Troubleshooting

Troubleshooting

This page is organized by symptom - what you see that looks wrong. For each symptom, it explains why it is happening and what to do about it.

Some of these "problems" are actually normal behavior that looks unexpected on first encounter. Others are genuine configuration issues with clear fixes. The distinction matters, and each entry tells you which one you are dealing with.

The indicator halted with a runtime error naming a slot

What you see: The indicator is not loading. TradingView shows a runtime error panel with a message that mentions a specific slot (for example, "DC 01 timeframe cannot be lower than the chart timeframe").

Why it happens: You set a slot's timeframe to something shorter than the chart's current timeframe. The indicator requires each slot's timeframe to be equal to or greater than the chart timeframe. When this rule is violated, the indicator halts with a labeled error rather than producing incorrect data.

What to do:

1. Read the error message - it names the offending slot.

2. Open the indicator settings and either increase that slot's timeframe to match or exceed the chart timeframe, or switch your chart to a lower timeframe.

3. The indicator will reload automatically once the settings are valid.

This is not a bug. It is a protective stop. Running a higher-timeframe Donchian Channel calculation at a timeframe lower than the chart would produce meaningless results. The indicator halts and names the slot rather than producing silently incorrect output.

The channels are flat and not updating

What you see: One or more slot channels appear as flat horizontal lines that do not change, even though price is moving.

Why it happens (most likely): On Bar Close is ON (the default), and the slot's timeframe has not completed a new bar yet. With confirmed-bar behavior, the channel only updates when a higher-timeframe bar closes. Between closes, the channel is flat.

On a 1-minute chart:

• A 5-minute slot updates every 5 bars (5 minutes)

• A 15-minute slot updates every 15 bars (15 minutes)

• A 60-minute slot updates every 60 bars (1 hour)

Between updates, the channel shows the last confirmed value. The flat line means "the last confirmed HTF bar's range has not changed because no new HTF bar has closed."

What to do:

• If this is expected cadence: Nothing. Wait for the next HTF bar to close and the channel will step to new values. This is normal confirmed-bar behavior.

• If you want to confirm it is working: Switch to a 1-minute chart and watch. After the slot's timeframe bar closes (for example, at the top of each 5-minute boundary for a 5m slot), you should see the channel step.

• If you want faster updates and accept the tradeoff: Turn On Bar Close OFF in PU Settings. The channels will update in real time within the HTF bar, but historical review stops lining up cleanly with what the indicator showed live. See MTF & Repainting.

The channels show na (no lines) for a large portion of the chart

What you see: The channel lines do not appear for the first section of the chart. They start drawing partway through the visible history.

Why it happens: The Donchian lookback requires enough higher-timeframe bars to compute the highest high and lowest low. If a slot's timeframe is 60 minutes with a length of 20, the indicator needs at least 20 completed hourly bars before it can produce a value. Until those bars are available, the slot returns na.

This is more visible with high timeframes and long lookback lengths. A Weekly slot with a length of 52 needs a year of weekly data. If the chart does not have that much history available, the slot stays na for the entire visible range.

What to do:

• Reduce the lookback length to require fewer bars. A length of 10 instead of 20 means the channel starts producing values sooner.

• Reduce the slot's timeframe to work with smaller bar intervals that accumulate faster.

• Load more history by scrolling left on the chart or switching to an instrument with deeper data availability.

• Accept it if the high-TF/long-lookback combination is what your analysis requires. The na period will fill in as more bars form over time.

I set an Optional Ticker and the channel looks wrong

What you see: After entering a symbol in the Optional Ticker field, the slot's channel appears unusually tight, unusually wide, or offset in a way that does not seem to match the chart's price action.

Why it happens: The cross-ticker logic pulls the other instrument's Donchian Channel and then tries to scale it into the chart's price space using a ratio of the two instruments' closes. When that ratio behaves normally, the slot carries the other instrument's structural shape in your chart's coordinates.

If the two instruments trade at very different price levels (for example, SPY at $550 and your stock at $15), the ratio compresses or expands the channel values. If they have very different volatility profiles, the scaled channel width will not match the chart's own volatility.

There is also a code edge worth knowing: if the script cannot compute a valid ratio, it does not force the slot to na. It leaves the alternate symbol's raw channel values in place. On a differently priced instrument, that can throw the slot far off-scale instead of simply hiding it.

What to do:

• Check if the behavior makes sense: Compare the slot to what the other instrument's own DC looks like on its own chart. The shapes should correspond even if the absolute levels look different on your chart.

• Try a more closely related instrument: If the scaling looks erratic, the instruments may not be correlated enough. Switch to a more closely related reference instrument and see if the slot behavior improves.

• Watch for the raw fallback path: If the slot suddenly looks absurdly high, low, or flat relative to your chart, the ratio may have failed and the script may be showing the alternate symbol's raw values instead of a cleanly scaled overlay.

See Under the Hood for how the scaling mechanism works and Limitations & Trust Boundaries for where it stops being trustworthy.

The blended channel does not match the visible slots

What you see: The blended channel (red lines, lime basis) sits in a position that does not seem to correspond to the visible individual slot channels.

Why it happens (most likely): A hidden slot is still contributing to the blend. Hiding a slot removes its visual lines from the chart but does not remove its weight from the blended calculation. The blend includes the hidden slot's values at its full configured weight.

What to do:

1. Check whether any slot has "Hide DC XX Plot" enabled. If so, that slot is still in the blend.

2. If you want to exclude the slot from the blend entirely, either set its Blended Weight to 0 or disable the slot.

3. If you intended the hidden slot to contribute to the blend, this is working as designed - the blend just looks "off" because you cannot see all the inputs going into it.

Other possible causes:

• Unequal weights: If one slot has a much higher weight, the blend sits closer to that slot than you might expect from a visual average of the three channels.

• Cross-ticker slot influence: If one slot uses a cross-ticker and is contributing to the blend, the blend is averaging values from a different instrument alongside the chart's own structural values. That can pull the blend to unexpected positions, especially if the slot is using the raw fallback path instead of a clean ratio-scaled overlay.

• Missing slot values being coerced to zero: The blend uses nz() when it pushes slot values into its arrays. If a slot is enabled but still warming up, its missing values become 0 inside the blend instead of staying na.

Turning On Bar Close off made history look different

What you see: After toggling On Bar Close from ON to OFF, the channel positions on historical bars appear to have changed.

Why it happens: With On Bar Close ON, historical bars show the previous confirmed HTF bar's values. With On Bar Close OFF, historical bars show the final confirmed HTF bar's values (the current bar's confirmed result rather than the previous bar's). The two are offset by one HTF bar. The visual result is that the entire channel history appears to shift by one HTF bar period.

This is not the indicator changing its past values. It is the indicator switching which HTF bar it reports - the previous confirmed one or the current confirmed one (which was the building bar when it was live). The two modes show genuinely different data series.

What to do:

• If you toggled it to experiment, toggle it back to ON and confirm the history returns to its previous state.

• Understand that the two modes are showing different things, not "correcting" each other. See MTF & Repainting for the full explanation.

The blended channel is missing or sitting somewhere impossible

What you see: The blended channel space looks empty, or the blend seems to have vanished off-scale, even though individual slots appear to be working.

Why it happens:

• Blended DC is disabled. Check that "Enable Blended DC" is ON in the settings.

• Blended DC plot is hidden. Check that "Hide Blended DC Plot" is OFF.

• No enabled slots are contributing. If all slots are disabled while the blend stays enabled, the blend math falls back to zero-valued lines. On most charts that means the blend is effectively off-screen, not truly absent.

• Enabled slots are still warming up. The blend does not wait for clean na handling. Missing slot values are coerced to 0 before blending, so a slot with insufficient history can pull the blended lines toward zero.

• All weights are zero. If every enabled slot has a Blended Weight of 0, the blend collapses to zero-valued lines and the blended alerts stop being meaningful.

• A cross-ticker slot is using the raw fallback path. If the script cannot compute a valid scaling ratio, the alternate symbol's raw values can yank the blend to a place that makes no visual sense on your chart.

What to do: Check the Enable and Hide settings first. Then verify that at least one enabled slot has non-zero blend weight and is producing meaningful values. If the blend still looks wrong, inspect each slot for insufficient history or a cross-ticker ratio failure.

Alerts are not firing

What you see: You set an alert based on one of the indicator's conditions, but it never triggers.

Why it happens:

• The condition has not been met on a confirmed bar. All alerts require bar-close confirmation. The condition must be true at the moment the chart-timeframe bar closes. If the condition was briefly true during bar formation but false at close, the alert does not fire.

• The slot is disabled. Per-slot alerts require the relevant slot to be enabled. An alert set for "DC 03 Is Above Basis" will never fire if DC 03 is disabled.

• The slot's values are na. If the slot does not have enough history to compute values, its basis is na, and the alert condition cannot be met.

• The blended channel is disabled. Blended alerts require the blended DC to be enabled.

• The blend has collapsed to junk values. If every enabled blend weight is 0, or enabled slots are still being coerced to 0 inside the blend, the blended alerts may technically evaluate but stop meaning anything useful.

What to do:

1. Confirm the relevant slot or blend is enabled and producing meaningful values on the chart.

2. Check whether the condition is currently being met by looking at the chart at bar close.

3. Remember that alerts fire only on confirmed bars - they will not trigger mid-bar.

Quick reference table

When multiple things seem wrong at once

If you are seeing several symptoms simultaneously - say, a blend that does not match the visible slots, flat channels, and an alert that is not firing - start by checking the most fundamental setting: is the slot enabled? Is On Bar Close on or off? Has the chart timeframe changed since you last configured the indicator?

Most compound issues trace back to a single configuration change that has cascading effects. Changing the chart timeframe, for example, can trigger a runtime error on one slot (if its TF is now lower than the chart's), produce na on another (if the new chart TF and lookback exceed available history), and make the blend appear empty or nonsense-sized because its inputs are missing, zeroed, or off-scale. Fix the root cause - usually the chart timeframe, slot state, or cross-ticker setup - and the downstream symptoms resolve.