DisplayPort Pin 20 Power Backfeed? How to Tell Backfeed, Wake Faults & Bad Cables Apart

| HK GPU Fix

DisplayPort Pin 20 Power Backfeed? How to Tell Backfeed, Wake Faults & Bad Cables Apart

Key Takeaways

DisplayPort pin 20 backfeed happens when a non-compliant cable joins the 3.3 V supplies at both ends, so a powered monitor can keep PC lights and fans alive after shutdown or block the next power-on. The cable-swap and unplug tests separate it from bandwidth, EDID, hot-plug and monitor-power faults.

What Pin 20 Is, and Why It Must Not Be Wired End to End

A standard external DisplayPort connector has 20 pins. Pin 20 is DP_PWR: a 3.3 V supply, within a tolerance of plus or minus 10 percent, up to 500 mA, which is about 1.65 W at nominal voltage. It is provided by the connector on each device: the source side, typically the graphics card, provides DP_PWR at its own socket, and the sink side, typically the monitor, provides DP_PWR at its own socket. Its legitimate purpose is local: powering a dongle, active adapter, or similar device plugged into that socket, so the dongle needs no separate power lead. The monitor itself powers from the mains; it does not need power from the cable, and the cable does not need to deliver power to it.

The VESA position, as quoted in its former FAQ and in vendor technical notes, is that a standard DisplayPort cable must not carry pin 20 from one end to the other. In the cable, pin 20 is a no-connect. The reasons follow in order. First, tying two supplies together forces them to contend. No two 3.3 V supplies are exactly equal, and each carries its own tolerance, so joining them through a wire creates a contention condition between supplies, which the sources describe as effectively a short-circuit condition between supplies, with potential damage if the mismatch is large. This is the specification rationale and vendor warning, not a measured failure rate, and no incidence figure should be attached to it. Second, backdrive, or backfeed, occurs when one side is off: if the PC or graphics card is off, or its power supply is switched off, while the monitor is still powered from the mains, the monitor's DP_PWR can feed backwards through a wired pin 20 into the graphics card and motherboard. Low-power rails, LEDs, fan controllers, and standby circuitry can then stay partly energised even though the PC is off. That rogue power can also disturb the next power-up sequence, producing no-boot, failure to power on until the cable is unplugged, or erratic wake behaviour. Third, there is no need for the connection at all: since each end already provides its own DP_PWR for a local dongle, carrying power through the cable serves no purpose for a normal monitor connection.

It is important not to confuse DP_PWR with other power interfaces. DP_PWR at 3.3 V and low current for local dongle power is a different interface from USB-C Power Delivery and from monitor USB charging, with different power levels and different purposes. Arguments borrowed from those interfaces do not apply to a standard PC-to-monitor DisplayPort cable.

Exceptions That Are Not Backfeed Faults

Not every cable that involves pin 20 is non-compliant, and condemning every such product would be as wrong as ignoring the problem. A captive or attached cable, permanently fixed to a device, may legitimately carry DP_PWR, because it is part of that device's design rather than a generic interconnect joining two independent supplies. Active cables and adapters may legitimately use DP_PWR locally at one end to power their own electronics. The distinction is local use at one socket versus a straight wire joining both ends' supplies. A compliant active product terminates or manages pin 20; a cheap passive cable that simply wires all 20 pins straight through is the problem case. Mobile DisplayPort and similar variants, where power delivery through the cable is part of that variant's design, are a different interface from a standard PC-to-monitor DisplayPort cable and must not be used as evidence that standard cables should wire pin 20.

Non-compliant cables exist for a mundane reason: multiple sources agree that some manufacturers wired all pins straight through without understanding or following the specification, and cheap unbranded cables were widely sold that way. Certification is the countermeasure, since VESA-certified cables are tested not to pass pin-20 power end-to-end, but certification is not universal. Older discussion notes that few manufacturers pursued formal compliance testing, so absence of a logo does not prove a cable is bad, and the presence of the word certified in marketing copy is not proof either. Only a verified swap test, or genuine certification that can be checked, addresses the wiring question. Price tier and version number do not: the fault is a wiring choice, not a price, and a version label such as 1.2 or 1.4 describes bandwidth and features, not pin-20 wiring.

The Hallmark Pattern: Symptoms That Track Power State

Backfeed symptoms track power state, not picture quality, and they vanish when the cable path is broken. That single observation separates backfeed from most other DisplayPort faults, which are visible while the PC is running. A wired pin-20 cable can carry a perfect picture all day and still hold standby rails up after shutdown, so picture quality does not clear pin 20. The table below pairs the hallmark symptoms with the test that implicates or clears the cable.

SymptomTestInterpretation
Monitor will not fully power off, its power LED stays lit, or PC LEDs, fans, or RGB stay on or spin slowly after shutdownUnplug the DisplayPort cable with the PC still off and the monitor still on mains. If the stray lights or fans stop immediately and return when the cable is re-plugged, the cable path is carrying the powerStrongly points to pin-20 backfeed. A monitor-internal fault would not switch on and off with the DisplayPort cable alone
PC will not power on, or will not power on until the DisplayPort cable is unpluggedUnplug the DisplayPort cable, power on, then re-plug after bootBackfeed holding standby or power-sequencing rails in a wrong state is the leading explanation. If the PC still will not start with the cable out, the fault lies elsewhere, such as the power supply, board, or power button
PC will not fully shut down: chassis or CPU fans keep running and the power supply stays active, and only the rear power-supply switch kills everythingRepeat shutdown with the monitor unplugged from the mains while the DisplayPort cable stays connected. If shutdown now completes, the monitor was the power sourceConfirms the monitor side as the feeder. If fans persist with the monitor fully unpowered, suspect power-supply or board standby behaviour, not the cable
Intermittent no-signal, wake failure, or monitor not detected at boot, but only sometimesSwap to a known VESA-certified cable, or a cable known not to wire pin 20, and repeat the same power-on and wake sequence several timesIf the intermittency disappears with the swap and returns with the old cable, backfeed is supported. If it persists identically on a certified cable, it is probably not pin 20
Problem appears only after a cable change, or only on the DisplayPort input, while HDMI on the same monitor and graphics card works normallyCross-check: same monitor on HDMI, same graphics card on another monitor or inputLocalises the fault to the DisplayPort path. HDMI working does not by itself prove pin 20; it only rules out a dead monitor or graphics card. The cable-swap test is what implicates pin 20 specifically

Safe Tests, in Order

No test here involves cutting a cable, opening a connector, or probing a live pin. The safe tests, in order, are as follows. First, the swap test, which is the primary test: replace the suspect cable with a VESA-certified DisplayPort cable of the correct bandwidth for the mode in use, then re-test shutdown, power-on, and wake. Change one variable at a time: do not change the monitor, graphics-card port, and cable together, or a pass will not say which change mattered.

Second, the unplug test: with the PC shut down, unplug the DisplayPort cable from either end and observe LEDs and fans. Then unplug the monitor from the mains, with the DisplayPort cable left connected, and observe again. The two tests together separate power coming through the cable from power coming from the PC's own standby rail. Third, the monitor-off and PC-off matrix, covering four states: PC on with monitor on, which is normal use; PC off with monitor on, which is the backfeed window; PC off with monitor mains off, which should be fully dark; and PC on with monitor off, which tests wake and hot-plug behaviour. Backfeed symptoms live in the second state, and a symptom that appears equally in all four states is not a backfeed signature.

Fourth, single-monitor isolation: disconnect other monitors, cables, docks, and adapters, and test one graphics-card port to one monitor through one cable. Multi-monitor setups multiply variables, and one bad cable can hold up a whole system's power sequencing. Fifth, the input cross-check: the same monitor via HDMI, or another input, with the DisplayPort cable fully removed. If the power-state symptoms disappear on HDMI, the DisplayPort cable path is implicated; if they persist on HDMI with no DisplayPort cable connected, they are not pin 20.

Sixth, and optional, a continuity check on a disconnected cable only: with the cable unplugged from both devices, a multimeter in continuity mode between pin 20 at each end would show continuity on a non-compliant cable and open circuit on a compliant one. The practical limits must be stated plainly. DisplayPort pins are small and hard to probe reliably without fine probes, a failed probe contact looks identical to a compliant open circuit, and active cables may legitimately show different behaviour. This test can confirm a suspect cable, but it cannot clear one on a single sloppy attempt. For most readers the swap test is the usable test, and multimeter probing should not be presented as easy or definitive. Never probe pin 20 while either device is powered.

Telling Backfeed Apart from Bandwidth, EDID, Hot-Plug, and Monitor-Power Faults

No signal alone proves nothing. A monitor showing no signal is powered and listening on that input but receiving no usable video, and that state is shared by backfeed-related sequencing failures, bandwidth failures, EDID and hot-plug failures, wrong input selection, and a dead graphics-card output. The article must not let no signal stand as a diagnosis for any of them. The table below gives each fault family its signature and its discriminator.

Fault familyTypical signatureDiscriminator
Pin-20 backfeedPower-state symptoms: lights or fans persist after shutdown, no power-on until the cable is removed, wake anomalies; picture may be fine while the PC is runningSymptoms follow the cable and the monitor's mains state; they disappear on a certified-cable swap
Cable bandwidth or link-training faultWorks at low resolution or refresh, but fails, flickers, drops out, or shows artefacts at high resolution, refresh, or bandwidth modes; the link may train down to fewer lanes or a lower rateMode-dependent: lowering resolution or refresh to a conservative mode restores a stable picture, and a certified cable of adequate bandwidth fixes it. Power-state symptoms are absent
EDID faultMonitor not detected, wrong or missing resolutions, generic monitor identity, or the operating system falling back to a low default mode; a picture may appear but at the wrong modeA detection problem, not a power problem: the operating system does not see the right display identity. Direct connection without a dock, adapter, or KVM, and an EDID read through display settings or driver tools, differ between good and bad paths. Backfeed does not change EDID content
Hot-plug detect (HPD) faultMonitor not detected at boot, or detection is a coin toss at each boot, but once detected the session is stable; recovery may need a restart rather than a cable wiggleBoot-time, all-or-nothing pattern. HPD is a separate 3.3 V sink-to-source signal: high means present, low for more than 2 ms means absent, and a brief low pulse means an interrupt or a request to re-read. A backfeed cable can disturb power sequencing and mimic this, so the cable swap is still the tiebreaker, but a fault that survives a certified-cable swap is not pin 20
Monitor power faultMonitor dead on all inputs, power LED dark, on-screen display unavailable, or monitor powers off by itself regardless of cableFails identically on HDMI and DisplayPort, and with the cable unplugged; on-screen display and power behaviour are the tell. Backfeed, by contrast, needs the DisplayPort cable connected to manifest
Graphics-card or driver faultArtefacts, driver crashes or resets, black screen under load, fan ramping under load: symptoms tied to load or driver events, not to shutdown statePersists across cables and inputs and correlates with load or driver logs; the power-state tests are negative

The teaching point running through the table is that the discriminator is a test, not a symptom description. Two faults can share a symptom, such as no signal at boot, and only the swap test, the unplug test, and the power-state matrix separate them. Guessing from the symptom alone, however confident it feels, is how a cable gets replaced for a monitor fault, or a monitor gets replaced for a cable fault.

Edge Cases That Complicate the Picture

Multi-monitor, dock, and KVM chains each add a pin-20 path and an EDID and hot-plug path. Isolate to a single direct cable first: a dock or KVM that re-drives DP_PWR or mishandles hot-plug detection can reproduce backfeed-like or EDID-like symptoms without any single cable being non-compliant. Chasing the cable in a chain, without first removing the chain, wastes the test.

Adapters, including DisplayPort to HDMI, DVI, or VGA and mini-DisplayPort adapters, legitimately draw DP_PWR at the source socket, since active adapters need local power. An adapter chain that appears to fix backfeed by changing connector type has not necessarily proved pin 20: it has changed the whole signal path. Do not cite an adapter swap as pin-20 proof unless the DisplayPort cable itself was also swapped in isolation.

Monitor USB hub and standby power can mislead in the other direction. Some monitors keep USB ports powered in standby by design, and USB power staying on is not, by itself, pin-20 backfeed. Check whether the PC-side lights and fans follow the DisplayPort cable, not merely whether the monitor's USB stays live. Similarly, a PC's own standby rails legitimately keep some LEDs and USB alive when the machine is off but plugged in, under standby arrangements such as ErP and the plus-5V standby rail. The discriminator is the power-supply-off or mains-unplugged test: with the PC's power supply switched off or unplugged from the wall, its own standby rail is dead, so any remaining PC-side lights must be coming from elsewhere, through the cable.

A certified cable that still shows symptoms has three explanations, and none of them is that all DisplayPort cables are bad. The certified claim may be marketing rather than genuine VESA certification, the fault may not be pin 20 at all, or another link in the chain, such as a dock, adapter, or second monitor, may be the feeder. Do not escalate to cable surgery. On damage claims, the honest statement is limited: vendor notes warn of potential damage from contending supplies and disturbed power sequencing, but no source in this research quantifies how often damage occurs, and forum reports of a card failing are uncontrolled anecdotes. The article may say damage is warned of and that the safe action is replacement; it must not claim a probability, a typical repair cost, or that a given card was damaged by its cable. Continued use of a known backfeeding cable because it still works is not advised: a picture that works today says nothing about the next power sequence.

The Fix, and What Not to Do

The fix is replacement with a compliant cable: a VESA-certified DisplayPort cable of the correct bandwidth for the mode in use, verified by repeating the shutdown, power-on, and wake sequences that failed before. Do not cut, lift, or tape over pin 20, splice the cable, or open a connector to disable the pin. Forum posts float these modifications, but they risk bent pins, shorts, intermittent contact, and a cable that fails in a new way, and they void any cable warranty. Replacement is cheaper than the risk and is the fix every vendor source recommends. Do not probe pin 20 while either device is powered, and do not treat a multimeter reading from a sloppy probe as clearing a cable.

When buying a replacement, remember what the labels do and do not certify. A version number describes bandwidth and features, not pin-20 wiring. A brand name or a price tier does not certify wiring either, since the fault is a wiring choice. Marketing use of the word certified is not verification. Genuine VESA certification, checked rather than assumed, plus a swap test that repeats the failing power sequences, is the combination that settles the question. A certified cable fixes the pin-20 class of fault only; bandwidth, EDID, hot-plug, and monitor-power faults survive a cable swap and need their own diagnosis, as the differential table above shows.

Misconceptions to Leave Behind

DisplayPort carries power to the monitor, so pin 20 must be wired, is incorrect. DP_PWR is provided locally at each socket for dongles, and the monitor powers itself from the mains. A standard cable carrying power end to end is the defect, not the design.

If the picture is fine, the cable is fine, is incorrect. Backfeed is a power-state fault. A wired pin-20 cable can carry a perfect picture while the PC is running and still hold standby rails up after shutdown. Picture quality does not clear pin 20.

Any no signal on DisplayPort means pin 20 is incorrect. Bandwidth, EDID, hot-plug, wrong input, and graphics-card or driver faults all produce no signal. Pin 20 is implicated only by the power-state tests, above all the cable-swap and unplug tests.

An expensive, branded, or version-1.4 cable cannot have the pin-20 problem is incorrect. The fault is a wiring choice, not a price tier or a version number. Only genuine certification, or a verified swap test, addresses the wiring question, and marketing use of the word certified is not verification.

Cutting pin 20 or taping the pin is a safe permanent fix is incorrect. Physical modification risks shorts, bent pins, and intermittent contact, and voids any cable warranty. Replacement with a compliant cable is cheaper than the risk and is the recommended fix.

HDMI is immune, so switching to HDMI proves the graphics card is faulty, is incorrect. HDMI has its own plus-5 V pin and its own handshake faults, and an HDMI swap changes the whole path. HDMI working only localises the fault to the DisplayPort path; it does not diagnose the graphics card.

The monitor is faulty because it powers the PC is incorrect as a blanket claim. The monitor is supplying DP_PWR exactly as its socket is specified to do; the fault is the cable joining the two supplies. Blaming the monitor, or the graphics card, without the cable swap misdirects the fix. A monitor-side fault remains possible, but it must be shown by the monitor failing on other inputs or with a certified cable.

Stop Points and What to Record Before Seeking Inspection

Stop and replace the cable rather than modifying it if the power-state tests point to backfeed, and stop testing if there are signs of electrical damage, such as a burning smell, scorch marks, or a melted connector. Before seeking inspection, record the symptom in power-state terms, the four-state matrix results, the cable-swap result with the same power-on, shutdown, and wake sequences repeated, the single-monitor isolation result, and the HDMI cross-check result. That record lets an inspection start from the cable evidence rather than from a no-signal description alone. For inspection in Hong Kong, the shop is Room 1209, Dali Plaza, Shek Mun. The telephone number is 2186 8010 and the WhatsApp number is 6558 6806.

Frequently Asked Questions

How can a monitor keep my PC partly powered after shutdown?

Through a non-compliant DisplayPort cable that wires pin 20 end to end. The monitor's own DP_PWR supply then feeds backwards into the graphics card and motherboard, keeping low-power rails, LEDs, and fan controllers partly energised even though the PC is off, and it can disturb the next power-up sequence.

My picture is perfect. Can the cable still be a pin-20 cable?

Yes. Backfeed is a power-state fault, not a picture fault. A wired pin-20 cable can carry a perfect picture while the PC is running and still hold standby rails up after shutdown. Picture quality never clears pin 20; only the shutdown, power-on, and wake tests do.

What is the simplest test for pin-20 backfeed?

The cable-swap test, supported by the unplug test. Replace the suspect cable with a VESA-certified cable and repeat the same shutdown, power-on, and wake sequences. Separately, unplug the DisplayPort cable with the PC off and the monitor on mains: if stray PC lights or fans stop and return on re-plugging, the cable path is carrying the power.

Does no signal on DisplayPort mean pin 20 is the cause?

No. No signal only means the monitor is powered and listening on that input but receiving no usable video. Bandwidth failures, EDID faults, hot-plug faults, wrong input selection, and graphics-card or driver faults all produce the same display. Pin 20 is implicated only by power-state evidence, not by no signal alone.

Will an expensive or version-1.4 cable guarantee there is no pin-20 problem?

No. The fault is a wiring choice, not a price tier or a version number. Version labels describe bandwidth and features, not pin-20 wiring. Genuine VESA certification, checked rather than assumed from marketing copy, plus a verified swap test, is what addresses the wiring question.

Can I fix it by cutting or taping over pin 20?

No. Physical modification risks bent pins, shorts, and intermittent contact, voids any cable warranty, and can create a cable that fails in a new way. The recommended fix is replacement with a compliant, VESA-certified cable of the correct bandwidth, verified by repeating the failing power sequences.

Why does my PC refuse to power on until I unplug the DisplayPort cable?

Because backfed power can hold standby or power-sequencing rails in a wrong state, so the normal power-on sequence cannot proceed cleanly. Removing the cable removes the rogue supply, and the sequence can then run. If the PC still will not start with the cable removed, the fault lies elsewhere, such as the power supply, board, or power button.

How do I tell backfeed from a monitor power fault?

A monitor power fault fails identically on HDMI and DisplayPort and with the cable unplugged, with a dark power LED or unavailable on-screen display. Backfeed needs the DisplayPort cable connected to manifest, and its symptoms switch on and off with the cable and with the monitor's mains state.

Can a dock, KVM, or adapter cause backfeed-like symptoms?

Yes. Each link adds a pin-20 path and an EDID and hot-plug path, and a dock or KVM that re-drives DP_PWR or mishandles hot-plug detection can reproduce backfeed-like or EDID-like symptoms without any single cable being non-compliant. Isolate to a single direct cable first before blaming one cable.

What if a certified cable does not fix the symptoms?

Then either the certified claim is marketing rather than genuine certification, the fault is not pin 20, or another link in the chain is the feeder. Return to the differential table: bandwidth, EDID, hot-plug, monitor-power, and graphics-card or driver faults all survive a cable swap and need their own diagnosis, rather than escalating to cable modification.

Related articles

Related services

If you need Power connector repair, see these service pages: