Why did you connect the mouse while the system was on? Serial mouse is not hot-plug capable.

Lets go back to: "so I disconnected the hard drive (idk why I did that), and it worked. Then I went and connected it again, and the PC stopped responding completely."

Does this mean the computer will POST with the drive removed? I would say the drive is bad. (or cable in backwards, but since you didn't move that, rule that out). Pretty random failure after simply hot-plugging a serial mouse, if true.

dominusprog wrote on 2025-07-13, 14:54:

Why did you connect the mouse while the system was on? Serial mouse is not hot-plug.

I always thought a serial mouse could be plugged in with the PC turned on. I only ever heard about PS2 devices that could damage something; meanwhile, I never damaged anything by plugging in a serial mouse.

Major Jackyl wrote on 2025-07-13, 15:07:

Does this mean the computer will POST with the drive removed? I would say the drive is bad. (or cable in backwards, but since you didn't move that, rule that out). Pretty random failure after simply hot-plugging a serial mouse, if true.

No, it was only a one-time thing. The drive is a CF-IDE adapter with a practically new CF card. The cable can't be plugged in backwards in this motherboard, nor can it be plugged in backwards to the adapter. Also forgot to mention, I got it set up in a way where I got the CF adapter on the primary IDE and a DVD+HDD on the secondary IDE. Disconnecting the adapter made the PC boot.

I always thought a serial mouse could be plugged in with the PC turned on. I only ever heard about PS2 devices that could damage something; meanwhile, I never damaged anything by plugging in a serial mouse.

Well.. The serial port and serial mouse aren’t exactly hot plug capable, but the RS-232 port on PC has protective diodes.
Normally, the used UART FiFo is even able to cope with shorts.
So *normally*, the serial port itself is quite safe and good to tinker with.
Completely different to parallel port, which goes "poof" if you just make weird looks. 😂
But can't speak about the serial mouse, since it's not standardized. It depends on the model.

Back in the late-90's a friend of mine had an XT form factor 286 PC and he did the same as OP and resulted in the same error OP experienced. So this could definitely happen. We were to young back then to properly troubleshoot. We were around 10. I wasn't present when he connected the mouse, but after that the PC couldn't recognize the hard drive. We never quite could fix it even with replacing the hard drive controller card. but booting from floppy and loading games from there still worked.

So things like this can happen.

Jo22 wrote on 2025-07-13, 15:41:

Well.. The serial port and serial mouse aren’t exactly hot plug capable, but the RS-232 port on PC has protective diodes.
...
But can't speak about the serial mouse, since it's not standardized. It depends on the model.

Even if I did fry the diodes, the PC should still boot up, right? I could replace the diodes myself without much issue if that happened, so that's not the issue.
The mouse is a Microsoft-branded one.

RandomStranger wrote on 2025-07-13, 16:13:

Back in the late-90's a friend of mine had an XT form factor 286 PC and he did the same as OP and resulted in the same error OP experienced. So this could definitely happen. We were to young back then to properly troubleshoot. We were around 10. I wasn't present when he connected the mouse, but after that the PC couldn't recognize the hard drive. We never quite could fix it even with replacing the hard drive controller card. but booting from floppy and loading games from there still worked.

The thing is, in my case the PC doesn't boot or give any errors through the PC speaker, even though it did beep at the beginning. After some tinkering, it stopped completely. I doubt I fried the IDE, and even if I did, I'm more concerned about why the motherboard seems dead.

RandomStranger wrote on 2025-07-13, 16:13:

Back in the late-90's a friend of mine had an XT form factor 286 PC and he did the same as OP and resulted in the same error OP experienced. So this could definitely happen. We were to young back then to properly troubleshoot. We were around 10. I wasn't present when he connected the mouse, but after that the PC couldn't recognize the hard drive. We never quite could fix it even with replacing the hard drive controller card. but booting from floppy and loading games from there still worked.

So things like this can happen.

Wait a moment, was it really a serial mouse or serial port?
Because, there had been bus mice with a DE-9 connector in that time .
My 286 PC had it on-board, too and the label was "MOUSE". It wasn't short-protected, also.
This port didn't work with my Genius mouse I got, so I never used that port, but ordinary "SERIAL" port (also DE-9).

Part of RS-232 is that shorting pins has to be tolerated. And Windows 95 even advised hot-plugging a serial mouse if no mouse was detected (but not PS/2 of course).

You could try reflashing the BIOS.

Another option, as stupid as it may sound, is to just disconnect everything and take all the components apart, put everything away and try again in 1 week.

jakethompson1 wrote on 2025-07-13, 18:28:

Part of RS-232 is that shorting pins has to be tolerated. And Windows 95 even advised hot-plugging a serial mouse if no mouse was detected (but not PS/2 of course).

Yeah, that's why I do it.

devius wrote on 2025-07-13, 20:08:

You could try reflashing the BIOS.

Another option, as stupid as it may sound, is to just disconnect everything and take all the components apart, put everything away and try again in 1 week.

I would love to be able to reflash the BIOS, but I don't own any device that is capable of flashing chips. When the PC was still up and running, I wanted to reflash the BIOS on my S3 Virge, but because there is not a way to do that via software, I just swapped the chip from some other S3 card I have. While doing research I found that there is a way to reflash the chip using some network card. I even got one that should be compatible, but until I get the PC running again, there is no point since the S3 Virge was meant to be used in this PC.

It does not sound stupid at all. Every time when I run out of options when troubleshooting, I just leave the hardware and come back after a few days. It works wonderfully almost every time.

dominusprog wrote on 2025-07-13, 14:54:

Why did you connect the mouse while the system was on? Serial mouse is not hot-plug.

Partially true - the serial port (RS-232) is definitely hot-plugable (speaking as someone with a LOT of serial port
experience - I designed the NewBridge 10xx series of serial data switches and multiplexers).

So, hardware-wise, no damage should occur - but software can't be guaranteed. A super-simple OS
(like DOS) generally needs the mouse to be connected when it's driver loads. More complex multi-threading
OSs (like Winblows, Linux) etc. may detect the mouse "on the fly - but this is often not done
because lots of other "stuff" could be connected to a serial port and you don't necessarily want to go
"twiddling bits" every time anything is connected.

DOS "knows" there's a serial mouse because you are loading a mouse driver..
Other OS's often look for one at startup - some keep checking. others no...

None of which explains why the system "died" ... if it doesn't even POST, could be a
hardware problem caused by static discharge (unlikely as RS-232 ports should be pretty well
protected) - otherwise most likely software has gotten itself confused...

DaveDDS wrote on 2025-07-13, 23:06:

Partially true - the serial port (RS-232) is definitely hot-plugable (speaking as someone with a LOT of serial port experience - […]
Show full quote

dominusprog wrote on 2025-07-13, 14:54:

Why did you connect the mouse while the system was on? Serial mouse is not hot-plug.

Partially true - the serial port (RS-232) is definitely hot-plugable (speaking as someone with a LOT of serial port
experience - I designed the NewBridge 10xx series of serial data switches and multiplexers).

So, hardware-wise, no damage should occur - but software can't be guaranteed. A super-simple OS
(like DOS) generally needs the mouse to be connected when it's driver loads. More complex multi-threading
OSs (like Winblows, Linux) etc. may detect the mouse "on the fly - but this is often not done
because lots of other "stuff" could be connected to a serial port and you don't necessarily want to go
"twiddling bits" every time anything is connected.

DOS "knows" there's a serial mouse because you are loading a mouse driver..
Other OS's often look for one at startup - some keep checking. others no...

None of which explains why the system "died" ... if it doesn't even POST, could be a
hardware problem caused by static discharge (unlikely as RS-232 ports should be pretty well
protected) - otherwise most likely software has gotten itself confused...

Good point, but why would anyone take a risk when it can easily be avoided?

ksiumaxx wrote on 2025-07-13, 17:13:

Even if I did fry the diodes, the PC should still boot up, right? I could replace the diodes myself without much issue if that happened, so that's not the issue.
The mouse is a Microsoft-branded one.

Yup should still boot. Have you tried swapping parts, particularly power supply or motherboard? Based on the other things you described, the problem is upstream, cause more tests show bad behavior. Power supply can be the problem. No way plugging in something is the problem.

Seriously, take the advice on disassembling it all and leave it for a week or so before reassembly and testing. May not work, sure, but you'd be shocked at how many times it solves apparently unexplainable issues

dominusprog wrote on 2025-07-16, 16:16:

Good point, but why would anyone take a risk when it can easily be avoided?

You could also say something like "why would anyone take the risk of getting a bus when everyone knows that busses can get into
accidents". Or "why would anyone eat a meal - they could choke".

The hardware risk of hot-plugging a serial port is less than either of the above.. Like busses and eating utensils RS-232 ports were
designed to be usable "on the fly". I've personally hot-plugged thousands of serial connections with never a hardware
problem.

And if you are referring to the risk of adding a mouse to a running computer - "why would anyone risk using a computer - everyone
knows that computers sometimes crash and lose data for unexplained reasons"

Yes, some "risks" should be avoided - but I don't see hot-plugging a serial cable as one of them.

DaveDDS wrote on 2025-07-17, 16:11:

You could also say something like "why would anyone take the risk of getting a bus when everyone knows that busses can get into […]
Show full quote

dominusprog wrote on 2025-07-16, 16:16:

Good point, but why would anyone take a risk when it can easily be avoided?

You could also say something like "why would anyone take the risk of getting a bus when everyone knows that busses can get into
accidents". Or "why would anyone eat a meal - they could choke".

The hardware risk of hot-plugging a serial port is less than either of the above.. Like busses and eating utensils RS-232 ports were
designed to be usable "on the fly". I've personally hot-plugged thousands of serial connections with never a hardware
problem.

And if you are referring to the risk of adding a mouse to a running computer - "why would anyone risk using a computer - everyone
knows that computers sometimes crash and lose data for unexplained reasons"

Yes, some "risks" should be avoided - but I don't see hot-plugging a serial cable as one of them.

Something I didnt know... my experience is hot plugging and old hardware rarely is helpful and avoided at most times. I just use that on most old hardware.

I doubt hot plugging a serial mouse would kill a board.

Back in the day I had a serial cable for linking 2 PCs for gaming and an LPT cable for copying files in Norton Commander across PCs. I don't think I turned off PCs to connect them. It was a Pentium MMX.

DaveDDS wrote on 2025-07-13, 23:06:

None of which explains why the system "died" ... if it doesn't even POST, could be a
hardware problem caused by static discharge (unlikely as RS-232 ports should be pretty well
protected) - otherwise most likely software has gotten itself confused...

Exactly.

I'd venture and say O/P did "something else" to the machine without realizing it while troubleshooting the blinking cursor / no OS loading error.

In cases like that, it's best to start from the bottom with all of the basics again.

First test the board with just CPU (+heatsink), RAM, and GPU (if no onboard) only. If that doesn't allow the board to POST, then it's very likely there was a random failure with one of these components that has nothing to do with hot-plugging the serial mouse.

That said, O/P mentioned the board is a MSI -branded one.
Old MSI stuff is synonymous with bad caps. Not saying this is necessarily the issue, but always worth mentioning and considering.
And it would help to know what power supply brand O/P has. No-name cheap garbage stuff can be problematic for the smallest of reasons.

As the3dfxdude suggest, swap PSU and see if that helps.
If you have another mobo that uses the same RAM type and/or CPU, test those too (though CPU is extremely unlikely to have failed, so probably can skip testing that.)

My money is either on mobo or PSU... or O/P did something to the hardware without realizing it while troubleshooting. In such cases, going back "1 week later" at it is a sound suggestion. Read the mobo manual again and check all jumpers are correct. Inspect for bad/bulging caps. Post PSU brand and model here or alternatively, check it for bad caps at the very least... or if it's a gutless wonder, even that won't be relevant.

dominusprog wrote on 2025-07-16, 16:16:

Good point, but why would anyone take a risk when it can easily be avoided?

In some situations, it CANNOT be avoided.
With some old industrial controllers, for example, you can't just afford to restart/reboot whole machine (and possibly have shut down whole production floor equipment) just because some serial device got disconnected, dropped out due to error (hard or soft) or malfunctioned in some other way.
Indeed RS-232 was DESIGNED to be HOT-PLUGGABLE.

Just because Win9x is an unstable piece of garbage doesn't change that fact.

momaka wrote on 2025-07-18, 00:31:

In some situations, it CANNOT be avoided.
With some old industrial controllers, for example, you can't just afford to restart/reboot whole machine (and possibly have shut down whole production floor equipment) just because some serial device got disconnected, dropped out due to error (hard or soft) or malfunctioned in some other way.
Indeed RS-232 was DESIGNED to be HOT-PLUGGABLE.

Just to point out ... it's not always because of accidental disconnect!

I've worked on many industrial controllers "back in the day" which didn't have (or normally need) video displays and
keyboards - but many (most) had an RS-232 port where you could connect a serial console/debug terminal "on the fly".