I have prepared a very elaborate article for next Monday (06.11.2023) that reflects in detail on my work over the last 6 months on this seemingly never-ending topic. And I can already spoil the fact that the pieces of the puzzle that are still missing from my point of view and have probably finally been found now provide a much more complete picture for assessing the problem. First of all, I would like to thank KrisFix for a lot of scorched test objects, the manufacturer for unused samples of the headers used, CableMod for the trustful cooperation in quality control and the board partners who followed the whole thing with interest, because some damaged users had also sent their cards to me. I was able to draw some interesting conclusions from over 50 cases.
Of course, I don’t want to give everything away, but as the article is so extensive and I’m translating it manually this time, there’s no time for anything else this weekend. So it’s time to whet your appetite and bridge the gap until Monday. In the picture above you can see the so-called headers, i.e. the part that is soldered onto the graphics card. I had already asked myself months ago why nobody had taken a closer look at this, because we only ever discuss the NVIDIA adapters, the power supply connection cables and the CableMod adapter – in other words, only what you ultimately plug into the headers.
In addition to an explanation of the terminology of the clamp/plug connection used, there will also be (or have to be) insights into the manufacture of individual components such as the *********, as well as detailed investigations and measurements on the physical unit of these headers. And I was surprised myself that certain carelessness could lead to such damage. I’m going to quote from the upcoming article and I’ve peeped out a few details, because I don’t want to give too much away just yet.
The fact is that manufacturing tolerances, the material and the general quality of workmanship are a certain focus of the problem. And after investigating more than 50 defects here in the laboratory, it has become clear that it was almost always a combination of several causes that led to total failure and not just a single one. Even if the story with the incompetent users shortly before NVIDIA’s announcement of the quarterly figures naturally suited the provider well (quite by chance): It doesn’t help anyone, not even NVIDIA, because it was probably a hasty snap decision….
…I can only advise manufacturers of these connectors to adhere exactly to the specifications and to check this permanently and frequently within a series. Yes, the costs are higher, but the customer will certainly be grateful. Part of the problem certainly also lies in the ************ used, as the incoming goods must already be checked more closely by quality control. As there are many manufacturers of connectors, the quality of the products also varies greatly. While the plastic housings are all molded very correctly, the ********** and its initial quality are more problematic.
So, I’m going back to work now and you can look forward to it. I don’t have to hire an external service provider to do anything, you can do it yourself 😀
55 Antworten
Kommentar
Lade neue Kommentare
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Moderator
Urgestein
1
Urgestein
Urgestein
1
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →