IMD SolutionComputing LibraryShop main store

Troubleshooting

Historian tag scaling for an unscaled PLC integer value

Edge computing and IIoT Published 7 min read

Short answer

Scale the tenths-encoded integer on the historian tag itself: linear scaling with gain 0.1 and offset 0 turns raw 144 into 14.4 while keeping the direct PLC acquisition. Check the tag's existing scaling first so a second divide-by-ten does not log 1.44, and repoint the tag at a SCADA-supplied value only if logging that stops with that runtime is acceptable.

On this page
  1. Why does the historian log 144 when the screen shows 14.4?
  2. Where should the divide-by-ten go: historian, SCADA or PLC?
  3. What scaling values fix a tenths-encoded integer tag?
  4. How do you configure the tag without breaking logging?
  5. Is it a scaling error or the wrong tag binding?
  6. What happens to history already stored at the wrong scale?
  7. Questions asked

A historian that stores 144 while every screen shows 14.4 is not a database fault: the tag acquires the PLC's tenths-encoded INT without the divide-by-ten the display applies. Fix it on the directly acquired historian tag — enable linear scaling so the value comes back as raw × 0.1 — rather than repointing the tag at a SCADA runtime and losing logging whenever that runtime closes. Before editing, read the tag's existing scaling properties, because a second divide-by-ten turns 144 into 1.44. The sections below separate the screen path from the history path, give the range values to enter, walk the configuration and verification steps, and state what happens to history already stored at the wrong scale. They assume access to SCADA and the historian, none to the controllers.

Why does the historian log 144 when the screen shows 14.4? #

Because the historian tag reads the master PLC's INT directly, it stores the encoded 144, while the SCADA screen divides the same integer by ten in its display expression and shows the operator 14.4 ft — that screen calculation never enters the historian's acquisition path.

Two independent paths carry one measurement. The operator path runs transmitter → transmitting PLC multiplies by 10 → INT to the master PLC → SCADA input tag → divide-by-ten display expression → 14.4 ft on screen. The history path runs master PLC INT → I/O server → historian tag → archive. A displayed number can be an expression rather than the stored value of its underlying tag: the historian sees the integer that entered the I/O server, and nothing the screen does changes that number.

In AVEVA Historian, each tag's definition records the I/O server that supplies its values, so the tag details show immediately whether the source is the controller item or a SCADA-supplied attribute. That is the first thing to confirm, because the fix differs: a controller-sourced tag is corrected with scaling on the tag itself, while a SCADA-sourced tag raises the runtime-dependency question in the next section. The unit label is a separate matter: whether the transmitter reports tonnes or feet is settled against the transmitter configuration, and the divide-by-ten correction works the same either way.

Where should the divide-by-ten go: historian, SCADA or PLC? #

Scale it on the directly acquired historian tag: linear scaling with gain 0.1 corrects the value while keeping the existing controller acquisition, and logging never depends on a SCADA runtime staying open — the one option you can also execute with access only to SCADA and historian.

Option Conversion runs in Effect on logging Decision
Linear scaling on the historian tag Analog-tag scaling properties Keeps the direct controller acquisition The fix for this installation
Historize a scaled SCADA tag SCADA expression, then collection Works only while the supplying runtime stays open Acceptable only if the scaled tag already exists and that runtime is always on
Repoint the tag at the SCADA value SCADA supplies the scaled number Logging stops when the runtime closes Only where that dependency is acceptable
Change the PLC encoding Controller logic or transmitted format Changes the source contract for every reader Cleanest long term, but it needs controller access you do not have
Convert in queries or at the display Retrieval expression or report Archive still holds 144 Fine for one report; it never repairs the history

Historian-side scaling wins here because it corrects the tag that carries the error and leaves acquisition untouched. Its one prerequisite is the check in the configuration steps below: if the tag already converts something, your divide-by-ten stacks a second stage, which is exactly the failure the checklist catches before you edit.

What scaling values fix a tenths-encoded integer tag? #

Set a linear map with gain 0.1 and offset 0: engineering value = raw integer × 0.1, so raw 144 comes back as 14.4 and 0 stays 0 — in range terms, the engineering limits are simply the raw limits divided by ten.

This is fixed-point decoding, not a unit conversion. The PLC multiplies the level by 10 so that a one-decimal measurement fits in an INT; the historian must reverse exactly that encoding. Setting the tag's engineering-unit string does not change 144 into 14.4 — only the scaling does.

AVEVA's Historian documentation describes the linear conversion for analog tags as a retrieval-time calculation of the form retrieved = MinEU + (value - MinRaw) × (MaxEU - MinEU) / (MaxRaw - MinRaw). Gain is (MaxEU - MinEU) / (MaxRaw - MinRaw); when the engineering limits equal the raw limits divided by ten, gain is 0.1 and offset is 0.

Worked example, with illustrative ranges — enter your transmitter's real span. A transmitter spanning 0 to 200.0 sends 0 to 2000 counts through the PLC. Enter raw 0 to 2000 and engineering 0.0 to 200.0. Gain = (200.0 - 0.0) / (2000 - 0) = 0.1, offset 0. The raw reading 144 then retrieves as 0.0 + (144 - 0) × (200.0 - 0.0) / (2000 - 0) = 144 × 0.1 = 14.4. Never enter the observed reading 144 as a range limit: it is one sample, not the transmitter's full scale. Other encodings follow the same rule — a ×100 integer keeping two decimals takes gain 0.01 and engineering limits equal to the raw limits divided by 100.

How do you configure the tag without breaking logging? #

Record the tag's current configuration and a short window of timestamped values before you edit, then change only the analog scaling properties on the acquired tag — one verification and one edit, in that order, is what keeps a scaling fix from becoming a data gap.

  1. Capture the before-state: the source I/O server item, the data type, the engineering-unit string, and any existing raw and engineering limits, plus a few timestamped values for reference.
  2. Open the tag's definition in the historian management console — in AVEVA Historian: Historian tab → Tag Management group → Tag Configuration → Tag Details, then select the server and the tag.
  3. Compare the two paths at one timestamp: historian value 144 against the screen's 14.4 ft confirms the encode-and-divide pair on the same transmitter.
  4. Check for existing scaling. If the raw and engineering limits already differ, the tag already converts something — resolve what is there before adding a second divide-by-ten.
  5. Enter the linear ranges: raw limits from the transmitter's span in counts, engineering limits equal to those raw limits divided by ten, offset 0.
  6. Verify after the edit: new values must come back as 14.4 while the screen shows 14.4 ft, and a timestamp from before the edit must come back as 14.4 as well. If the trend client instead shows gaps or refuses to plot the tag, that is a retrieval fault, not a scaling one — the troubleshooting historian trend client connection failures page covers those symptoms.

Is it a scaling error or the wrong tag binding? #

A steady ten-to-one gap between screen and history is scaling, but confirm both paths read the same transmitter first: match the acquisition item in the historian tag details against the display binding before you trust the ratio as the whole diagnosis.

Symptom Mechanism Check that separates the causes
Screen 14.4, history 144 Display-only divide-by-ten; acquisition carries the encoded integer Compare the screen expression, the live SCADA tag value and the historian source item
History shows 144 with the right unit label Unit string set, numeric scale untouched Read the scaling properties separately from the unit string
Corrected tag now shows 1.44 Two stacked divide-by-ten stages Look for existing scaling before applying yours
SCADA-sourced history tag stops updating Supplying runtime closed Compare last timestamps against the runtime stop

Two tags can hold plausible numbers while reading different instruments. Matching the device item behind the historian tag to the item behind the display binding rules that out in one look; only then does the ratio become a diagnosis.

What happens to history already stored at the wrong scale? #

In AVEVA Historian, linear scaling for analog tags is a retrieval-time conversion, so once the tag's raw and engineering ranges are set, history written before the fix comes back as 14.4 as well — no rewrite is needed on this installation.

The stored counts were always the transmitter's true tenths; the retrieval formula now maps them once, and every client reading through the tag scaling — trends, reports, exports that respect it — gets the correction automatically. A raw export that bypasses the tag scaling still returns 144, so write anything permanent against the scaled retrieval.

The other model exists: a historian that stores post-conversion values, or a tag fed by an already-scaled SCADA value, keeps the pre-fix span at 144 permanently. There the decision is overwrite or annotate — rewrite the affected span by dividing the stored values by ten with the historian's data-repair or import tools, or leave it and mark the discontinuity for anyone trending across the change. Verify which model you have with one retrieval: after setting the ranges, pull a timestamp from before the fix — 14.4 ends the job, and 144 means your installation stores scaled values and needs the overwrite.

Questions asked

Can the UOM function fix the scaling instead?

No. A retrieval-time unit-conversion call such as UOM() in AVEVA Historian changes the unit a query or report presents; it does not reverse a tenths encoding, and the trend still comes back at 144. Set linear scaling with gain 0.1 on the tag for the correction, and keep UOM() for genuine unit changes, such as showing the same level in another unit on a report.

What gain applies when the PLC keeps two decimal places?

Gain 0.01, with offset still 0. A ×100 encoding preserves hundredths in an integer, so the engineering limits equal the raw limits divided by 100: raw 1442 then retrieves as 14.42. The rule generalizes — gain is one divided by the encoding factor — but confirm the factor from the PLC logic or the SCADA expression before entering the ranges, because a mismatched factor silently shifts every value.

Would it be better to remove the multiply-by-ten in the PLC?

Yes, in principle: transmitting engineering units or a floating-point value removes the local correction every reader must remember. It needs controller access, a change to the master-to-remote mapping, and coordination with every existing consumer of the INT, so it is a project rather than a fix. With access only to SCADA and historian, tag-level linear scaling is the correct and complete correction.

Does editing the tag scaling interrupt data collection?

No. Scaling properties change how values are interpreted, not whether acquisition runs, and the historian keeps collecting from the same I/O server item through the edit. Make the change outside a critical batch window anyway and check the trend immediately afterwards, so an accidental edit to the source item or the ranges shows up in minutes rather than at month-end.

Keep reading

  • Troubleshooting

    Fixing a historian device shortcut that will not appear

    A ping, an open web page and an online programming session prove IP reachability, not a working data path. When the historian drops a device shortcut, the RSLinx Enterprise tree usually holds the gateway interface instead of the controller behind it. Stop browsing and the related services, delete the interface entry, define the SLC 5/03 manually under ENI Devices at the same IP address, point a fresh shortcut at the controller icon, restart, and confirm with a live item read.

    Edge computing and IIoT

  • Selection guide

    Plant historian selection by scale and support needs

    Pick the tier first: a control-room or automation-vendor historian for one site's process trends, an enterprise platform only when several sites share one operational history, and a self-built time-series stack only when your team owns connectors, dashboards, backups and upgrades. Size by stored events per second — the sum of each tag group's tags times its stored events per second — times retention, never by tag count alone.

    Edge computing and IIoT

  • How-to

    Showing a historian maximum on an operator text field

    Bind the text field to a History Summary custom property on the Industrial Graphic: set the attribute, choose Max, and set the end boundary to the current time with the start one hour behind it. The property queries the historian with no scripting. Attribute Statistics is the wrong tool here — it keeps running figures over the last 60 values, not over an hour.

    Edge computing and IIoT