VOGONS


Keep GeForce 4 to cards from dying

Topic actions

First post, by candle_86

User metadata
Rank l33t
Rank
l33t

So in the last 6 months I've had 5 Ti's fail.

2x Ti-4200 64mb
1xTi-4200 128 mb
1x Ti-4200 8x
1x Ti-4600

I've swapped boards and psus, but the deaths are all the same, after a few weeks I start getting artifacts and continue to do so until the card is unusable.

Athlon XP 2800 based hp a430n and I've swapped boards to one with new caps, and swapped psus to a dell 350W that checks out on a psu tester. What could do this, I've got another ti-4600 that I'm terrified to try. Ironically for the last 4 weeks it's been running a Radeon 8500le with zero issues

Last edited by Stiletto on 2021-01-31, 23:24. Edited 1 time in total.

Reply 2 of 71, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie

How are you storing these cards? ESD could be one of the reasons why they all die...
I store ALL my spare components in high quality antistatic bags (bought hundreds of them, of all sizes, when I first got into this hobby), in a clean and dry closet.
I also try my best to only use good quality PSUs and I ALWAYS replace their capacitors with Nichicon Low-ESR.
All in all, I only had a few components die while in storage (Voodoo cards). Had no issues with other devices/manufacturers.

1 x PLCC-68 / 2 x PGA132 / 5 x Skt 3 / 9 x Skt 7 / 12 x SS7 / 1 x Skt 8 / 14 x Slot 1 / 5 x Slot A
5 x Skt 370 / 8 x Skt A / 2 x Skt 478 / 2 x Skt 754 / 3 x Skt 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k

Reply 3 of 71, by mockingbird

User metadata
Rank Oldbie
Rank
Oldbie

A PSU's voltages may check out ok with a PSU tester, but it could be passing hundreds of millivolts of AC ripple to system components.

Do not use old PSUs unless they have been re-capped.

mslrlv.png
(Decommissioned:)
7ivtic.png

Reply 4 of 71, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
bloodem wrote on 2021-01-31, 18:24:
How are you storing these cards? ESD could be one of the reasons why they all die... I store ALL my spare components in high qua […]
Show full quote

How are you storing these cards? ESD could be one of the reasons why they all die...
I store ALL my spare components in high quality antistatic bags (bought hundreds of them, of all sizes, when I first got into this hobby), in a clean and dry closet.
I also try my best to only use good quality PSUs and I ALWAYS replace their capacitors with Nichicon Low-ESR.
All in all, I only had a few components die while in storage (Voodoo cards). Had no issues with other devices/manufacturers.

I was also going to mention ESD.
And I also bought hundreds of brand new antistatic bags, it's one of the best purchases a person in this hobby can make 😁

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 5 of 71, by framebuffer

User metadata
Rank Member
Rank
Member

Uhm but he said that the card(s) died after few weeks of usage, how this can be due to ESD? I mean, unless he touched them while electrostatically charged when they were running, but I guess he didn't

Windows 98 and SAMBA | Quake CPU Benchmarks | GeForce2: GTS vs MX400

Reply 6 of 71, by tannerstevo

User metadata
Rank Member
Rank
Member
framebuffer wrote on 2021-02-15, 21:45:

Uhm but he said that the card(s) died after few weeks of usage, how this can be due to ESD? I mean, unless he touched them while electrostatically charged when they were running, but I guess he didn't

Well not related, but I had a usb port die a few months after I plugged in a thumb drive.
There was a pop and a spark so yeah.
ESD can kill even after a few months.

Reply 9 of 71, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie

At the risk of stating the obvious, removing the heatsink and cleaning the old and applying new thermal paste is always a good move. When I acquire a new to me vintage card I generally makes sure it works, then both do this, and clean and lubricate the fan.

"I'd rather be rich than stupid" - Jack Handey

Reply 10 of 71, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie
framebuffer wrote on 2021-02-15, 23:07:

OK interesting, so you mean if you discharge on a component you can "wear it" and reduce it's life drastically to days?

Yes. ESD damage is often subtle and that can make cause and effect difficult to establish. That's why some people remain skeptical that it happens at all. And when they've been careless enough to cause a complete component failure, they assume it was a faulty part when they received it. https://www.rs-online.com/designspark/what-is … es-can-it-cause

"I'd rather be rich than stupid" - Jack Handey

Reply 11 of 71, by Velociraptor

User metadata
Rank Member
Rank
Member

This problem does scare me. It's made me set things up with less expensive cards so that they could be replaced.

I do change the paste, and put on huge aftermarket coolers, and I do run them as cold as I can. But I'm concerned that if prices keep going up and the cards keep dying then it won't be possible to have them all the time.

Memory chips on older cards rarely have heatsinks on them, so one of the tasks I'll need to do is fit some kind of heatsinks to them as well.

I'm very interested to hear if Candle is just simply unlucky or if it's a common thing and all these early 00s cards are doomed.

Reply 12 of 71, by auron

User metadata
Rank Oldbie
Rank
Oldbie
Repo Man11 wrote on 2021-02-16, 00:38:

At the risk of stating the obvious, removing the heatsink and cleaning the old and applying new thermal paste is always a good move. When I acquire a new to me vintage card I generally makes sure it works, then both do this, and clean and lubricate the fan.

it's not always a good move; in fact when doing it results in prying off the heatsink with force and damaging the BGA, it's not a good move at all. there's been posts with examples of this here and in general it has probably happened more than people care to admit, because this TIM replacement thing is stuck as "common sense" now, as indicated by your post.

i've seen a method shown on video where the removal is done with dental floss, and that seems to me like the least stressful way for the BGA, but i doubt this is how most people go about this. actually i have my own doubts about this being necessary for every kind of old TIM; i have an original xbox where i didn't do this and the CPU temps it shows are absolutely fine, around 50°C.

i'm not against replacing old TIM in general, but people should probably put more thought into it than blanket statements.

Last edited by auron on 2021-02-16, 21:17. Edited 1 time in total.

Reply 14 of 71, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie
auron wrote on 2021-02-16, 01:18:

it's not always a good move; in fact when doing it results in prying off the heatsink with force and damaging the BGA, it's not a good move at all. there's been posts with examples of this here and in general it has probably happened more than people care to admit, because this TIM replacement thing is stuck as "common sense" now, as indicated by your post.

In general, there is no need to pry off the heatsink, because it’s not usually glued to the chip. The only cards I came across with glued heatsinks are Voodoo 3 (and custom Voodoo 1 / Voodoo 2), some early GeForce 256 / Creative & Visiontek GeForce 2 GTS and a Gainward GeForce FX 5900.
I also typically replace the thermal paste (and sometimes the cooler altogether) on all of my cards immediately after the initial testing, and it’s usually a pretty easy process that takes 10 minutes.
Now, when it comes to glued heatsinks, I usually don’t bother. I find that the heat transfer is still very good anyway, so I just add a fan or replace/lubricate the existing one.

1 x PLCC-68 / 2 x PGA132 / 5 x Skt 3 / 9 x Skt 7 / 12 x SS7 / 1 x Skt 8 / 14 x Slot 1 / 5 x Slot A
5 x Skt 370 / 8 x Skt A / 2 x Skt 478 / 2 x Skt 754 / 3 x Skt 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k

Reply 15 of 71, by Miphee

User metadata
Rank Oldbie
Rank
Oldbie

Not ESD, CES. Common Ebay Scam.
Those cards were all dead way before you bought them, maybe "repaired" in an oven so they work for another month.
When the seller has tested and untested stuff it always means that the "untested" ones didn't work for him.
Untested means that he knows that it's shit but it's your risk to buy it. They usually say "no warranty because..."
Instascam.
You just ran a streak of bad luck, it happens to many of us in this hobby.
Some people want to offload their junk that way so they mix tested/untested in a single pack so people will buy everything at once.
I just bought a huge pack of tested/untested and naturally the untested ones don't work at all or show signs of tinkering and bad repair jobs.
Business as usual.
ESD is not a serial-killing monster, I haven't lost anything to ESD and I haven't used an ESD bag in years. Too inconvenient not to mention it's useless without ESD gloves or flooring and proper ESD grounding. I doubt many of us have the whole package necessary to combat a phenomenon that's really a minor risk.

Last edited by Stiletto on 2021-02-16, 23:23. Edited 1 time in total.

Reply 16 of 71, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie

I strongly disagree. ESD is not a 'minor' risk - if it were, factories would not invest millions of dollars in ESD protection equipment. They could very well just use that money for R&D.... right?

Now, the truth is that ESD risk assessment is tricky and it's also very dependent on the environment, even on what clothes you're wearing. Sometimes just having the wrong carpet in a room can cause A LOT of issues.
I for one managed to kill an S3 Virge DX video card by wearing the wrong sweater (which I actually knew was dangerous). I had just turned off the PC, tried to remove the video card, the sweater touched the video card, there was a very visible spark and... when I turned it back on, the screen was filled with artifacts (one or more memory ICs were probably damaged).

Also, even worse, as others have pointed out, ESD is not necessarily immediately noticeable, so you might not even realize that at some point in time it was ESD that caused a failure that you just attributed to bad luck. Again, some environments are ESD death traps, while others are inherently much safer.
I don't know if ESD was truly the problem in the OP's case, but it's definitely something worth considering.

1 x PLCC-68 / 2 x PGA132 / 5 x Skt 3 / 9 x Skt 7 / 12 x SS7 / 1 x Skt 8 / 14 x Slot 1 / 5 x Slot A
5 x Skt 370 / 8 x Skt A / 2 x Skt 478 / 2 x Skt 754 / 3 x Skt 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k

Reply 17 of 71, by Miphee

User metadata
Rank Oldbie
Rank
Oldbie

So ESD is the secret force that can be blamed when something goes bad without an immediately visible reason.
Or it's just taken waaaaay too seriously and it's not a real threat with all the chip protections in place since the 1980's.
Again, it's all useless unless it's 100% efficient (like in big factories). Or do you all handle your equipment with ESD gloves all the time and have ESD grounding and flooring in your work area?
Because that's the only way to be sure, a bag won't protect anything after you take the hardware out of it.

Reply 18 of 71, by Tetrium

User metadata
Rank l33t++
Rank
l33t++
framebuffer wrote on 2021-02-15, 23:07:

OK interesting, so you mean if you discharge on a component you can "wear it" and reduce it's life drastically to days?

ESD damage is a scientifically proven fact. This phenomenon can damage the microscopic traces in a somewhat similar way to how electro-migration can also cause damage.
So if for example a trace is damaged by an ESD and cut by 90%, the remaining 10% may fail at a later date due to (for instance) electromigration. ESD damage may (and probably will) shorten transistor lifespan, so it's not surprising it may take a while to permanently fail.

I mean, using older TIM will probably not kill a GPU or CPU right away either. Things may take a while to die.

The ESD protection bags are relatively cheap and down the line will save you a lot of time (if only to protect the component from scratches, dust and bugs and to form a bit of an insulative barrier). At one point I got like 600 in batches of 100 and the really large ones and medium sized ones seemed to be the most useful (especially if large enough to at least hold a single harddrive. Smaller ones turned out to be less useful to me).

But I digress. This may be not the cause of OPs issues with his GF4 cards. But I've noticed quite a lot of new threads about graphics cards spontaneously giving up the ghost in the last few years.

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 19 of 71, by bloodem

User metadata
Rank Oldbie
Rank
Oldbie
Miphee wrote on 2021-02-16, 09:29:

Or do you all handle your equipment with ESD gloves all the time and have ESD grounding and flooring in your work area?

Can't speak for everyone else, but no, I don't wear gloves (I used to wear them when I was an employee in a laptop service center more than a decade ago), however in recent years I made a habbit out of always discharging myself before touching sensitive components (I have a grounded radiator in my room). It's not 100% foolproof, but it's better than nothing. I also stopped wearing clothes that are prone to static buildup 😀
And, no, ESD is not just the invisible boogeyman that you blame when there's no apparent cause of failure, but it's still better to be safe than sorry and take all the reasonable precautions. And one of those precautions is keeping ESD sensitive devices in an antistatig bag, especially when storing them for a long period of time in a dusty old closet. Yes, static buildup can also occur in a closet when certain conditions are met.

1 x PLCC-68 / 2 x PGA132 / 5 x Skt 3 / 9 x Skt 7 / 12 x SS7 / 1 x Skt 8 / 14 x Slot 1 / 5 x Slot A
5 x Skt 370 / 8 x Skt A / 2 x Skt 478 / 2 x Skt 754 / 3 x Skt 939 / 7 x LGA775 / 1 x LGA1155
Current PC: Ryzen 7 5800X3D
Backup PC: Core i7 7700k