Troubleshooting

This page is for when something on the chart is not doing what you expect. It maps the symptom you are actually seeing to the most likely cause and to the fix. It also separates two very different buckets that readers...

Written By Axiom Admin

Last updated 22 days ago

Troubleshooting

This page is for when something on the chart is not doing what you expect. It maps the symptom you are actually seeing to the most likely cause and to the fix. It also separates two very different buckets that readers tend to lump together: "you configured something unexpected" and "the tool is operating as documented and your expectation was wrong." Both buckets are real. Neither one means the tool is broken.

Read this page as you would a field guide. Scan the table, find the symptom that matches, follow the row to the likely cause and the fix. If your symptom is not in the table, the bottom of the page has a quick general diagnostic procedure.

Symptom β†’ likely cause β†’ fix

#

Symptom

Likely cause

Fix

1

The chart is blank and a red error strip at the top reads MA NN timeframe cannot be lower than the chart timeframe.

A slot's TimeFrame: is strictly below the chart's timeframe. The tool has triggered its runtime-error guard naming the offending slot (MA 01, MA 02, or MA 03). This is the tool operating as documented.

Open the named slot's group in the inputs dialog. Raise the slot's TimeFrame: to the chart's timeframe or above. Click OK. The chart returns. See MTF & Repainting for why the guard exists.

2

A slot's line is drifting mid-bar on its higher timeframe β€” for example, Slot 03's purple line is moving inside a 60-minute bar instead of stepping only at the 60-minute close.

On Bar Close? is OFF, which changes every slot's posture to live-HTF-bar reads. This is the documented OFF behaviour; the drift is the cost of responsiveness.

If the drift is not what you want, flip On Bar Close? back to ON in the PU Settings group. If you specifically need live Slot 03 while keeping Slot 01 and Slot 02 confirmed, Base does not support mixed posture β€” route to Axiom MA CTX.

3

The blended line is not moving even when the slot lines are clearly moving.

One of: (a) every enabled slot has Blended Weight: at 0; (b) Enable Blended MA is off; (c) Hide Blended MA Plot is on, making the blend invisible rather than absent; (d) unusual negative weights leave no usable positive total weight.

In the inputs dialog: confirm Enable Blended MA is on; confirm Hide Blended MA Plot is off; confirm at least one enabled slot has a positive Blended Weight: and that the total voting weight is positive. If the blend is genuinely stuck because no slot is weighted, set at least one slot's weight above zero. See Settings.

4

The blended line is moving even though none of the visible slot lines appear to move.

A slot is enabled with Hide MA NN Plot on and a non-zero Blended Weight:. The hidden slot is steering the blend while not drawing on the chart. This is by design and is documented as the "visibility-only hidden slot" workflow in Workflows.

If you want the blend to stop reflecting the hidden slot, set that slot's Blended Weight: to 0, or toggle its Hide MA NN Plot off to make it visible. The hidden-plot mechanic is also covered in Visuals & Logic.

5

The All MA Slots Uptrend alert is silent even though two of the three visible slot lines are lime.

The usual cause is a hidden enabled slot with a non-na MA value that is reporting downtrend. Alignment counts enabled slots with non-na MA values, including hidden ones. A slot that is still na is not counted yet; it lowers the active count rather than blocking alignment by itself.

Open the Data Window (Alt+D) on the bar in question. Check the values of MA 01, MA 02, MA 03, Active Uptrend Count, and Active Downtrend Count. If a hidden valid slot is disagreeing, disable it (to remove it from alignment) or accept that it is part of the alignment vote. If one slot is na, read the active count you actually have and wait for enough history. See Alerts.

6

An enabled slot's line is missing from the chart entirely.

One of: (a) Hide MA NN Plot is on; (b) the slot's Source: or Type: produces na on the current chart range; (c) the slot's Length: is greater than the available history and the MA has not yet produced a value; (d) the slot is running off-chart because its TimeFrame: is very high on a very short history.

Toggle Hide MA NN Plot to off. If the line is still absent, scroll the chart back far enough for the slot to accumulate enough HTF bars for Length: to produce a value. Consider shortening Length: if the chart history is thin on the slot's timeframe.

7

A slot's colour is flipping several times inside a single chart bar on a slow higher timeframe, or several times across adjacent chart bars.

Trend Length: on that slot is very small (for example, 1 or 2) relative to the slot's natural noise, and the MA is crossing its own recent past value repeatedly. The flicker is computationally real and documented.

Lengthen Trend Length: on the slot until the flicker matches the cadence you want the slot to report. Remember that Trend Length: is the rising-versus-falling comparison, not the smoothing window β€” see the banner at the top of Settings.

8

The blended line looks almost flat but the Blended MA Is Uptrend alert is still firing on every bar.

The blended colour and alert are decided by the weight-majority vote, not by the line's slope. A flat-looking blended line can report uptrend if the up-weighted slots outvote the down-weighted ones β€” and because ties resolve to uptrend, a tied vote also reads uptrend.

Confirm per-slot trend states in the Data Window. If the vote is resolving as documented, the alert is operating correctly; update your mental model of what the alert confirms. See Alerts and Visuals & Logic.

9

A per-slot MA NN Is Uptrend alert fired, but by the time you looked at the chart the slot's colour was already a different state.

The alert fired on a closed chart bar when the slot's trend was up. Since then, either a subsequent closed chart bar flipped the slot's trend, or (under On Bar Close? OFF) the slot's live HTF read drifted. The alert is a snapshot, not a live readout.

Open the bar where the alert fired in the Data Window. Confirm the slot's trend state at that bar. Alerts report the state at firing time and are not retroactively amended. See Alerts.

10

Changing Line Width: on a slot changed how the chart looks but did not change the line's trend colour or position.

Line Width: is cosmetic. It does not affect computation. This is the documented behaviour.

No fix needed. If you meant to change the trend colour, adjust Trend Length: (slower or faster comparison). If you meant to change the line's position, adjust Source:, Length:, or Type:.

11

You expected to find per-slot On Bar Close?, a Symbol: input for cross-ticker reads, or Up Color: / Down Color: inputs on a slot, but there is no such input.

Those features are not in the Base trim. They live in the CTX trim of the Axiom MA family.

Route to Axiom MA CTX. The Base pack's README.md and Settings document the trim-ladder decision.

12

Alerts fire far more often than you expected, or on every closed chart bar regardless of a state change.

Per-slot and blended trend alerts fire every closed chart bar while the condition holds, not only when the state changes. The tool does not ship state-change alerts.

Wire the alert in TradingView with Once per bar close if that reduces the volume to your taste. If you need "state-change only" semantics, use a separate companion indicator or external alert handling that compares the current bar's trend state to the prior bar's.

13

A slot whose Trend Length: is set to 0 is drawing in the down-state colour even when price and the MA are clearly rising.

Trend Length: of 0 disables the trend read. The slot's computed trend returns false, which drives the line's colour to the down-state variant regardless of the MA's actual behaviour.

Set Trend Length: to a positive integer. If you want a very fast read, start from 1 and lengthen until the flicker matches your cadence. See Settings.

14

Two slots look like they are at the same price level but their colours disagree.

The two slots are computing independent MAs on independent timeframes with independent Trend Length values. Their current prices can coincide even when their own self-comparisons against their Trend Length past values resolve in opposite directions.

No fix needed. Disagreement between slots at a shared price level is coherent and is information. See Visuals & Logic on reading disagreement.

"You configured something unexpected" versus "the tool is operating as documented"

A large fraction of the symptoms that readers report fall cleanly into one of these two buckets. Telling them apart saves you from blaming the tool for something it is telling you honestly.

You configured something unexpected β€” fix the configuration

  • The runtime error (symptom 1). Fix the slot's TimeFrame:.

  • Blend not moving because no slot has a non-zero weight (symptom 3). Fix the weights.

  • Slot missing because Hide MA NN Plot is on (symptom 6). Unhide.

  • Trend colour inverted because Trend Length: is 0 (symptom 13). Fix the trend length.

  • Trend flicker from a too-small Trend Length: (symptom 7). Raise it.

The tool is operating as documented β€” update the expectation

  • OFF posture drifting inside the HTF bar (symptom 2). The posture is doing what it says. Flip it ON or move to CTX if you need mixed posture.

  • Blend moving because a hidden slot is weighted (symptom 4). Hidden is visual-only. Fix with weight, not with hide.

  • Alignment silent because a hidden valid slot disagrees (symptom 5). Alignment counts enabled slots with non-na MA values; na slots are not in the active set yet. Adjust the configuration if the expectation needs to shift.

  • Blend colour not tracking blend slope (symptom 8). Colour is a weighted vote, not a slope read. Update the mental model.

  • Alerts firing on every closed chart bar (symptom 12). The alerts report a state. They do not ship as state-change events.

  • Disagreement between slots at a shared price (symptom 14). Two independent self-comparisons can coincide in price and diverge in direction.

The first bucket asks for a keystroke. The second bucket asks for a paragraph from the relevant page in this pack.

A quick general diagnostic procedure

When the symptom is not in the table:

  1. Open the Data Window (Alt+D). Hover the bar in question. Confirm the values of MA 01, MA 02, MA 03, MA Blended, Active Uptrend Count, and Active Downtrend Count. These are the numeric truth of the chart at that bar.

  2. Re-read the inputs. In the inputs dialog, confirm every slot's Enable, Hide Plot, Source:, TimeFrame:, Length:, Type:, Trend Length:, Line Width:, and Blended Weight:. Confirm the MA Blended group. Confirm On Bar Close? in PU Settings.

  3. Compare configuration to chart. Does the slot's value at the bar match what you would expect from its TimeFrame:, Length:, Type:, and Source:? If the chart timeframe is below the slot's, the slot steps; if the chart timeframe equals the slot's, the slot is effectively the native MA on the chart's own bars.

  4. Check the trust boundary. If the read seems off but the configuration checks out, the likely cause is a state like "hidden slot contributing," "weight-majority vote carrying the blend," or "OFF posture reading live HTF." Limitations & Trust Boundaries has the full enumeration.

  5. If all of the above check out and the tool is still producing unexpected numbers, capture the chart state (timeframe, ticker, inputs, the bar in question, the values from the Data Window) and escalate through Axiom Charts support as indicated on the site.

A short note on reporting something that still looks wrong

If you have worked through steps 1–4 and the tool still appears to be producing a number or a state that contradicts its documentation, the most useful report includes:

  • The ticker and chart timeframe.

  • The exact values of every slot's inputs (defaults plus any deviations).

  • The state of On Bar Close?.

  • The values from the Data Window at the specific bar in question.

  • A description of what you expected and what you saw.

That is the shape of a report someone can actually reproduce. Vague reports take longer to diagnose and often resolve to one of the symptom rows above.

Where to go next

  • MTF & Repainting β€” if the symptom is about the global repaint switch or the runtime-error guard.

  • Visuals & Logic β€” if the symptom is about what a line means rather than how a line is configured.

  • Alerts β€” if the symptom is about an alert that did or did not fire.

  • Limitations & Trust Boundaries β€” if the symptom has resolved into "the tool is honest and I was asking the wrong question."

  • Axiom MA CTX β€” if the symptom is the absence of a feature (per-slot repaint, cross-ticker, colour inputs, wider palette) that Base does not ship.