC) - would be nice if it had AC amps,
which would be handy measuring line current draw - but that won't be essential for most of what you will likely want
it for.

A couple of "extras" you may find useful:

It has a 10a scale - many little cheap ones can't go that high, and 10a will be enough to check the power draw of most system
rails (high-end gaming rigs with video may draw more)

It has a 200ua scale, which will let you measure very low currents -199.9 to 199.9 microamps.
This means you should be able to see draw of 0.1-0.2 microamps. (induced currents in lines etc. often cause the
lowest digit to bobble a bit in very low ranges... but you should be able to tell if it's bobbling more to one side
that true no-current (hold the leads together) - this could be useful when trying to see if there's much "power
off" draw on something like a CMOS RAM backup battery!

The fact that it has a diode test setting means it can put out enough voltage to cause a diode to conduct
in forward direction (typically >0.7v) the ability to test diodes can be very useful when finding problems in
power supplies.

The Continuity setting means it will make a noise if the leads are connected in that setting.
Very useful when you are tracing circuit connections - not having to look at the meter to see if
there's a connection.

Hopefully the above isn't a bunch of "boring stuff you already know" - I get the impression that you may not have
had/used such a device before and want to let you know a couple things it might do that may not be obvious.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Sorry for the delayed response. Depression.

I did boot it up again yesterday and everything has remained fine still. So fingers crossed the problem is sorted. But if it does do it again, I'll have to use my compressed air machine and some cleaning alcohol.

I didn't even notice the battery for that multimeter. I'm not sure what kind it needs exactly. Do they often need replacing? Or can they just be recharged? And thanks for trying to explain those points. Rest assured, I didn't/don't know any of that already. If anything, I'm confused. :p It's all new to me, so I don't really understand it. ^^;

Also, are those metres, and that one I linked to specifically, any good at finding out what amperture components need? Specifically the CPU? I understand Pentium 3s, like what I have, need 20+A on the 3.3V and 5V rails...? I don't think my current PSU even has those. Plus, the wattage is quite low. I'm surprised I've not had blue screens of death more often or the machine failing to boot completely.

DustyShinigami wrote on 2025-01-31, 14:45:

I did boot it up again yesterday and everything has remained fine still. So fingers crossed the problem is sorted. But if it does do it again, I'll have to use my compressed air machine and some cleaning alcohol.

Here's hoping!

I didn't even notice the battery for that multimeter. I'm not sure what kind it needs exactly. Do they often need replacing? Or can they just be recharged?

All depends on the meter ... but I've got probably a dozen or more different multimeters, and I haven't noticed any of them
draining batteries at all quickly. Chances are it will go many months!

Also, are those metres, and that one I linked to specifically, any good at finding out what amperture components need?
Specifically the CPU? I understand Pentium 3s, like what I have, need 20+A on the 3.3V and 5V rails...? I don't think my current PSU even has those. Plus, the wattage is quite low. I'm surprised I've not had blue screens of death more often or the machine failing to boot completely.

You won't be able to test the current draw of a single chip (like a CPU) unless you build a specific test platform (this would be VERY
difficult as you would have to have the device powered and working under a load - all the while drawing it's power from separate sources
that you could measure).

You will be able to measure the system DC power consumption on any one of it's rails (from the power supply), but that meter will only
go as high as 10 amps (and usually these little meters are made so you can do a measurement at such high currents, but
may get damaged if you run that much current through it for a longer time.

And thanks for trying to explain those points. Rest assured, I didn't/don't know any of that already. If anything, I'm confused. :p It's all new to me, so I don't really understand it. ^^;

You're welcome, and rest assured - if you do get more "into" this kind of stuff, you will be exposed to and catch-on to a lot more of it fairly
quickly - but can truly be a "lifetime hobby".

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal