How do you know that those are fake? Did you open them and test?

Not saying you are wrong, I am just interested in knowing how you determined they are fake.

cyclone3d wrote on 2021-06-02, 08:15:

How do you know that those are fake? Did you open them and test?

Not saying you are wrong, I am just interested in knowing how you determined they are fake.

Yes, the guy I paid for tested all 4 of them and when installed the board didn't boot, while once those were took off the board worked like a charme.
I needed just 1, I bought 4 because I knew there could be defective parts, but those are all fakes.
Also as stated I doubt they get out from the factory wrapped with tape and paper, that's not how the original stuff is sent out.

Contact UTSource and settle this issue. I think they bought from chinese sellers without checking the sources is the mistake.

Paper and cardboard is good ESD material, but tapes on components are not ESD. If sandwiched in paper or cardboard then bundle taped, that's acceptable.

Cheers,

Am I right in assuming this is about the cache upgrade on that DFI motherboard?
If so what did you do apart from installing the second cache chip?

Doornkaat wrote on 2021-06-02, 18:03:

Am I right in assuming this is about the cache upgrade on that DFI motherboard?
If so what did you do apart from installing the second cache chip?

Asus motherboard, it was sold as with 512kb of cache and 1024, no other things are needed since the tag ram is already populated.

Nemo1985 wrote on 2021-06-02, 18:15:

Doornkaat wrote on 2021-06-02, 18:03:

Am I right in assuming this is about the cache upgrade on that DFI motherboard?
If so what did you do apart from installing the second cache chip?

Asus motherboard, it was sold as with 512kb of cache and 1024, no other things are needed since the tag ram is already populated.

Which one?

Sorry if you don't want to discuss this. I'll stop then. I just know that on the DFI board some resistors need to be reconfigured as well. Maybe there's a similar problem here and an easy fix.

Doornkaat wrote on 2021-06-02, 19:17:

Nemo1985 wrote on 2021-06-02, 18:15:

Doornkaat wrote on 2021-06-02, 18:03:

Am I right in assuming this is about the cache upgrade on that DFI motherboard?
If so what did you do apart from installing the second cache chip?

Asus motherboard, it was sold as with 512kb of cache and 1024, no other things are needed since the tag ram is already populated.

Which one?

Sorry if you don't want to discuss this. I'll stop then. I just know that on the DFI board some resistors need to be reconfigured as well. Maybe there's a similar problem here and an easy fix.

No problem to discuss about it, it's the Asus asus sp98agp-x.

Utsource is flooded with Chinese rebranded refurbished and fake stuff same as all the other available sources for old chips. I wouldn't expect much trouble getting a refund.

firage wrote on 2021-06-02, 19:29:

Utsource is flooded with Chinese rebranded refurbished and fake stuff same as all the other available sources for old chips. I wouldn't expect much trouble getting a refund.

How actually? They refuse to complain despite the warranty.

Nemo1985 wrote on 2021-06-02, 19:24:

No problem to discuss about it, it's the Asus asus sp98agp-x.

Alright! 😃
All you did was solder on the chip, right?
My reasoning is if they just sent you plastic with legs the board would probably post normally just with the same amount of cache as before; electrically nothing changed.
So either you got chips that are defective or just too slow for the board,
or you got completely different chips,
or they are the real deal but something else is wrong.

I looked up some pictures of the board. I assume you populated the pads directly next to the chip with the correct capacitors? Without them the chip will most likely not work right.
Also I think you need to move the resistor from R24 (one cache chip) to R23 (two cache chips).

I don't know wether this is all the difference between the 512k and the 1024k variant and I can certainly not verify wether my observations are correct at all but I believe it's still worth a shot.
If you're going to go for it I'm rooting for you! 😃✊

Doornkaat wrote on 2021-06-02, 22:31:

My reasoning is if they just sent you plastic with legs the board would probably post normally just with the same amount of cache as before; electrically nothing changed.

This.

Also, in my experience even broken cache chips won't prevent the board from booting unless there's a short between Vcc and GND, which you can easily check. All the blown cache chips I've encountered either weren't detected at all (as if they were not there) or produced instability after POST when the L2 is enabled.

Of course, if these chips have a different pinout from what the motherboard expects, a short can occur. So make sure you verify that.

I didn't solder them my self, I wouldn't have the knowledge. But since we are gone to guessing.
There isn't any empty spot for others capacitors, the board is fully popoluted, the only thing that are missing are 2 tiny resistors around the second cache chips.
The pin layout is the same (just checked to be sure) but there are standards. Also the chips I bought are 5ns while the installed one is 6ns, for sure that is not a reason for incompatibility.
When Tiido tried to use fake cache chips on my motherboard (not that one) it just stopped booting, so things are different between 486 mb (where when you select the wrong cache amount it boots and then hangs before loading the OS) and those socket7, it came back to life when those fake chips were desoldered.

If those fake chips were just plastic packages with legs, then them being soldered on the board or not would not have made any difference whatsoever unless the legs are all connected... aka, a metal plate with legs sandwiched in between two layers of plastic.

cyclone3d wrote on 2021-06-03, 05:18:

If those fake chips were just plastic packages with legs, then them being soldered on the board or not would not have made any difference whatsoever unless the legs are all connected... aka, a metal plate with legs sandwiched in between two layers of plastic.

Should I try to open one of the chips? 😀

Nemo1985 wrote on 2021-06-03, 04:24:

I didn't solder them my self, I wouldn't have the knowledge. But since we are gone to guessing.
There isn't any empty spot for others capacitors, the board is fully popoluted, the only thing that are missing are 2 tiny resistors around the second cache chips.

If you mean the solder spots marked blue in the attached picture, i am pretty sure those are for capacitors. The values are the same as on the capacitors next to the other chip.

The pin layout is the same (just checked to be sure) but there are standards. Also the chips I bought are 5ns while the installed one is 6ns, for sure that is not a reason for incompatibility.

I agree.

When Tiido tried to use fake cache chips on my motherboard (not that one) it just stopped booting, so things are different between 486 mb (where when you select the wrong cache amount it boots and then hangs before loading the OS) and those socket7, it came back to life when those fake chips were desoldered.

So what exactly did the board do? Did it turn on (fans spinning, drives powering up) but no beeps and no picture? Did it POST (count RAM, list drives etc.) but not boot the OS? Or was there no sign of life at all?

You did not adress the possible issue with not moving the resistor (marked green). From pictures online it appears the resistor is installed at R24 when only one cache chip is installed and in R23 with two cache chips.

Again, I am not saying those chips are certainly real and working. How would I know?
I'm just trying to explore the possibility of the problem being the result of an oversight in installing them.

Btw. since this was the original question it is my experience that PayPal sides with the buyer in 99% of cases. Usually you'll be required to return the item at your won expense though.

Doornkaat wrote on 2021-06-03, 06:21:

If you mean the solder spots marked blue in the attached picture, i am pretty sure those are for capacitors. The values are the same as on the capacitors next to the other chip.

Wait, what pictures? I didn't see any attachment other than the chips picture I uploaded on the main post.

Doornkaat wrote on 2021-06-03, 06:21:

When Tiido tried to use fake cache chips on my motherboard (not that one) it just stopped booting, so things are different betwe […]
Show full quote

When Tiido tried to use fake cache chips on my motherboard (not that one) it just stopped booting, so things are different between 486 mb (where when you select the wrong cache amount it boots and then hangs before loading the OS) and those socket7, it came back to life when those fake chips were desoldered.

So what exactly did the board do? Did it turn on (fans spinning, drives powering up) but no beeps and no picture? Did it POST (count RAM, list drives etc.) but not boot the OS? Or was there no sign of life at all?

You did not adress the possible issue with not moving the resistor (marked green). From pictures online it appears the resistor is installed at R24 when only one cache chip is installed and in R23 with two cache chips.

Again, I am not saying those chips are certainly real and working. How would I know?
I'm just trying to explore the possibility of the problem being the result of an oversight in installing them.

Btw. since this was the original question it is my experience that PayPal sides with the buyer in 99% of cases. Usually you'll be required to return the item at your won expense though.

This motherboard specifically started to turns on as soon the atx power connector was plugged, that was already worrying.
Then it showed a black screen with fan spinning and so on.
Unlucky I can't remember what the debug card said.

Can you please reupload the pictures? They sure help me to clarify.
Thanks

Sorry, I edited the post on my phone and that must mave somehow messed things up.
Reuploaded from PC. 😁

Doornkaat wrote on 2021-06-03, 07:16:

Sorry, I edited the post on my phone and that must mave somehow messed things up.
Reuploaded from PC. 😁

I was referring exactly to those things, thank you for the clarification. I suppose it's worth a try but i'm not confident.
I should move the resistor from r23 to r24 and then add two caps on the pads near the cache chips.
Do you know how can I find the value for the caps?

Nemo1985 wrote on 2021-06-03, 08:12:

I was referring exactly to those things, thank you for the clarification. I suppose it's worth a try but i'm not confident. I sh […]
Show full quote

Doornkaat wrote on 2021-06-03, 07:16:

Sorry, I edited the post on my phone and that must mave somehow messed things up.
Reuploaded from PC. 😁

I was referring exactly to those things, thank you for the clarification. I suppose it's worth a try but i'm not confident.
I should move the resistor from r23 to r24 and then add two caps on the pads near the cache chips.
Do you know how can I find the value for the caps?

Those are caps that provide a local "energy pool" for the chips in case they need a sudden spike of energy, and delivering it from somewhere else in the board wouldn't be fast enough. So the exact value of the caps doesn't matter that much. It's important that these caps have a low resistance, so they can deliver the energy with minimal "friction". Multilayer ceramic capacitors (MLCC for short) do a great job at this, so industry convention is to use 100nF MLCC capacitors. The typical size of SMD components on PC mainboards is called "0603".