Trashbytes wrote on 2023-07-16, 01:59:

… crazy TBW figures of 1,200 Terrabytes …

Remember:
- 1 kilobyte: 1000 bytes
- 1 kibibyte: 1024 bytes
- 1 megabyte: 1000 kilobytes
- 1 mebibyte: 1024 kibibytes
- 1 gigabyte: 1000 megabytes
- 1 gibibyte: 1024 mebibytes
- 1 terabyte: 1000 gigabytes
- 1 tebibyte: 1024 gibibytes
- 1 terrabyte: 1 byte, from planet earth

SCNR 😀

OK, OK, I'm gonna use TRIM from R. Loew, you've convinced me.

BTT - Just ordered 4x4MB 30-pin SIMM, a 486DX2-66 and CL GD5429 ISA VGA card for my 4386-VC-HD motherboard. I'll try to get as much out of that board as is technically possible 😁

Right now I'm configuring MS-DOS mode on my WIn98SE machine in order for it to run X-Wing CD.

rasz_pl wrote on 2023-07-16, 04:13:

For a 1TB drive, so a modern fancy drive with cache and best NAND gets 1200 cycle warranty from manufacturer nowadays compared t […]
Show full quote

Trashbytes wrote on 2023-07-16, 01:59:

Unless you are buying cheap nasty QLC drives, most modern drives are using MLC or TLC with massive SDram caches with crazy TBW figures of 1,200 Terrabytes*

For a 1TB drive, so a modern fancy drive with cache and best NAND gets 1200 cycle warranty from manufacturer nowadays compared to 10K for the worst NAND flash in 2008.

Trashbytes wrote on 2023-07-16, 01:59:

The drive will happily take care of itself for many years so long as you leave some free space for garbage collection and provisioning/leveling, I think 25% of the drive capacity is fine but 50% wouldn't be weird either due to partition limits.

and the only way to let drive know that some space is free without trimming are hacks I described above (zeroing/secure erase, smaller partition, hpa).

Never knew a smaller partition is a hack for most retro machines that's a limitation of the BIOS itself.
I 'll stand by what I wrote .. the whole SSD Trim thing is overblown FUD for a retro rig that isn't under 24/7 daily use. Many SSD drives functioned perfectly fine before the advent of Windows 7 and Trim and many of the very same drives still function just fine today. (Unless its an old OCZ drive ...their Nand wasn't the best)

OSkar000 wrote on 2023-07-08, 22:19:

I forgot to answer this post a while ago. […]
Show full quote

80386SX wrote on 2023-06-13, 23:20:

OSkar000 wrote on 2023-06-08, 20:55:

Tried to get some more life signs from my HP Vectra XU 5/90 today.. not much to report unfortunately. […]
Show full quote

Tried to get some more life signs from my HP Vectra XU 5/90 today.. not much to report unfortunately.

received_790597562572372.jpeg
Signs of life!

I would say its orange and yellow. Not mentioned in the manuals that I have found so far.

The cpu gets warm
Chips on the motherboard gets warm
Keyboard blinks at start
Soft power on/off works most of the time.

IMG20230608223838.jpg

I have tested with one and two known good Pentium 133 and two sticks of FPM memory, also tested and working.

Any ideas 😀

Maybe check the internal of the power supply (capacitors, etc).
Keep the move, HP Vectra XU 5/90 it's a very nice machine.

I forgot to answer this post a while ago.

The PSU looks fine and gives correct voltages under load. But it is probably a good idea to spend some more time and check that everything is ok.

I did put a post-card in it and it doesnt show anything. The cpu and other chips gets warm but no other good signs of life.

Good voltages do not necessarily mean that the capacitors are ok 😀
It's worth spending time on this rare machine, keep us posted 😉

80386SX wrote on 2023-07-16, 13:25:

OSkar000 wrote on 2023-07-08, 22:19:

I forgot to answer this post a while ago. […]
Show full quote

80386SX wrote on 2023-06-13, 23:20:

Maybe check the internal of the power supply (capacitors, etc).
Keep the move, HP Vectra XU 5/90 it's a very nice machine.

I forgot to answer this post a while ago.

The PSU looks fine and gives correct voltages under load. But it is probably a good idea to spend some more time and check that everything is ok.

I did put a post-card in it and it doesnt show anything. The cpu and other chips gets warm but no other good signs of life.

Good voltages do not necessarily mean that the capacitors are ok 😀
It's worth spending time on this rare machine, keep us posted 😉

As soon as I have some time to spend on it I will create a thread here 😀

rasz_pl wrote on 2023-07-16, 04:13:

and the only way to let drive know that some space is free without trimming are hacks I described above (zeroing/secure erase, smaller partition, hpa).

For the record, I can confirm cheap Sandisk drives don't account the HPA zone for trimming provision. Firmware Bug? I don't know.
Ironically, them actually account unpartitioned space as trimming provisioning just fine (it has to be in the end of the drive to work though). Tested in 2 cheap 256GB Sandisk drives from different manufacturing dates/batches.

Some fancy drives don't need end user provision, as them already have 1 or 2 extra NAND chips marked in firmware for such purpose. Check in the datasheets for "Factory Provisioned" feature.

rasz_pl wrote on 2023-07-15, 20:42:

lti wrote on 2023-07-15, 19:06:

(I've filled it a few times with video captures - down to 10GB free out of 1TB). I thought the "leave unpartitioned space" thing was just paranoia

It is paranoia if you have trim working and some free space left.

I didn't intentionally fill the drive (more than once), but that's an NVMe drive that won't go in an old computer. It will have an OS installed with TRIM support. I just looked that drive up today, and write endurance is only 400TB.

For my old stuff, I have one SSD on an IDE to M.2 SATA converter. It's a low-end ADATA SU650 120GB drive with reject-grade chips (SpecTek), but it still has a write endurance rating of 70TB. I was a little surprised. I never really paid attention to write endurance. ADATA even has an industrial model (IM2S3168) with Kioxia chips and lower capacity options (down to 32GB) on the same PCB.

Today I modified the Dallas battery of my recently acquired intel batman's revenge motherboard. I desoldered the dallas, and replaced it with a socket. then I used a multitool to reveal the battery points in the dallas and soldered on the CR2032 holder. Placed the modified Dallas on the socket and connected the motherboard again. After some nailbiting the releaving post beep was there. and, succes. The cmos battery empty message was gone and the bios settings were saved again.

Another socket 4 motherboard saved 😀

Trashbytes wrote on 2023-07-15, 08:39:

....
I love the cheap Intel S series SSDs, they are damn reliable too and Intel does provide a software toolkit for these drives that includes Trim.

Any links to this one ?
And does it work with Win7 I wonder ?

win7 doesnt need any tools, trims on its own

My retro activity for the day has been arguing with an eBay seller. He is of the opinion that a motherboard with non-functioning ps/2 and COM ports and a barely functioning keyboard counts as fully functional.

Worked on the 486

* Fixed power supply fan (oiled it a little, no more brwrwrwrwrwrwrwrw noise and no more failure to start to spin... so far)
* Lubricated the 3 1/2 floppy drive (no more weird noise between track ~20 & 60)
* Tried to fix weird grinding noise in my 5 1/4 Teac FD55 (failed to fix, no idea where that come from. noise come up gradually after about 30 seconds of "latched" spinning. No weird noise if it's spinning when drive is not latched. I could not fully dismantle that drive yet, need more research about how it's made. Look like the previous hodler did use some strong glue on the latch so I was not able to remove it (I want to retro bright it too, which require removal or all the plastic parts)
* Fixed CDROM ejection mechanism. I guess it just needed some lubrication... Now it eject every time I press the button if the drive is positioned flat. Still won't eject if it's placed sideway but it's a non-issue for me because it's not placed sideway in the 486 case.

Rav wrote on 2023-07-16, 21:51:

* Lubricated the 3 1/2 floppy drive (no more weird noise between track ~20 & 60)

It's very rare for any floppy drive to require lubricating the head sled rails. These usually are meant to run dry (bronze bushings move over hardened, polished steel rods), but must be clean. There's very little slack here and even thin nicotine residue for example can seize them up. Lubricating the rails will work, for a time, until the lubricant attracts dirt and causes the bushings to wear. It's your drive but be aware what you are doing is a quick fix and not a lasting repair.

Rav wrote on 2023-07-16, 21:51:

* Tried to fix weird grinding noise in my 5 1/4 Teac FD55

The rotating plastic spindle on the clamping arm - this might actually require lubrication, but it needs to be a proper one. Thin oils will not work, I had some success with graphite grease (note, it's a very dirty stuff). In general FD55 are easy to service but the one thing that can be annoying is the clamping arm. It's a bit flimsy IMO, can get bent out of shape a bit over time. It might help to slack the 2 screws that hold the spring at the back, clamp it properly and re-tighent the screw.

EDIT: Forgot to explain that the plastic part needs to be disassembled. The thing to lubricate is the metal shaft but not at the point where it connects to the clamping arm, but between the shaft and the spindle, or even not there but the flat surfaces of the spindle - it's plastic but usually with a metal insert, it's that insert sides vs the washers held by a spring that cause noise. But lubricating the shaft a bit is probably a good idea as well.

luckybob wrote on 2023-07-13, 13:57:

Veeb0rg wrote on 2023-07-13, 03:57:

A while back I found a chipset heatsink off a server that was being scrapped that I thought would make a great 486 heatsink. I had some time so I sat down and modified a 3d printed heatsink retainer to fit how I needed. Turned out pretty well.

OOO!

Are you willing to share that 3d printed part file? I'm tired of using thermal tape on 486 chips.

https://www.thingiverse.com/thing:6120080

Been restoring a few of the items from yesterday's haul.

On a side note: anybody know how to fix a Latitude C610 that WILL NOT read any hard drive you install in it?

PcBytes wrote on 2023-07-17, 11:27:

On a side note: anybody know how to fix a Latitude C610 that WILL NOT read any hard drive you install in it?

Once I mounted a HDD upside down (or backwards, can't remember) in its bracket. It closed in the chassis alright, but was not detected of course.
Does the original drive work? Not some kind of nasty firmwared disk you need?

Nice cards you got there by the way. Fairly quiet I think?

Original drive works (tested on Gericom-ECS G730) although it's wiped clean. As for mounting, it only goes one way due to the case bracket installed on the end opposite of the IDE pins.

For the cards, only the 9800GT is loud - the fan header has no control due to the card being originally designed for a silent passive heatsink.

I guess that activating an XP installation via Microsoft's automated dial-up service counts as a retro activity? An oscilloscope running XP at work all of the sudden was able to power up again, just to require Windows to be activated. Pretty cool that it actually worked, though ... 😎

Wow, I'm surprised that XP can still be activated by any method 😀

debs3759 wrote on 2023-07-18, 15:30:

Wow, I'm surprised that XP can still be activated by any method 😀

To be honest - as was I ... Wonder if it's some embedded version, that makes it "activatable" I was also kind of surprised it was an automated thing: 🤣: