ULTRAMAT 23 CO2 Detector Replacement: Resolving IR Channel Fault and Phase Not Found
The Siemens ULTRAMAT 23 is a multi-component NDIR gas analyzer capable of measuring up to three infrared-active components (CO, CO2, NO, SO2, CH4, etc.) plus a fourth electrochemical or paramagnetic O2 channel. Because the IR measurement principle relies on a tuned optical bench, replacing an IR receiving/detector chamber is not a drop-in operation when the spare part is delivered without factory calibration data for the host analyzer. Two of the most common post-replacement faults are "Fault of addresses for IR channels" and "Phase not found". This article documents the field-proven resolution procedure for the C79451-A3468-B26 (original) versus C79451-A3468-B526 (spare part) detector mismatch, with verified jumper, resistor, and motherboard header settings.
1. Problem Description
An end user is performing routine calibration on a ULTRAMAT 23 analyzer that measures CO2. During the calibration sequence, the CO2 measured value drifts away from the expected test-gas concentration. The technician suspects a degraded IR receiving chamber (the photometric detector block tuned to the 4.26 µm CO2 absorption band) and orders a replacement.
The replacement chamber arrives with Siemens part number C79451-A3468-B526. The chamber that was originally fitted in the analyzer is C79451-A3468-B26. After physical swap, the analyzer raises two simultaneous fault indications:
- Fault of addresses for IR channels — the analyzer cannot identify which physical IR channel the new detector occupies on the internal measurement board.
- Phase not found — the lock-in phase detection in the IR frontend cannot synchronize with the chopped IR source signal through the new detector, indicating an electrical or address mismatch rather than an optical fault.
Visual inspection of the new detector PCB shows two jumper positions (X1, X2) instead of the three (X1, X2, X3) found on the original B26 detector. Moving the jumpers on the new B526 does not clear the faults.
2. Affected Hardware and Part Numbers
| Part number | Designation | Factory context | Jumper positions |
|---|---|---|---|
C79451-A3468-B26 |
Original CO2 detector | Tested and serialized with the delivered analyzer | X1, X2, X3 (three jumpers) |
C79451-A3468-B526 |
Spare-part CO2 detector | Tested at the factory with a different analyzer; must be reconfigured to the host | X1, X2 (X3 not used) |
Both part numbers refer to the same physical detector hardware. The suffix change (B26 → B526) does not indicate an engineering revision. It is a logistics designation: B26 is the as-shipped-in-analyzer configuration, while B526 is the bare spare-part configuration that must be re-keyed to the host analyzer before commissioning.
3. Root Cause: Detector Address and Phase Synchronization
The ULTRAMAT 23 IR frontend uses address-coding jumpers on each detector PCB to tell the analyzer's measurement board which of the three physical IR channel slots is occupied. The motherboard exposes three matched headers:
| Header | IR channel | Display line | Gas |
|---|---|---|---|
X35 |
Channel 1 | Line 1 | First configured IR gas (e.g. CO) |
X36 |
Channel 2 | Line 2 | Second configured IR gas (e.g. CO2) |
X37 |
Channel 3 | Line 3 | Third configured IR gas (e.g. NO) |
| — | Channel 4 | Line 4 | Reserved for O2 (paramagnetic or electrochemical) |
The original B26 detector carries three jumper pads so Siemens can encode one of three physical slot addresses plus a phase-relationship value into the detector at the factory. The B526 spare part is delivered without the address/phase jumpers populated; only the X1 and X2 pads exist on the PCB, and X3 is intentionally not used. When the B526 is installed without re-keying, the analyzer sees an unknown device on its address bus (Fault of addresses for IR channels) and its lock-in amplifier cannot recover a phase-coherent signal from the chopped IR source (Phase not found).
4. Hardware Configuration Reference
4.1 Detector-side jumpers (C79451-A3468-B526)
On the B526 detector PCB, the three pads are physically laid out left-to-right as X3, X1, X2. Only X1 and X2 are functional:
| Jumper state | Encoded channel |
|---|---|
| X1 closed (middle), X2 open | Channel 1 → plug into X35
|
| X1 open, X2 closed (right) | Channel 2 → plug into X36
|
| X1 closed, X2 closed | Channel 3 → plug into X37
|
| X3 (left, not used on B526) | Do not fit a jumper here |
For a CO2 channel configured as the second IR component in the analyzer, fit the jumper on X2 only and install the detector on motherboard header X36.
4.2 Original resistor transfer
The B26 detector carries a calibration resistor that encodes the factory-measured gain and zero offset for the specific analyzer it was originally paired with. When installing a B526 spare, that original resistor must be removed from the B26 and soldered onto the B526 in the same location. The B526 ships with a generic resistor; without the original, the analyzer will pass the address/phase check but will fail zero and span calibration.
5. Step-by-Step Resolution Procedure
- Isolate the analyzer. Close the sample-gas block valves, vent the optical bench to a safe atmosphere, and de-energize the unit. Allow the IR source to cool per the manual.
- Open the analyzer housing and identify the IR detector stack. The detectors are mounted in the optical bench, each on a small PCB that plugs into a motherboard header.
-
Record the original configuration. Photograph the B26 PCB showing every jumper position, the calibration resistor, and which motherboard header the detector is plugged into (typically
X36for CO2 as channel 2). - Unplug the B26 detector from the motherboard header.
- Remove the calibration resistor from the B26 PCB and solder it onto the B526 PCB in the identical footprint and orientation.
- Configure the B526 jumpers to match the original B26 encoding. For a CO2 detector in channel 2: fit a jumper on X2 only. Do not fit a jumper on X3.
-
Install the B526 on the same motherboard header that the B26 was removed from (typically
X36). Confirm the connector is fully seated. - Re-apply power and let the analyzer complete its self-test (typically 2–5 minutes for IR bench stabilization).
- Clear the fault log from the operator menu. The Fault of addresses for IR channels and Phase not found messages should clear automatically once a valid detector is recognized and lock-in synchronization is achieved.
- Run zero and span calibration with certified test gases. Expect a small offset correction; the value should now track the test-gas concentration without the pre-swap drift.
6. Verification
Use the following acceptance checks to confirm the swap is clean before returning the analyzer to service:
| Check | Method | Pass criterion |
|---|---|---|
| Fault log clear | Operator menu → diagnostics → fault memory | No "Fault of addresses for IR channels", no "Phase not found" |
| Signal level | Service menu → raw IR signal for the CO2 channel | Within the manufacturer's specified range for the configured optical path length |
| Phase lock | Service menu → phase display | Phase value stable; lock-in PLL indicator green |
| Zero drift | Flow zero gas (typically N2) for ≥ 5 min | Reading stable within analyzer zero-noise specification |
| Span response | Flow span gas of certified CO2 concentration | Reading within ±2 % of nominal after one calibration cycle |
| Cross-sensitivity | Expose to the other configured IR gases in turn | No spurious response on the CO2 channel beyond published cross-sensitivity |
7. ULTRAMAT 23 Technical Specifications (relevant excerpt)
The following values are summarized from the manufacturer datasheet and the EPA-published performance specification. Always confirm against the as-installed analyzer nameplate and the current manual revision.
| Parameter | Value |
|---|---|
| Measurement principle | NDIR (non-dispersive infrared) for IR components; paramagnetic or electrochemical for O2 |
| Number of IR components | 1 to 3 |
| Total components measured | Up to 4 (3 IR + 1 O2) |
| Common IR components | CO, CO2, NO, SO2, CH4 |
| Operator interface | 4-line display, line 1–3 = IR channels, line 4 = O2 |
| Detector configuration | Address-coded, plug-in to motherboard headers X35 / X36 / X37 |
For full electrical, environmental, and communications specifications (analog outputs, PROFIBUS, Ethernet, sample gas flow, pressure, and ambient temperature limits), consult the ULTRAMAT 23 Operating Manual (Siemens) and the ULTRAMAT 23 NDIR specifications sheet (EPA archive).
8. Related Fault Codes and Diagnostics
| Displayed message | Likely root cause | First-line corrective action |
|---|---|---|
| Fault of addresses for IR channels | Detector PCB jumpers not set, or detector plugged into wrong header (X35/X36/X37) | Re-key the detector jumpers per the B526 table above; verify header position |
| Phase not found | Lock-in amplifier cannot synchronize; address mismatch or open detector cable | Confirm detector seating; verify the calibration resistor was transferred from the B26 |
| IR signal low / signal out of range | IR source aged, sample cell contamination, or wrong detector installed for the configured gas | Check IR source intensity; clean the optical path; confirm the detector part number matches the configured gas channel |
| Zero drift excessive | Generic calibration resistor on a B526 replacement | Transfer the original calibration resistor from the B26 to the B526 |
9. Preventive Recommendations
- Document the jumper pattern on every detector before removal. A photograph of the original PCB eliminates ambiguity during a swap.
- Retire and label the B26 after the resistor is transferred. The B26 cannot be re-used as a generic spare in another analyzer without re-calibration.
- Order the B526 only as a like-for-like spare and verify on receipt that X3 is not populated and that the calibration resistor footprint is present but empty.
- After any IR detector swap, always run the full zero + span calibration sequence. Do not rely on the auto-calibration routine alone.
- Record the detector serial and the analyzer serial in the maintenance log so that a future B26 → B526 swap on a sister analyzer can be cross-referenced.
10. Field-Commissioning Checklist
- Confirm part numbers received match the purchase order (B26 vs B526).
- Photograph original B26 jumpers and motherboard header position before removal.
- Transfer the calibration resistor from B26 to B526, preserving orientation.
- Set B526 jumpers per Section 4.1 to match the original channel encoding.
- Install B526 on the same header (typically X36 for CO2).
- Power up, clear faults, run zero/span calibration with certified gas.
- Log the swap date, part number, and post-calibration span value in the maintenance file.
What is the difference between C79451-A3468-B26 and C79451-A3468-B526?
The B26 is the original CO2 detector shipped and factory-tested in the specific analyzer; the B526 is the same hardware delivered as a generic spare part tested with a different analyzer. Both part numbers refer to the same physical detector, but the B526 requires jumper reconfiguration and transfer of the original calibration resistor from the B26 before it will operate correctly in a host unit.
Why does the B526 have only two jumpers when the B26 has three?
The three IR channel addresses (1, 2, 3) can be encoded with two binary jumpers, so the B526 only needs X1 and X2. X3 on the B26 is a legacy pad carried over from earlier motherboard revisions and is not used on the B526. Closing X1 selects channel 1 (header X35), X2 selects channel 2 (header X36), and both closed selects channel 3 (header X37).
What causes "Fault of addresses for IR channels" after a detector swap?
The analyzer's measurement board polls the three IR headers (X35, X36, X37) for an address-coded detector. A B526 installed without its jumpers set presents an unknown or default address, so the analyzer cannot bind the detector to a configured gas channel. The remedy is to set the jumpers per the channel position and to transfer the original calibration resistor from the B26 to the B526.
What causes "Phase not found" after a detector swap?
The lock-in amplifier in the IR frontend derives its reference phase from the chopper and the detector's encoded phase relationship. If the address jumpers are wrong, the calibration resistor is missing, or the detector cable is unseated, the phase-locked loop cannot lock and the analyzer reports "Phase not found". Re-keying the jumpers, transferring the resistor, and confirming the connector seating clears the fault.
Do I have to recalibrate after replacing an ULTRAMAT 23 CO2 detector?
Yes. Even with the original calibration resistor transferred from the B26 to the B526, a full zero and span calibration with certified test gases is required to confirm the analyzer tracks concentration correctly. The drift symptom that originally prompted the detector swap should be gone after a single calibration cycle.