VOGONS


First post, by Trev-MUN

User metadata
Rank Newbie
Rank
Newbie

A few months ago, I was asking about identifying the RAM used in my Gateway 2000 P5-120 so that I could give my old rig an upgrade. I had a lot of great feedback, but there have been some, uh, interesting developments since yesterday and I could use some additional advice. Figured it would be better to start a new thread about it.

The electronics surplus store where I had a chance to buy SIMM RAM on the cheap (we're talking $1, $2 a module at most) is shutting down for good, and the 16th (tomorrow as of this post) is their final day of operation. I was able to take my computer there yesterday to try and test out RAM before buying anything. (I'd also bought a header and spare IDE cable due to my reported issues with the SATA/IDE to USB adapter getting in the way of itself when trying to hook it up to the hard drive for backing up to an external source.)

Memory Check Not OK? — From the factory, my P5-120 came with 16 MB, identified by folks here at VOGONS as two sticks of Mitsubishi MH2M325 EDO 8 MB SIMM RAM. The motherboard has four SIMM slots but it seemed that any RAM added to the two unpopulated slots went undetected. Removing the existing Mitsubishi modules WOULD be noticed, as booting with one stick would lead to 8 MB being reported. However, putting other modules I bought into a slot originally populated by one of the modules that came with the computer would cause the memory test to report 16 MB again.

Now it might just be that I happened to grab a bunch of 8 MB modules from the bins on sale. I thought I'd grabbed modules of higher capacity than that, though. One of the customers at the store had pointed me to this Github table of known chip types and modules and I used that to identify and purchase one unmarked module that had 16 HY5117404B J-60 chips (the chart says such a module is 32 MB EDO) and four Hyundai modules with stickers saying HYM536410, with two saying they're CM-60, one BM-60, and another AM-60. These Hyundai modules have eight chips on one side and four on the other, and they're all HY5117400 J-60 of some type. (The AM-60 module has HY5117400A J-60 chips, the BM-60 one has HY5117400B J-60 chips, the two CM-60 ones have HY5117400C J-60 chips.) I thought these were 16 MB modules.

Also, occasionally while testing the modules I bought, the computer would beep upon trying to power on rather than actually starting up. What could be going on there? Did I just buy a bunch of 8 MB modules despite thinking I'd found otherwise? Is it normal behavior for the motherboard? (PC Hoarder Patrol identified it as potentially being a Gateway Thor, OEM variant of the Intel Advanced / ATX) Or is something more serious happening?

Bent Out of Shape — One of the guys shopping at the store had been showing me how to install and remove SIMM ram (definitely a different experience compared to DDR) and he happened to notice that one of the unpopulated SIMM slots has a bent pin. From the way he described it, I was going to need tweezers to try and bend the pin back into shape, but he had also recommended I buy some spare SIMM sockets while I still had the chance.

For the record, I didn't use that slot for testing the SIMM RAM I'd bought.

I don't know if the pin somehow got bent during my time at the store or if it had always been like that, since the slot in question hadn't been messed with at all until that point, to my knowledge. Could it be the reason behind the memory test weirdness?

Power Button Funny Business — I also started noticing strange behavior from my computer's power button. Now it seems like I have to hold the power button until the computer reaches POST or else the machine shuts off instantly. This didn't use to happen. In fact, now it seems like pressing the power button to turn the machine off is not always a guarantee and the computer will instead reboot instead of shut off. I have NO idea what's going on there, and when pointing this behavior out to that guy who was helping me, he said these sort of ATX machines shouldn't behave that way.

Survival Kit Supplies — One last thing that's on my mind. The guy had warned me I should buy up anything I think I might need in the future because, in his words, "this is it, baby." I did wind up buying four SIMM sockets on his advice, but I'm wondering if there's any other parts I should think about getting for making repairs to the P5-120 in the future. As of this post I have one more day I can go back and get anything I need from there. Are there any other components I should purchase to have on hand? Capacitors, maybe? None of the ones in my computer have bulged or popped.

(On that note, I also have a Pentium II-era Compaq Deskpro and Dell Optiplex GX1 I got for free a year ago and I'm wondering if I should think about buying anything THEY could use before it's too late ...)

Reply 1 of 6, by majestyk

User metadata
Rank Oldbie
Rank
Oldbie
Trev-MUN wrote on 2022-07-15, 15:42:

Power Button Funny Business — I also started noticing strange behavior from my computer's power button. Now it seems like I have to hold the power button until the computer reaches POST or else the machine shuts off instantly. This didn't use to happen. In fact, now it seems like pressing the power button to turn the machine off is not always a guarantee and the computer will instead reboot instead of shut off. I have NO idea what's going on there, and when pointing this behavior out to that guy who was helping me, he said these sort of ATX machines shouldn't behave that way.

This sounds like a dying PSU and might get worse when there´s more RAM installed due to the higher power consumption.
An ageing PSU can produce the strangest errors, so it´s best to make sure the PSU is 100% o.k. before troubleshooting the other issues.

Reply 2 of 6, by Trev-MUN

User metadata
Rank Newbie
Rank
Newbie

Crap. I don't suppose a modern or fairly modern power supply would work as a replacement? I have a spare Antec Earthwatts 380W power supply that came with another computer I picked up at a Fry's Electronics liquidation which still works.

Or maybe I could look for a replacement while I'm down there. Or is this something that will require repairing the PSU itself?

Reply 4 of 6, by Trev-MUN

User metadata
Rank Newbie
Rank
Newbie

Alright, so ... I swapped the power supplies and tested the mainboard by itself and it seemed like it worked fine. I thought that was evidence that the power supply had indeed gone bad, so I went ahead and completely mounted the Antec Earthwatts supply in the computer and hooked everything up to it.

Except now that I have done so, the computer doesn't output anything to video and emits a long beep and two short beeps on POST. Which seems to be "video card memory issue" but no such issue happened when I was at the shop on Thursday, hooking up a random VGA monitor to the machine to test RAM. Also, there does now seem to be a return to requiring that I keep the power button pushed in, in order to have the computer boot up.

I'm going to put the original power supply back in the computer to see what happens.

Reply 6 of 6, by Trev-MUN

User metadata
Rank Newbie
Rank
Newbie

I'm not sure if that's the issue. I just finished swapping the original PSU back into the computer and tested things out ... no POST error beeps this time and the monitor works.

I wonder if there's something about that Earthwatts PSU that provides too much juice for the video card to work?

Also, the CMOS battery has long since gone kaput, but I noticed when booting to Windows that there was an additional message in BIOS that "memory size decreased." Yet in Windows, the machine still sees the 16 MB of RAM that's reported upon booting.

It's all so bizarre.