A newly built device stops working because of static electricity, yet the reason remains unclear.
A newly built device stops working because of static electricity, yet the reason remains unclear.
When preparing the setup, I verify that the chassis is correctly grounded and there are no floating voltages. A multimeter proves useful in this process.
First, disconnect all USB headers from the motherboard and observe if the problem continues. Reinserting them one by one helps identify if a faulty or damaged cable is responsible. This could introduce unwanted voltage at an incorrect point, disrupting the ground connection and potentially activating safety mechanisms.
The painted chassis and plastic components can cause gaps where metal-to-metal contact should exist.
Ensure the motherboard screws are firmly attached to metal surfaces—typically small solder pads designed for this purpose.
Verify that the standoffs are made of metal.
Confirm the motherboard tray is securely bonded to the chassis, without any insulation from paint or similar materials.
If the chassis includes a removable I/O shield, check that the metal tabs align properly with the rear I/O housing.
Lastly, confirm the PSU mount has solid metal-to-metal contact with the chassis; insufficient contact might be the root cause. Many modern chassis use sleds or carriers to simplify PSU installation, which could also lead to contact problems.
I checked the Fractal Terra description and this phrase raised some worries.
Inspired by the living spaces of today’s gamers, Terra features unique design elements like robust panels cut from anodized aluminum, CNC-milled details, and solid walnut that comes from FSC certification.
From my experience, certain anodized metal surfaces can act as insulators. Even a thin coating might prevent proper electrical contact between the panels if they’re not properly bonded, potentially leading to grounding issues.
The structural layout of Terra differs from typical computer cases I’m familiar with—it’s not the conventional shape. This unconventional design could contribute to the issue.
I’m unsure how much walnut is used inside the case, possibly only a few visible pieces, but wood is still an insulator, no matter its certification status. It might just add around $20 to the cost.
To verify electrical continuity across all metal components, I’d need a professional bonding tester capable of detecting milliohms. If unavailable, a multimeter could help, though it would be less reliable.
https://www.amptec.com/products/bonding-testing/620lk/
In a last resort, I might consider swapping all parts for a standard metal case to see if the problem resolves. If it does, it would confirm the Terra is the source of the issue.
It's likely not necessary to push to that level for testing continuity, though this highlights an important aspect of anodization.
For a SFF Mini-ITX setup, the risk of a faulty cable being involved in the issue is even higher.
You should be checking for specific details when testing it. Since the multi-meter is delayed, you might want to consider returning or replacing the PSU and motherboard to rule out other issues. Another option is a case with mesh material, but that’s not ideal right now. It’s disappointing because this was a project I’ve been working on for my wife’s birthday this week—it doesn’t seem to be happening yet. Thanks for the advice, everyone. I’ll keep checking these comments and update once I have more information or replace the components.
Currently, I'm still uncertain whether the issue lies with your Terra setup or factors such as the absence of a mains earth, the motherboard, the PSU, or the presence of numerous static-generating carpets in your space.
A multimeter is useful for measuring Voltage, Current, and Resistance (along with other optional functions).
When testing for unwanted voltages on the computer's metal components, adjust the meter to the AC Volts range, usually 200V or 600V. Setting it to 200V allows you to detect any AC voltage up to that level. This should help confirm whether an elevated AC voltage might be coming from your PSU chassis (or any attached case metalwork) due to leakage from the Y-Class capacitor network, especially if your PSU isn't properly earthed.
You'll need to position the computer near a solid earth connection for accurate testing. As previously mentioned, connect one probe—regardless of color—for AC to the earth. This could be a metal water pipe, a metal gas line, or a verified earth in a wall outlet. Link the other probe to the exterior of the ATX PSU case. If you observe several tens of Volts AC, your PSU likely lacks proper earthing. You won't detect leakage from the Y-Class capacitor network if the case is correctly earthed.
Where I reside, all residential and office mains outlets are legally required to be earthed. Unless the earth wire has somehow detached behind the wall outlet or the main cable's earth connection is missing or disconnected, most desktop computers in my country can be considered safely earthed. Problems are more likely in regions where mains earths aren't mandatory for domestic use.
You don't require an earth connection for a Class II Double Insulated PSU (marked with a two-square nested symbol), as these units (such as laptop PSUs with a 2-pin figure-of-eight input) are built to operate without one. ATX PSUs with a 3-pin mains input, however, are designed to need an earth connection.
https://en.wikipedia.org/wiki/Appliance_classes
To verify continuity between the PSU case and the computer's metal parts, detach the mains plug from the back of the ATX PSU. Set the meter to the lowest Ohms setting, then firmly press one probe tip against the outside of the PSU chassis. Press the other probe tip firmly on an exposed (bare) metal surface inside the computer case. A reading close to zero (or near zero) indicates a short circuit. Most inexpensive multimeters may display several Ohms even when probes are connected, suggesting the parts aren't electrically bonded.
The milliOhmeter I referenced is meant to confirm that all metal components in your system are properly bonded to the safety earth. In a 19-inch rack configuration, this bonding is typically achieved via separate safety earth wires or copper braid straps from the rack to the equipment. Therefore, if the main earth fails, the devices remain separately earthed.
If the Meshroom S panels are constructed from metal (not plastic), they qualify as a metal case in my view. Even with numerous small ventilation openings (for dust control?), they still function as a Faraday cage, shielding internal components from static electricity. Electrical disturbances—such as conducted emissions entering or leaving through connectors—present a different concern.
Stay safe inside a Faraday cage exposed to artificial "lightning".
https://prepare-and-protect.net/2014/03/...day-cages/
Okay thank you. I'm still waiting on everything to come in so in the meantime I took deeper look at the cables and noticed this large indent on the riser cable. Would this be damaged enough to cause this issue? Considering it crashes the gpu mainly.
https://imgur.com/a/7QAaUnZ