Troubleshooting

Symptoms are easier to notice than causes. This page walks the surprises this pane produces and names the likely cause for each, a fast check you can run, and the fix. The sections are grouped so you learn to triage rather than memorize fixes. Most issues are misreads. Some are setup errors. A few are honest limits.

A note before you start

When the pane does something you did not expect, resist the urge to change a dial immediately. The single most common way a minor oddity on this pane becomes a major misconfiguration is a reader who "fixes" a symptom by pushing sensitivity, shortening a length pair, or flipping On Bar Close? off — all in the first five minutes, all without a written account of what changed or why. Three symptoms later, the original issue is still present, two new artifacts have joined it, and nobody can remember the baseline.

Most symptoms on this pane are not bugs. They are warm-up artifacts, chart-timeframe collapse, weight-zero observers behaving exactly like weight-zero observers behave, or the three slots actually disagreeing with each other in a way that is inconvenient rather than broken. The triage habit at the bottom of this page names the four questions to ask before reaching for a dial.

Group 1 — Setup errors (the script is telling you something specific)

Symptom: "MACD 0N timeframe cannot be lower than the chart timeframe"

A runtime error, with the offending slot named.

• Cause. The slot's TimeFrame input is set to a value lower than the chart's timeframe. Common triggers: you moved the chart to a higher timeframe, or you typed a slot timeframe that ended up below the chart (for example, 5 on a 15-minute chart).

• Check. Look at the offending slot's TimeFrame input and your chart timeframe.

• Fix. Either raise the slot's timeframe to match or exceed the chart, drop the chart to a timeframe below the slot, or disable the slot.

• Next step. If you hit this repeatedly, your workflow is chart-timeframe-dependent in ways your slot configuration does not reflect. A single chart timeframe plus a single slot ladder above it is the shape the tool was designed around.

Symptom: "MACD 0N Fast Length must be less than Slow Length"

A runtime error, with the offending slot named.

• Cause. The slot's Fast Length is greater than or equal to its Slow Length.

• Check. Compare Fast Length and Slow Length on the named slot.

• Fix. Set Fast Length strictly below Slow Length. The textbook 12/26 is a baseline; any pair where fast is below slow is valid.

• Next step. If you wanted a non-textbook MACD shape where fast equals slow or fast exceeds slow, this indicator is not the tool. The script enforces the rule because a MACD is, by definition, fast minus slow.

Symptom: The indicator loads but nothing plots

• Cause — possibility A. Every slot is disabled. Look at the three Enable MACD 0N inputs.

• Cause — possibility B. Every enabled slot has weight 0 and the slots are also hidden, or the slots are still warming up so their lines are na. Either way, the blend has no data and the slots have no lines.

• Cause — possibility C. Plot Blended K/D and Plot Blended Histogram are both off and every slot has Hide MACD 0N Plot on.

• Check. Scan the Enable flags, the weights, the Hide Plot flags, and the Display group.

• Fix. Depends on the cause. Enable at least one slot. Give at least one enabled slot non-zero weight and a visible plot. Turn on the blend display toggles if you want the blend drawn.

Group 2 — Misreads (the pane is behaving correctly, you are reading it wrong)

Symptom: A slot is pinned at 100 or 0 for bars at a time

• Cause. The ATR Length / ATR Sensitivity pairing on that slot's timeframe has pushed the sigmoid into its flat region. This is not intensity; it is saturation.

• Check. Are only one or two slots pinned while others breathe? Is the raw chart action on the slot's timeframe outsized, or average? If only one slot is pinned and the chart looks average, you are looking at saturation.

• Fix. Two honest options.

• Lengthen ATR Length to make the volatility yardstick steadier.

• Lower ATR Sensitivity modestly to pull the flat region outward.

• Wrong fix. Raising sensitivity because the pane looks "stuck at the top." This makes the flat region wider and produces more pinning, not less.

Symptom: The blend sits near 50 while the slot lines look active

• Cause. The slots are disagreeing. A bullish slot and a bearish slot with equal weights produce a weighted average near the midline.

• Check. Are the three slot lines on different sides of their own signals? Is one in full tone and another faded?

• Fix. This is not a bug. The blend is doing what a weighted average does. The reading is "the three timeframes disagree on direction." That is often a more informative reading than a clean three-way agreement would be, because it names the disagreement instead of averaging over it.

• Next step. If you want the blend to favor one timeframe's story more, adjust the weights deliberately. But do not set weights to hide disagreement; set them to reflect which timeframe you trust more for the decision you are making.

Symptom: Alignment alerts fire on every bar during a trend

• Cause. Your three slots are probably near-identical (same length pair, same MA family, different only by timeframe). Alignment across similar slots fires continuously during any trending session.

• Check. Compare the length pairs, the MA types, and the sources across slots 01, 02, 03. If they match on most or all of those, you have the configuration that produces continuous alignment.

• Fix. Differentiate the slots. Workflows walks a genuinely differentiated configuration in Scenario 1.

• Next step. If you wanted continuous alignment alerts during trending sessions and treat them as a persistence indicator rather than a breadth indicator, the current configuration is doing what it will do. If you are using them as a breadth/consensus indicator, Limitations and Trust Boundaries §1 is the section to read.

Symptom: The histogram just turned faded green and I thought that meant "still up"

• Cause. You are reading a two-dimensional color code as one-dimensional.

• Check. Is the column above 50? Is it smaller than the previous column?

• Fix. Faded green means above 50 and falling — bullish but decelerating. It is the regime-warning state on the bullish side. Treat a faded-green stretch the same way you would treat faded red on the bearish side: as evidence that thrust is leaving the current direction.

• Next step. The four-state table on Visuals and Logic is worth pinning somewhere. The table is the antidote to this exact misread.

Symptom: The 60-minute slot "never moves"

• Cause. With On Bar Close? on, the 60-minute slot updates once per 60-minute bar close. Between those closes, the value is the previous 60-minute bar's confirmed reading.

• Check. Note the chart time and wait until the next top of the hour. The slot should step to a new value on the 60-minute close.

• Fix. Not a bug. If you want the 60-minute slot to update more often, you are considering the repaint trade. Read MTF and Repainting before flipping the switch.

Symptom: The three slot lines all look glued together

• Cause — possibility A. Chart-timeframe collapse. Your chart is at or above your slowest slot timeframe. MTF and Repainting walks this in full.

• Cause — possibility B. Your three slots are near-identical. See the alignment-alerts symptom above.

• Check. Compare the chart timeframe to the slowest slot timeframe. Compare the three slots' length pairs and MA families.

• Fix. For cause A, lower the chart or widen the slot ladder. For cause B, differentiate the slots.

Symptom: An alert fired but the chart looked unremarkable

• Cause. The alert is a state alert, not a transition. A per-slot bull/bear alert can fire on every confirmed chart bar during a held state.

• Check. What was the state on the prior bar? If the state held, the alert is firing because the condition is still true, not because something new happened.

• Fix. Decide whether you want repeated state firings or transition-like behavior. The TradingView alert-frequency setting ("Only once" vs "Once per bar close") changes the notification cadence; genuine transition detection needs external wiring.

Symptom: A slot line is colored before its signal state seems ready

• Cause. First-bar fallback. The indicator does not plot a separate per-slot signal line, but it does compute one internally. Before that internal signal value is available, the slot colors itself by whether its normalized MACD value sits above or below 50 instead of above or below its own signal.

• Check. Give the slot enough bars on its own timeframe for the MACD and signal calculations to form. On the default 60-minute slot, that can mean many hours of session history, not a few minutes.

• Fix. Nothing. This is a first-bar nicety so the line does not sit colorless. Do not try to "fix" the color by changing sensitivity or length; the fallback resolves itself as soon as the internal signal value is available.

Group 3 — Genuine limits (the tool has reached an honest edge)

Symptom: I want an alert on a crossover and there isn't one

• Cause. No crossover or crossunder alerts ship. Every alert in the script is a state alert.

• Check. Alerts enumerates the ten conditions and names the absent transition alerts explicitly.

• Fix. Build the transition externally. Wire alerts on the states and handle the change in your downstream system, or use TradingView's native alert-on-condition-change on the plotted series.

Symptom: I want an alert when a slot crosses 70 or 30

• Cause. The reference lines are visual zones. No alerts fire on them in this indicator.

• Check. Alerts documents this design choice.

• Fix. Wire a TradingView threshold alert against the plotted slot series manually.

Symptom: The blend timing changed after I enabled master smoothing

• Cause. Working as intended. Master smoothing changes the blend by applying one more MA pass after the weighted blend. Blend-based alerts now fire from that smoothed series, so their timing can move and short-lived raw blend states may disappear.

• Check. Compare the blend-fire timing against the per-slot alert timing. The per-slot alerts are untouched by master smoothing; only the blend is smoothed.

• Fix. If you wanted the raw blend timing, turn master smoothing off. If you wanted the slower regime view, accept the changed timing and read the slot alerts separately when you need earlier evidence.

Symptom: I want per-slot repaint control (some slots on close, others live)

• Cause. Base exposes one global repaint switch. Per-slot repaint control lives on the CTX and STR versions.

• Fix. If per-slot control is a workflow requirement, CTX is the context-breadth version to evaluate; STR carries that control too, then trades five CTX slots for blended-line structure. This is a version-scope difference, not a bug.

Symptom: I expected HMA, DEMA, TEMA, ALMA, Jurik, KAMA, FRAMA, Laguerre, or VAMA in the MA dropdown

• Cause. Base is keyed to the Lite moving-average library. The imported local library exposes SMA, EMA, RMA, WMA, VWMA, and SWMA only.

• Fix. Use one of the Lite families in this trim. If a workflow truly depends on another MA family or family-specific parameters, that is a different-trim or different-tool requirement, not a hidden setting in Base.

Symptom: I wanted the raw MACD magnitude and the pane shows 0-to-100

• Cause. The pane is a normalized convergence/divergence read by design. Raw MACD magnitude is lost by construction in exchange for cross-timeframe comparability.

• Fix. If raw magnitudes are a workflow requirement — for classical divergence counting, for example — a classical MACD indicator is the right tool and this one is not. The two can live on the same chart without conflict.

Triage habit

When you hit a symptom, ask these four questions in order:

1. Is the script telling me something (a named runtime error)? If yes, you are in Group 1.

2. Am I reading the pane correctly (histogram first, slots second, blend last; four-state color; reference lines as zones not triggers)? If no, you are in Group 2.

3. Is the configuration still coherent with the chart timeframe (slot ladder above the chart, slots differentiated, weights reflecting intent)? If no, you are in Group 2 or the chart-timeframe collapse subsection of MTF and Repainting.

4. Is what I want from the tool something it does not do (transitions, threshold alerts, per-slot repaint, Pro-library families)? If yes, you are in Group 3 and the answer is a version-fit question, a TradingView-side wiring question, or a "different tool" question, not a tuning question on this indicator.

Where to go next

• Knobs and the tradeoffs attached to each — Settings.

• Pane elements and the reading order that prevents most misreads — Visuals and Logic.

• The alignment trap explained at length — Limitations and Trust Boundaries.

• The repaint switch, the slot-timeframe rule, and the chart-timeframe collapse case — MTF and Repainting.