GPU Works Plugged Straight Into the TV but Black-Screens or Loses 4K 120Hz Through an AV Receiver? Direct-Connection Test, Receiver Bandwidth, EDID & HDCP Repeater Diagnosis Guide

2026-11-03 | GPU Repair | 16 minutes | HK GPU Fix Editorial

GPU Works Plugged Straight Into the TV but Black-Screens or Loses 4K 120Hz Through an AV Receiver? Direct-Connection Test, Receiver Bandwidth, EDID & HDCP Repeater Diagnosis Guide

Key Takeaways

When a GPU works perfectly plugged straight into the TV but black-screens or drops to a lower mode through an AV receiver, the receiver chain itself is the suspect: this guide uses a direct-connection comparison, port-by-port checks, and content tests to separate bandwidth limits from EDID and HDCP repeater problems.

What Exactly Counts as This Symptom

This guide deals with a narrowly shaped symptom, and the shape has to be matched first. The graphics card, connected directly to the television with a single cable, produces a stable picture at the expected resolution and refresh rate. Insert an AV receiver into the chain, with the card feeding the receiver and the receiver feeding the television, and something is lost: the screen goes black, the picture drops to a lower resolution or refresh rate, or a feature that worked direct, such as HDR or variable refresh rate, quietly disappears. The defining feature is the contrast between the two arrangements. Direct works; through the receiver does not.

Several lookalike situations must be ruled out first, or the whole diagnosis will head in the wrong direction. First, a card that fails even when plugged directly into the television is not a receiver problem at all, and the receiver section of this guide will not help it. Second, a picture that is present but silent is a different symptom: this guide is about a lost picture or a lost picture mode, and a purely silent picture belongs with audio-output diagnosis. Third, a television that has simply been left on the wrong input, or a receiver left on the wrong input, is a selection mistake rather than a chain fault, and it should be checked before any deeper test. Fourth, a duplicated display feeding a capture device or a second screen at the same time is a different arrangement from the single serial chain covered here. Match the symptom before changing any setting, because a receiver setting changed for the wrong symptom will only add confusion.

The direct-connection result is itself powerful evidence, and it is worth stating plainly what it proves. If the same card, the same television, and a known-good cable produce a correct picture direct, then the card output and the television input are basically working. The fault, or the limitation, lives in what was added: the receiver, the second cable, or the receiver settings. That single inference is what separates this guide from general no-signal diagnosis, where the card, the cable, and the display are all still suspects. Here, two of them have already been cleared by the direct test.

The Mechanism: Why Adding One Box Can Break a Working Picture

A direct chain has two ends: the graphics card as source and the television as display. Adding a receiver turns it into three segments: card to receiver input, receiver output to television, with two cables instead of one. The capability of the whole chain is set by its weakest segment, not by its strongest end. A television that accepts a demanding mode and a card that can output it are not enough if any segment in between cannot carry that mode.

A receiver is not a passive cable. It is an active repeater, and that word matters. It reads the television EDID, which is the display capability list covering resolution, refresh rate, HDR formats, colour depth, and similar items, and then it reports a capability set back to the graphics card. It may also process the picture, for example by upscaling it, scaling it, or overlaying its own on-screen menu. Every one of those steps can change the signal that was originally sent. A misreported capability list, an extra processing step, or an input port that simply does not support the requested mode can each produce a black screen, a downgraded mode, or a silently missing feature.

Bandwidth is a hard limit, not a preference. A picture mode, defined by its resolution, refresh rate, colour depth, and HDR content, needs a certain amount of bandwidth, and that bandwidth must fit within every segment at once: the card output, the receiver input, the receiver output, the television input, and both cables. The published HDMI standards give useful reference points: HDMI 2.1 is specified at 48 Gbps and HDMI 2.0 at 18 Gbps. Those are standard specifications, not promises about any particular receiver. A receiver case marked with a high standard does not mean every input port on that receiver supports every mode at that standard. Ports on the same receiver can differ in bandwidth and in their enhanced-mode settings, so each port must be checked in the manual and in the receiver menu individually. Never assume.

HDCP, the copy-protection system for high-bandwidth digital content, adds a second chain on top of the picture chain. For protected content, such as a 4K stream from a streaming platform, the source, the receiver as repeater, and the television must all complete an authentication handshake, and every link in that chain must be compatible. If any link fails or uses an incompatible version, the picture can go fully black or quietly downgrade. Ordinary game and desktop pictures generally do not travel through this protection chain, so a pattern in which games work but protected streams go black is itself a diagnostic clue, not a contradiction. It points at the protection chain rather than at the card.

Handshake timing is the third mechanism. When devices power on, switch inputs, or change modes, the three devices must complete EDID reading and HDCP authentication in sequence. If one device is not ready yet, if the power-on order is scrambled, or if the receiver is in standby passthrough rather than powered processing, the result can differ from one attempt to the next. Intermittent behaviour on the same receiver therefore does not automatically mean a loose connection. It can mean a handshake sequence problem, and the power-on order test later in this guide exists precisely to separate those two explanations.

ObservationMore likely explanationWhat to do next
Direct to the television works, through the receiver the screen is blackReceiver chain, second cable, or receiver setting is the suspect; card and television are basically clearedKeep the direct result as the baseline and add the receiver back one variable at a time
Direct works, through the receiver the picture drops to a lower resolution or refresh rateA segment of the chain cannot carry the requested mode, or the reported capability list is narrowerStep the mode down one item at a time to find where stability returns
Direct works, through the receiver HDR or variable refresh rate quietly disappearsThe capability list passed through the receiver omits that feature, or a processing step strips itCheck the enhanced mode of that exact input and switch video processing to passthrough
Games and desktop work through the receiver, only protected streaming goes blackHDCP repeater chain problem, not a general bandwidth or card problemFollow the content split in this guide rather than replacing cables or the receiver
The result changes with power-on order, or works in standby passthrough but not when poweredHandshake timing or processing-path problemCompare a fixed power-on order against a random order, and test both receiver states separately
One receiver input fails while other inputs on the same receiver workThat input setting or port capability, not the whole receiverCompare the failing input settings against a working input, port by port

Step One: The Direct-Connection Comparison

This is the decisive first step, and it must be done cleanly. Use the same graphics card and the same television, connect them directly with one cable, bypassing the receiver entirely, and use a known-good cable if one is available. Test the same mode that failed through the receiver. If the direct picture is stable and correct, the receiver chain is confirmed as the area to investigate, and the card and television are basically excluded. If the direct picture also fails or also downgrades, stop: this is not a receiver problem, and continuing to adjust receiver settings will waste time. The fault then lies with the card, the cable, the television, the driver, or the mode itself, and general no-signal or mode-limit diagnosis is the right path instead.

Record the direct result precisely before moving on. Note the resolution, refresh rate, HDR state, and variable refresh rate state that worked direct, because every later comparison is measured against that baseline. A vague memory such as direct seemed fine is not enough: later, when the receiver chain drops one feature, the question will be exactly which feature was lost relative to the direct baseline, and that question can only be answered from a written record.

Step Two: Add the Receiver Back One Variable at a Time

With a working direct baseline, add only the receiver, changing nothing else at first. Do not simultaneously change cables, ports, and settings. If the fault returns as soon as the receiver is added, the receiver chain is confirmed, and the next job is to narrow down which part of it is responsible. Work through the variables one at a time, testing after each single change, and writing down the result of each change.

First, try a different input port on the same receiver, keeping everything else identical. If a different input works while the original input fails, the problem points at the original input setting or port capability, not at the receiver as a whole. Second, replace one cable segment at a time: either the card-to-receiver cable or the receiver-to-television cable, but never both together. If replacing one segment restores the picture, that segment was the weak link, and the record should say which segment it was. Third, check the enhanced mode of the input in use. Receivers commonly label this with names such as 4K Enhanced, 8K Enhanced, or HDMI Enhanced, and the setting is often made per input rather than once for the whole receiver. Enter the receiver menu and check the exact input being used; an enhanced mode left off on that input alone is enough to block a demanding mode while other inputs appear fine. Fourth, switch the receiver video processing or upscaling to passthrough, so the receiver passes the picture through without altering it, and test again. If passthrough restores the picture, a processing step was changing the signal.

Each result has a specific meaning, and the meaning should not be overstated. A different port working points at the old port setting; it does not prove the receiver is otherwise perfect. A cable segment fixing the fault identifies that segment; it does not prove the other segment is flawless. Passthrough working shows processing was involved; it does not by itself identify which processing step was responsible. One variable at a time is slower than changing everything at once, but it is the only way the result can identify a cause. Change several things together and a recovery teaches nothing, because the next change of equipment will bring the same fault back.

Step Three: The Mode Step-Down Comparison

On the same receiver chain, lower the demands of the picture mode one item at a time, testing after each step: first lower the refresh rate, then lower the resolution, then switch HDR off, each as a separate step. Note the step at which the picture becomes stable. If lowering one item restores stability, a bandwidth or capability limit is the strongest suspect. If even a low mode stays black, the suspicion shifts toward handshake, EDID, HDCP, or connection problems, because a pure bandwidth limit should ease as demands fall.

A step-down result must not be misread as a repair. Running at a lower resolution or refresh rate, or without HDR, is a way of locating the bottleneck, not a fix for it, and it carries a real cost in picture quality. The value of the test is in what it narrows down: it shows how much demand the chain can carry, and it gives the next test a concrete starting point. A chain that is stable only after a step down still has an unidentified weak segment, and the port, cable, and setting tests remain the way to find it.

Step Four: The Content Split

Test two kinds of content through the same receiver chain: ordinary desktop or game pictures, and protected streaming content such as a 4K stream from a streaming platform. If both kinds fail together, bandwidth, EDID, or handshake problems are the stronger suspects. If desktop and games are fine and only protected content fails, the HDCP repeater chain becomes the prime suspect, and the investigation should turn to the protection chain and the platform requirements rather than to cables or the receiver as a whole.

This split matters because the two branches lead to different actions. A general failure calls for port, cable, bandwidth, and setting checks. A protected-content-only failure calls for checking the receiver and television compatibility with the protection chain, checking firmware notes for known fixes, and checking the platform own requirements for the content in question. Mixing the branches up, for example by replacing a receiver because a protected stream went black, can waste effort on a chain that was never the problem.

Step Five: The Power-On Order Comparison

Power the chain up in a fixed order, television first, then receiver, then the computer as source, and compare the result against a random order. Also test the receiver two states separately: standby passthrough, in which the receiver passes the signal while in standby, and powered processing, in which the receiver is fully on and processing the picture. If a fixed order is stable while a random order is not, handshake timing is implicated. If standby passthrough works but powered processing fails, the processing path is implicated rather than the cable. If powered processing works but standby passthrough fails, the standby path is the suspect. The two states must be tested separately and never lumped together as the receiver works or the receiver fails.

An Alternative Arrangement After Diagnosis

Once the receiver video chain has been identified as the weak segment, one workable arrangement is to connect the graphics card directly to the television for picture, and send sound back from the television to the receiver using the television ARC or eARC return channel. This rearranges the chain so the receiver no longer sits in the video path, while surround sound is retained through the return channel. It is important to describe this accurately: it is a chain rearrangement, not a graphics card repair, and the sound formats that pass through it are limited by what the television ARC or eARC capability supports. Not every format can be assumed to pass. A purely silent-picture problem, with the picture itself fine, still belongs with audio-output diagnosis rather than with this arrangement.

Boundary Cases: When It Is Not the Receiver

First, direct connection also fails or also downgrades: not a receiver problem, whatever the receiver settings say. Second, only protected streaming content fails while games and desktop are fully normal: prioritise the HDCP chain and platform requirements before considering receiver replacement. Third, only one receiver input fails while other inputs work: that input setting or port is the suspect, not the whole receiver. Fourth, a firmware update may be a legitimate step when the receiver or television maker has published a known fix, but it must be handled with care: note the original settings before updating, understand the power-loss risk during an update, and never treat a firmware update as a universal cure. Fifth, standby passthrough working while powered processing fails, or the reverse, points at the processing path or the standby path respectively, not at the cable. Each of these boundaries exists to stop one symptom from being forced into the wrong explanation.

CheckWhat a pass provesWhat a pass does not prove
Direct connection worksCard output and television input are basically workingNothing about the receiver chain, which has not been tested yet
A different receiver input worksThe original input setting or port was involvedThat the receiver is fully healthy in every mode
Replacing one cable segment worksThat segment was the weak linkThat the other segment can carry every mode
Passthrough restores the pictureA processing step was changing the signalWhich processing step, or that the receiver is faulty
Stepping the mode down restores stabilityThe chain cannot carry the original demandWhich segment is the bottleneck, without further segment tests
A fixed power-on order worksHandshake timing was involvedThat the chain can carry the demanding mode in every state

Common Misconceptions

Misconception one: a case marking of a high HDMI standard means every port and every mode works. A marking can describe the highest capability or a particular port only; the actual capability and enhanced-mode switch of each port is what counts. Misconception two: an expensive cable must have enough bandwidth. Price is not certification, and the certification marking plus an actual mode comparison is what counts. Misconception three: a black screen means the graphics card is faulty. Before the direct comparison has been done, the chain has not been excluded at all, and this whole guide exists because a direct test can clear the card. Misconception four: running at a lower resolution counts as fixed. A downgrade only locates the bottleneck, and it costs resolution, refresh rate, or HDR. Misconception five: the receiver only handles sound, so it cannot affect the picture. In a serial chain the receiver is also the video repeater, and the capability list it reports plus its processing path directly decide what mode the card outputs. Misconception six: changing cables, ports, and settings all at once and getting a picture is good enough. A recovery with several changes teaches nothing about which change mattered, and the fault will return with the next equipment change. Misconception seven: adding an external EDID manager or splitter will certainly solve it. Such a device is only a possible aid with its own compatibility and setting questions, and it must not be presented as a guaranteed purchase or a promised result; the point of this guide is location, not accessory promotion.

When to Stop and Seek Repair

Stop experimenting with settings if any of the following appears. The direct connection also fails, which removes the receiver from the picture entirely. A firmware update is being considered but the maker guidance is unclear, or the update process itself looks risky. The picture stays black in every setting, with not even a low mode getting through. The receiver or TV shows signs of a hardware problem beyond settings, such as a loose port, a burning smell, intermittent power loss, or an input that never responds in any configuration. Or the picture fault comes together with other symptoms, such as crashes or driver resets, which go beyond a chain problem. Bring four items to inspection: the direct-connection baseline with its exact mode, the result of each single-variable test, the step at which a mode step-down became stable, and the content split result showing whether protected content alone failed. Those four let repair staff start from the chain directly, without repeating tests already done.

Readers nearby are welcome to bring their records to HK GPU Fix for advice, at Room 1209, Dali Plaza, Shek Mun, telephone 2186 8010, WhatsApp 6558 6806. Describing the direct baseline, which receiver input and setting was tested, and which step first restored stability is enough for a quick judgement on the next step.

Frequently Asked Questions

Why does the picture work direct but not through the receiver?

Because the receiver is an active repeater, not a passive cable. It reads the television capability list, reports a capability set to the card, and may process the picture. Any of those steps can narrow or change the signal, so a chain that works with two ends can fail with three segments.

Does a black screen through the receiver mean the graphics card is faulty?

No, not on this evidence. If the direct connection works, the card output has already been shown to work. The receiver chain, the second cable, or a receiver setting is the area to investigate first.

Why can one receiver input fail while another works?

Inputs on the same receiver can differ in bandwidth and in their enhanced-mode settings, which are often set per input. The failing input may have its enhanced mode off or a lower capability, while another input is configured differently.

What does the enhanced mode setting actually do?

It allows that input to carry a more demanding mode. With it off, the input may refuse or downgrade a high resolution, high refresh rate, or HDR mode even though the receiver as a whole is marketed for that standard.

Why do games work but protected streaming goes black?

Games and desktop pictures generally do not travel through the HDCP protection chain, while protected streaming content does. A failure limited to protected content points at the protection chain through the receiver, not at general bandwidth or at the card.

Is running at a lower resolution a fix?

No. A step down only locates the bottleneck by showing how much demand the chain can carry. It costs resolution, refresh rate, or HDR, and the weak segment still needs to be identified by port, cable, and setting tests.

Can power-on order really change the result?

Yes. Devices must complete EDID reading and HDCP authentication in sequence, and a device that is not ready yet can produce a different result. A fixed order of television, then receiver, then source is a legitimate comparison test against a random order.

What should I bring if I seek inspection?

Bring the direct-connection baseline with its exact mode, the result of each single-variable test, the step at which a step-down became stable, and whether protected content alone failed. That record lets repair staff start from the chain rather than repeating the whole diagnosis.

Related articles

Related services

If you need Chip-level GPU repair, see these service pages: