Power-related damage is the single most common issue we see on industrial PCs returned to us. Not overheating, dust, or software problems. Power.
In most cases, the customer doesn’t know what caused the failure. The PC was connected to what seemed like a perfectly adequate power supply, in an environment where everything else was running fine. But industrial power environments are not as clean or stable as they appear, and the assumptions engineers make during specification often do not account for what actually happens on the supply rail when everything is running at once.
The shared DC supply problem
The most common scenario is straightforward. An industrial PC is connected to a 24V DC power supply that also feeds a network switch, some I/O modules and a few solenoids or contactors. The steady-state power consumption of everything on the rail adds up to well below the power supply’s rating. On paper, there is plenty of headroom.
but the problem is not steady state. It’s transient. When a contactor or solenoid switches, it draws a high current momentarily. That inrush can pull the supply voltage down by several volts. After the dip, the voltage can overshoot above the nominal 24V as the supply recovers. If several contactors or solenoids switch at the same time, the combined effect is larger.
An industrial PC with a power input rated for 24V with a 6% overvoltage tolerance has a maximum of about 25.4V. A transient overshoot on a shared 24V rail can easily exceed that, particularly when multiple inductive loads switch simultaneously. The result is damage to the PC’s power input circuit.
This is why we recommend 9-36V DC wide-range power input for any industrial PC connected to a shared DC supply. The wider input range provides genuine headroom for the voltage dips and surges that are normal on a shared industrial supply rail. The cost difference is modest. The protection it provides is significant.
Lightning and surge protection
Lightning surge damage is another issue we see regularly, particularly at remote or exposed sites. The mechanism is often misunderstood.
A direct lightning strike is rare. What is common is induced surge current from a nearby strike, coupled onto long cable runs that are not well shielded. This affects both power supply cables and signal lines. A long RS-485 cable run, an Ethernet connection to a remote sensor, or a DC power feed from a distant supply can all carry induced surge currents into the PC.
Effective surge protection requires properly rated devices installed on both power and signal lines, at both ends of long cable runs. A surge protector rated to 48V DC will not clamp a 12V supply well enough to prevent damage. The other critical factor is the earthing of the surge protection devices, which requires a short, low-impedance path to ground. A surge protector connected to a long, undersized earth wire will not do its job effectively. The earthing design matters as much as the protection device itself.
We have seen failures where either the power lines or the signal lines were protected, but not both. A surge will find the unprotected path. Both need to be addressed to protect the equipment properly.
Vehicle and marine applications
Vehicles and small boats present a particularly harsh power environment that is easy to underestimate. The nominal supply might be 12V or 24V DC, but the actual voltage can vary dramatically.
- Load dump is the most severe event. If the vehicle battery is disconnected while the alternator is running, or if the battery terminals have a poor connection, the alternator output can spike as high as 100V briefly. Similar spikes occur when inductive loads such as motors are suddenly switched off.
- Engine cranking causes the supply voltage to swing from well below nominal to well above it within seconds, as the starter motor loads and then releases the electrical system.
A standard 9-36V wide-range power input may not provide sufficient protection against these extremes. For vehicle and marine applications, we recommend an external DC-DC converter between the vehicle supply and the PC. These converters are designed specifically for automotive and marine power environments, with input ranges and transient protection that go well beyond what an internal PC power supply can handle.
What fails inside the PC
The extent of the damage depends on the severity and type of power event.
Overvoltage on a standard DC input most commonly blows a power supply capacitor. This is a relatively straightforward repair.
Surges on a wide-range input can damage the internal DC-DC converter module. These can also usually be replaced without replacing the entire PC.
Severe surges, such as lightning-induced transients or major load dump events, can cause extensive damage to the motherboard itself. In many cases, repairs aren’t economically viable, or they require a return to the overseas factory. At that point, the cost is not just the repair but the downtime and the need to source, configure and commission a replacement.
The pattern is clear: the worse the power event, the more expensive the recovery. And in almost every case, the damage was preventable with the right specification and protection.
AC power is less of a problem
Power quality issues are less common for AC powered industrial PCs. Good quality industrial power supplies from manufacturers like Meanwell, Bicker, FSP and Zippy are designed for wide input voltage ranges, often 90-264V AC, and include built-in protections appropriate for continuous 24/7 industrial operation. They are significantly more resilient than consumer-grade power supplies.
That said, we do occasionally get asked about AC power data logging from customers experiencing repeated equipment failures on mains-powered gear, usually non-industrial equipment that lacks the same built-in resilience. Often the underlying issue is a mains supply that is consistently over voltage, and the customer needs data to present a formal complaint to their supply utility.
Getting it right at specification time
Power problems are the easiest class of industrial PC failure to prevent, because the solutions are known and the cost of protection is low relative to the cost of failure.Here’s how:
- Use wide-range DC input as the default. Unless the PC is connected to a dedicated, isolated power supply with no other loads, specify 9-36V DC input rather than standard 12V or 24V. The additional cost is small. The protection is substantial.
- Audit the shared supply. Before connecting a PC to an existing 24V rail, identify every other load on that supply, including inductive loads like solenoids, contactors and relays. Consider whether simultaneous switching could cause transients that exceed the PC’s input tolerance.
- Protect both power and signal lines. At remote or exposed sites, install surge protection with a suitable rating on all cable runs entering the equipment, not just power. Ensure the surge protection earth path is short and low impedance.
- Use external DC-DC converters for vehicles and marine. Don’t rely on the PC’s internal power input to handle automotive or marine power transients. An external converter designed for these environments provides a level of protection that internal power circuits can’t match.
- Talk to your supplier. Our sales team checks the customer’s intended power source on every order, because the wrong power input specification is the single most preventable cause of failure we see. A two-minute conversation at quoting stage can prevent an expensive failure months later.
How ESIS can help
We have seen every type of power-related failure across industrial, utility, vehicle and marine applications. That experience informs how we specify power input options, what protection we recommend, and what we check before any system leaves our workshop.
If you are specifying an industrial PC and are unsure about the power environment, talk to us before you order. We would rather spend five minutes getting the specification right than process a warranty return three months later!
Contact us now to get the right advice and equipment for your project.
