Lenovo ST150 Will Not Power On Elimination Order Before You Blame the Board
The motherboard is the most expensive part in an ST150 and the part blamed first whenever one stops starting. This page works the elimination order technicians use — power, settings, minimal boot, memory, processor — so the board is convicted only when every cheaper suspect has been honestly cleared.
✍️ Technical Anshu🖥️ Lenovo ThinkSystem ST150⏱️ About 1 hour with basic tools
🎥 Watch the ST150 dead-board diagnosis in the video — the checks performed in order
There is a pattern in server repair: the part that gets replaced is the part someone suspected first, and the motherboard is always suspected first. The elimination order exists to break that pattern. It is deliberately boring — power before settings, settings before parts, parts before verdicts — and it is the difference between a one-hour diagnosis and a month of buying and returning boards. Follow it on your ST150 and whatever verdict you reach at the end, you will be able to defend it.
ST150 elimination order — cheapest suspect first
1. Socket, cable, and power supply railsThe most common "dead board" of all
2. CMOS battery and BIOS resetMinutes to test, nearly free
3. Minimal boot — one ECC stickRemoves drives and cards from suspicion
4. Known-good RAM swapECC sticks fail quietly
5. Known-good CPU swapLast and least likely
6. Board verdictOnly after 1 to 5
👆 Work top to bottom. Each step either finds the fault or removes a suspect from the list permanently.
1
Why Elimination Order Saves Real Money
Every dead ST150 arrives with the same sentence attached: "the motherboard is finished." Sometimes it is true. Very often it is a guess made because the board is the biggest part inside. The elimination order exists to protect your wallet: it tests the cheap, common impostors first and convicts the board only when nothing else is left standing. The full order takes about an hour with basic tools — far less time than returning a wrongly bought board.
2
Steps 1 and 2 — Power and Settings
Start where electricity starts:
Try a different wall socket and a known-good power cable. Server rooms hide dead sockets and half-broken cables surprisingly well.
Open the case and look before you touch: burn smell, scorch marks, bulged capacitors, or a fan that never even twitches all point in different directions.
Test the power supply output with a multimeter if you have one — main rails within about ±5% of their rating. No meter? A known-good supply swap is the same test in different clothes.
Replace the CMOS coin battery and clear CMOS by the jumper method in the ST150 manual. A flat battery plus corrupted settings can produce a machine that behaves as if its board died in the night.
3
Steps 3 and 4 — Minimal Boot and the RAM Swap
Now reduce the server to board, CPU, heatsink and a single memory stick:
All drives disconnected, all PCIe cards out, all USB devices unplugged. Front-panel and fan connectors stay as they are — photograph them before touching anything else.
Power on with no RAM fitted first: beeps or error indications mean the board is alive and complaining correctly. Add one ECC stick in Lenovo's specified single-stick slot and try again.
Swap that stick for a known-good ECC UDIMM. The ST150 uses DDR4 ECC unbuffered memory — ordinary desktop non-ECC sticks are not a valid test, because a platform that expects ECC can refuse to start for the wrong reason.
If two known-good sticks in two slots change nothing, RAM is cleared of suspicion and you move down the list with confidence.
4
Step 5 — The CPU, and Knowing Its Real Odds
Processors are the most reliable part in the machine, which is why they are tested last. If you have access to any working Xeon E3-1200 v5/v6 or compatible LGA1151 processor, one careful swap finishes the elimination: the same dead behaviour with a known-good CPU leaves the board with no alibi. Handle the swap like surgery — bent socket pins (inspect the LGA1151 socket under good light while the CPU is out) are themselves a classic "dead board" that a careful eye catches for free. A socket with crushed pins is a board-level repair or a replacement board, and now you know exactly which.
5
The Verdict — And What to Do With It
If steps 1 to 5 are done honestly and the server still shows no POST life, write the verdict without flinching: the board is guilty, and your money can now move to the repair-versus-replace decision, which this blog series covers in its own pages. One more honesty note: if you could not perform the RAM or CPU swap for lack of spares, your verdict is "board, most likely" — and a repair shop's bench test is a legitimate, cheap way to remove the doubt before buying parts.
⚠️ Safety, once, clearly: work with the server unplugged, never open the power supply unit (its capacitors hold charge), and photograph every connector before you remove it. Server front-panel wiring is not forgiving of guesswork on the way back.
🎥 Watch the Diagnosis in the Video
See the elimination checks performed live on the ST150, in the same order as this page.
My ST150 fans spin but there is no display. Is the board dead?
Not proven yet. Fans spinning only proves some power exists. Run the elimination order: a failed RAM stick, a corrupted BIOS setting, or a shorted add-in card can all produce exactly this picture. The minimal boot with one known-good ECC stick is the test that separates them.
Can I test the ST150 with ordinary desktop RAM?
It is a poor test. The ST150 is an ECC UDIMM platform, and mixing or substituting non-ECC desktop memory can make a healthy board refuse to start — giving you a false dead-board verdict. Test with known-good DDR4 ECC unbuffered memory.
How long does the full elimination take?
About 45 to 90 minutes with a multimeter and one spare ECC stick on hand. Most of that is careful disassembly and note-taking. It is far cheaper than discovering, after fitting a replacement board, that the power supply was guilty all along.
This guide is based on the Lenovo ST150 motherboard video by Technical Anshu • Stuck on a step? Ask in the video comments — answers come there 🙏