Siemens Ultramat 23 NO Sensor Replacement & mA Output Trim

David Krause10 min read
Sensor IntegrationSiemensTroubleshooting
Licensed PE Working through this on a live machine? A Maine-licensed engineer can take it from here — included with IMD hardware, by the hour for everything else. Book an engineer

Siemens Ultramat 23 NO Sensor Replacement & Analog Output Calibration

The Siemens ULTRAMAT 23 is a multi-channel NDIR gas analyzer capable of measuring up to four gas components simultaneously, including up to three infrared-active gases (CO, CO2, NO, SO2, CH4) plus one electrochemical O2 channel. The platform is widely deployed in CEMS, process control, and combustion monitoring applications, where field service frequently involves sensor end-of-life replacement, jumper/revision migration, and 4–20 mA loop re-calibration. This technical reference addresses three recurring field problems encountered on the ULTRAMAT 23 platform: NO sensor part-number migration with jumper differences, SO2 channel reporting 0 ppm at site but generating sub-4 mA output, and analog output span error at the 4 mA and 20 mA endpoints.

Safety warning: ULTRAMAT 23 sensor modules contain dilute sulfuric acid electrolyte. Do not use metallic tools that can damage or pierce the cell when removing or installing a sensor. If electrolyte contacts skin, flush immediately with water and seek medical attention. Refer to the ULTRAMAT 23 operating manual (Siemens Support) Section 5 before performing any sensor replacement.

1. Problem Overview

Three independent issues are documented against a single ULTRAMAT 23 unit configured for four channels:

# Symptom Affected Component
1 NO channel not detected after sensor swap IR sensor module, motherboard connector X36
2 SO2 reading 0 ppm but mA output stuck at 1.7 mA SO2 IR channel + analog output stage
3 mA output out of spec at both ends (3.95 mA / 20.5 mA) Analog output DAC trim

Channel assignments on the affected motherboard:

  • Channel 1 (X35) — CO2
  • Channel 2 (X36) — NO
  • Channel 4 — O2 (electrochemical)
  • SO2 channel — typically occupies one of the remaining IR slots

2. NO Sensor Part-Number Migration: C79451-A3468-B22 → C79451-A3468-B522

Siemens periodically revises IR sensor modules without changing the base gas type. The part-number suffix indicates the hardware revision:

PN Suffix Hardware Revision Jumper Positions Jumper Installed Resistor Identification
C79451-A3468-B22 Original / older revision 3 positions 2 of 3 Discrete resistor on PCB for gas/range ID
C79451-A3468-B522 Current revision 2 positions Both installed Resistor retained from old module for backward compatibility

The two revisions are functionally interchangeable when the migration procedure below is followed. The differences are mechanical (jumper header count) and identification-related (resistor value must be transferred to maintain channel detection).

2.1 Root Cause of "Channel Not Detected" After Swap

The ULTRAMAT 23 motherboard reads the sensor identification through a combination of jumper positions and a precision resistor mounted on the sensor PCB. The resistor encodes gas type and measurement range to the analyzer's channel-sensing circuit. If the resistor is left on the old (B22) module or is omitted entirely on the new (B522) module, the analyzer will fail the ID handshake and report "channel not detected" even though the gas type is correct.

The CO2 and O2 channels continue to be detected because their sensors were not disturbed. Only the migrated NO module loses ID.

2.2 Required Migration Procedure

  1. Power down the ULTRAMAT 23 and isolate the analyzer from any sample gas. Allow the sensor chamber to vent.
  2. Remove the original C79451-A3468-B22 NO sensor from connector X36 on the motherboard.
  3. Desolder the identification resistor from the original B22 PCB. Verify the resistor is intact and not cracked (the original report indicated the resistor was still in good condition).
  4. On the new C79451-A3468-B522 sensor, install both jumpers across the available 2-position header.
  5. Solder the transferred resistor onto the equivalent identification pads on the B522 PCB.
  6. Seat the B522 module onto connector X36, confirm the gas-path seal is intact, and verify the optical bench is clean and dry.
  7. Power up the analyzer and observe the channel detection sequence in the diagnostic menu.
Do not use any tools (screwdrivers, picks, pliers) inside the sensor chamber that can scratch the optical surfaces or pierce the electrolyte reservoir. Refer to the ULTRAMAT 23 service manual for the approved extraction tool.

2.3 Verification After NO Sensor Migration

  • Channel 2 shows the correct gas identifier (NO) in the status screen.
  • Zero-gas calibration on the NO channel completes without an "identification error" or "sensor missing" flag.
  • Span-gas response on NO is within the tolerance band documented in the EPA ULTRAMAT 23 specification sheet.

3. SO2 Sub-4 mA Output at Zero ppm Reading

3.1 Symptom Description

  • Site displays SO2 = 0 ppm.
  • Loop mA measured at the analog output terminals = 1.7 mA.
  • Configured output range = 4–20 mA.
  • "Suppression of negative measured values" option is not available on this firmware revision.
  • Calibration mode (Configuration test) correctly drives the mA output to the commanded value; only measurement-mode output is wrong.

3.2 Root Cause Analysis

The mA output in measurement mode represents the live process measurement, not a clamped value. With a 4–20 mA span, the relationship between mA and measurement is linear:

I_out (mA) = 4 + 16 × (Measured − Zero) / (Span − Zero)

A measured output of 1.7 mA corresponds to:

(1.7 − 4) / 16 × 100 = −14.375 % of span below zero

This is the signature of a small negative drift in the SO2 measurement — typically −2 to −3 ppm of zero-point offset on a range such as 0–200 ppm. The analyzer is reporting 0 ppm on the front panel after rounding, but the underlying digital value is approximately −14 % of span. Because the firmware on this unit does not offer the "suppress negative measured values" toggle, the analog output reflects the negative value directly and falls below the 4 mA zero point.

mA Output Equivalent % of Span Interpretation
4.000 mA 0 % Zero point, no drift
3.200 mA −5 % Mild zero drift
1.700 mA −14.4 % Sub-zero drift (this case)
0.000 mA −25 % Hard under-range / fault

3.3 Solution Path

Since firmware-side negative suppression is unavailable on this unit, the corrective options are:

  1. Re-zero the SO2 channel with clean zero gas. Most field instances of 1–2 mA sub-zero drift are eliminated by a proper zero calibration using certified N2 or scrubbed air. Confirm zero-gas path is leak-free and the cell is fully equilibrated before accepting the new zero.
  2. Re-span after the zero adjust to confirm span response is unchanged.
  3. External loop conditioning (last resort): if a small persistent negative offset remains after zero/span, install an external 4–20 mA signal conditioner/clamp that limits the output to the 4–20 mA band. The ULTRAMAT 23 community of practice acknowledges this as a valid workaround when firmware-side suppression is not present, but it is inferior to a true in-instrument zero correction because it masks drift rather than correcting it.
The "Configuration test" mode forces the DAC to the commanded mA value without referencing the live measurement, which is why the test mA was correct while the measurement mA was not. This is normal behavior, not a hardware fault in the output stage.

3.4 Verification

  • After re-zero, output on certified zero gas reads 4.00 mA ± 0.02 mA.
  • After re-span with a known SO2 span gas, output reads within ±0.5 % of the expected mA value.
  • Front-panel SO2 ppm reading equals zero-gas concentration ± quantisation step.

4. Analog Output Span Trim at 4 mA and 20 mA Endpoints

4.1 Measured Error

Commanded mA Measured mA Error % of Span (16 mA)
4.000 3.95 −0.050 mA −0.31 %
20.000 20.50 +0.500 mA +3.13 %

The errors are not symmetric — the upper endpoint is approximately ten times worse than the lower endpoint — which indicates a DAC gain (span) error rather than a pure offset (zero) error.

4.2 Recommended Trim Procedure

  1. Enter the ULTRAMAT 23 service menu via the front-panel key sequence documented in the operating manual.
  2. Navigate to Analog output adjustment (menu path varies by firmware, commonly under Settings → Outputs → mA → Trim).
  3. Select the 4 mA endpoint trim and use the +/- adjustment to drive the loop mA to exactly 4.000 mA as measured by a calibrated mA reference (resolution 1 µA or better).
  4. Select the 20 mA endpoint trim and adjust to exactly 20.000 mA.
  5. Iterate between 4 mA and 20 mA at least twice because adjusting span can shift the zero point slightly.
  6. Save the trim values and exit service mode.

4.3 If the On-Board Trim Cannot Reach Specification

On some ULTRAMAT 23 hardware revisions the on-board DAC trim has a limited correction range. If the adjustment saturates before the error is removed, the residual error must be absorbed by one of the following:

  • An external 4–20 mA signal conditioner with user-adjustable zero and span trims.
  • Re-zero and re-span the gas measurement to compensate mathematically for the residual DAC offset (not recommended for high-accuracy CEMS applications).
  • Replacement of the analog output board if the DAC has drifted beyond the trim window.

4.4 Verification

  • Commanded 4 mA → measured 4.000 mA ± 0.010 mA.
  • Commanded 20 mA → measured 20.000 mA ± 0.020 mA.
  • Commanded 12 mA (mid-scale) → measured within ±0.050 mA of 12.000 mA.
  • Loop linearity check at 4 / 8 / 12 / 16 / 20 mA confirms monotonic response.

5. Diagnostic Cross-Reference Matrix

Observed Behavior Likely Root Cause Corrective Action Reference
NO channel not detected after B522 install ID resistor not transferred to B522 PCB Transfer ID resistor from old B22 sensor Section 2
CO2/O2 detected, NO missing Sensor ID resistor missing or wrong value Verify resistor on B522 pads, re-seat on X36 Section 2
SO2 = 0 ppm, mA = 1.7 mA Sub-zero measurement drift, no negative suppression Re-zero SO2 channel with N2/scrubbed air Section 3
Test mA correct, measurement mA wrong Live reading below zero, output reflects negative Re-zero, do not trust test mode for field loop check Section 3
4 mA command → 3.95 mA actual DAC zero error Use 4 mA trim in service menu Section 4
20 mA command → 20.5 mA actual DAC span error Use 20 mA trim in service menu Section 4
Asymmetric error at both ends Predominantly DAC gain error Adjust 20 mA first, then re-touch 4 mA Section 4
Trim saturates before spec DAC beyond trim window External conditioner or AOUT board replacement Section 4.3

6. Sensor Handling and Safety Notes

  • Sensors may contain dilute sulfuric acid electrolyte; PPE (chemical-resistant gloves, safety glasses) is required during swap.
  • Do not invert sensor modules with liquid electrolyte still inside.
  • Dispose of used sensors per local hazardous-waste regulations; do not drain electrolyte into any drain.
  • Observe the analyzer warm-up time (typically 30–60 minutes for IR channels) before evaluating drift or performing trim.
  • Always re-zero after a sensor swap before re-spanning.

7. Spare-Part and Revision Reference

Gas Legacy PN (suffix) Current PN (suffix) Connector
NO C79451-A3468-B22 C79451-A3468-B522 X36 (Ch 2 in this unit)
CO2 C79451-A3xxx-Bxx C79451-A3xxx-B5xx X35 (Ch 1)
SO2 C79451-A3xxx-Bxx C79451-A3xxx-B5xx Per analyzer config
O2 (electrochemical) — Per channel Ch 4 in this unit

Confirm the exact current PN against your Siemens spare-parts catalog or the ULTRAMAT 23 spare parts list before ordering.

8. Related Documentation

FAQ

Are the C79451-A3468-B22 and C79451-A3468-B522 NO sensors directly interchangeable in the ULTRAMAT 23?

Yes, they measure the same gas (NO) and are functionally interchangeable once the identification resistor from the old B22 sensor is transferred to the new B522 PCB and both jumpers on the 2-position header are installed. Without the resistor transfer, the motherboard will not detect the NO channel.

Why does my ULTRAMAT 23 SO2 channel read 0 ppm but output 1.7 mA on a 4–20 mA loop?

A 1.7 mA output on a 4–20 mA scale equals roughly −14 % of span, indicating a small negative zero drift. Because this firmware revision does not expose "suppression of negative measured values," the analog output reflects the negative digital reading and falls below 4 mA. Re-zero the SO2 channel with N2 or scrubbed air to eliminate the offset.

How do I trim the 4–20 mA output on the ULTRAMAT 23?

Enter the service menu, navigate to the analog output adjustment (typically under Settings → Outputs → mA → Trim), and use the 4 mA and 20 mA endpoint adjustments with a calibrated mA reference on the loop. Always iterate between the two endpoints at least twice because span adjustment can shift the zero point.

What causes "channel not detected" after a sensor swap on the ULTRAMAT 23?

The motherboard identifies each sensor through a combination of jumper configuration and an identification resistor on the sensor PCB. If the resistor is not present on the replacement sensor, the ID handshake fails and the channel is reported as missing. Verify resistor presence on every migrated sensor module.

Is it safe to use metal tools inside the ULTRAMAT 23 sensor chamber?

No. The sensor modules contain dilute sulfuric acid electrolyte and have sharp internal edges. Use only the approved plastic extraction tool specified in the ULTRAMAT 23 operating manual, and wear chemical-resistant gloves and eye protection during sensor handling.

Back to blog