The platform: hydraulics, detectors, reagents, electronics, printer and LIS

Trace one XP-300 sample from the pipette to the printout, name the assembly and board that does each step, and know which pressure, vacuum and reagent each step depends on.

What sits on the bench

The XP-300 is a compact three-part-differential haematology analyser. It is the successor of the KX-21 and KX-21N, and the service manual says so openly: the hydraulics, the sequence software and much of the mechanics were carried over from the KX-21N, while the user interface and data handling came from a newer small analyser. For you this matters twice. First, many parts carry KX-21 names in the parts list: the SRV unit, the detector block, the regulator unit, the pneumatic unit and the chamber unit. Second, much of what experienced engineers know about KX-21 faults transfers to the XP, so field notes and colleagues' experience on the older model are worth reading, as long as you check them against the XP documentation.

Everything is in one housing: the main unit with its own compressor, the built-in lyse container, the colour touch screen, the thermal printer and the start switch behind the aspiration pipette. Diluent and waste are outside in their own containers. There is no autoloader and no closed-tube piercer. The operator opens the tube, holds it under the pipette and presses the start switch. That simple front end is why so many XP faults start with the sample and the pipette rather than with electronics.

The analyser reports eight measured parameters (WBC, RBC, HGB, HCT, MCV, MCH, MCHC, PLT) and a group of histogram-derived ones: the three WBC size populations as ratios and counts, RDW, PDW, MPV, P-LCR and PCT. It has a whole blood mode and a pre-diluted mode for small samples that the laboratory dilutes beforehand. Throughput is about one sample per minute and the manual's design life assumes a small laboratory workload. Keep that in mind when a busy site runs it far above the design load: maintenance intervals by cycle count will arrive much faster than by calendar.

From pipette to chambers: the SRV and the dilutions

The heart of the hydraulics is the sample rotor valve, the SRV. It is a sandwich of two fixed, precision-lapped valve plates with a rotor between them, turned by a small air cylinder. A metering channel in the rotor holds an exact volume of blood. When the rotor turns, that volume is cut off from the aspiration line and lined up with the diluent path, and a diaphragm pump flushes it out with a fixed volume of diluent. This is how the XP measures volumes without a syringe: the SRV defines the blood volume, the diaphragm pumps define the reagent volumes.

In whole blood mode the analyser takes roughly 50 µL of anticoagulated blood through the pipette. The SRV portions it for two branches. For WBC and HGB, a small portion is diluted with diluent and lyse to roughly 1:500 directly in the WBC detector chamber. For RBC and PLT, a portion is first diluted in the mixing chamber, a small part of that first dilution is taken back through the SRV, and it is diluted again in the RBC detector chamber, giving roughly 1:25,000. In pre-diluted mode the laboratory has already diluted the blood 1:26, the analyser aspirates a larger volume of that, and the RBC branch needs only one dilution step.

Mixing is done by air bubbles blown into the chambers, not by stirrers. Draining and filling are done by master valves switching pressure and vacuum. There are seven diaphragm pumps. Each has a pneumatic side and a liquid side: vacuum on the pneumatic side draws in a fixed volume of liquid, pressure pushes it out. If the drive pressure or vacuum is weak, the pump strokes short and the dilution is wrong, without any part being 'broken'. Remember this sentence for lesson 4: on the XP, wrong counts are often a pneumatic fault.

Pneumatics: one compressor, three levels

A single piston compressor in the pneumatic unit makes both pressure and vacuum. The regulator unit turns them into the working levels. According to the service manual there are three that matter. A higher drive pressure of about 0.20 MPa works the master valves and the SRV air cylinder through solenoid valves. A regulated 0.05 MPa pressure drives the diaphragm pumps, the bubble mixing and the waste drain. A regulated vacuum of about −0.0333 MPa, around 250 mmHg, fills the reagent chambers, drains the waste chamber and pulls the sample through the apertures. The raw vacuum from the pump, stronger than the working level, is reduced by a bellows regulator with a needle valve and a small air reservoir.

The XP has no pressure or vacuum tanks. The manual describes lengths of tubing inserted in the lines to smooth pulsation instead. Only the two working levels are monitored by sensors on the analogue board; the source levels are not. The machine therefore tells you when the 0.05 MPa or the −0.0333 MPa level is out of range, but it does not tell you whether the cause is a weak compressor, a dirty regulator needle, a leaking relief valve or a cracked tube. For that you have four mechanical gauges: two on the pneumatic unit for the source levels and two on the main unit for the working levels. Read all four before you adjust anything.

Twenty-four solenoid valve outputs drive the solenoid valves at 12 V DC. The solenoid valves do not switch liquid; they switch air to the master valves, which switch liquid, pressure and vacuum lines. A solenoid valve that clicks has proved only that its coil is energised. Whether the master valve behind it moves, and whether that master valve seals, is a separate question.

Detectors: impedance and HGB

Counting uses the DC impedance method you met in Level 1. The WBC detector has a larger aperture, about 100 µm in the manual's description, and counts a metered half-millilitre of the lysed sample. The RBC detector has a smaller aperture, about 75 µm, and counts a metered quarter-millilitre of the high dilution; the same pulses are split electronically into RBC and PLT by size. The XP has no separate PLT channel and no optical counting. A clog, a dirty aperture or bubbles in the RBC detector therefore damage RBC and PLT together, while the WBC detector is separate.

Each detector block also carries what the counting needs around it: electrodes for the aperture current, a clog-monitoring signal derived from the DC voltage across the aperture, isolation chambers that separate the detector liquid from the rest of the hydraulics so that noise does not ride in on the tubing, and a thermistor near the detectors. The temperature is used for HCT compensation and for the clog calculation, and it is the source of the room temperature errors.

HGB is measured in a flow cell on the WBC branch. An LED at 555 nm shines through the cell to a photodiode. Before each sample the analyser reads the blank (diluent and lyse without blood) and then the lysed sample, and calculates HGB from the difference. The method is cyanide-free, based on sodium lauryl sulfate in the lyse. A dirty flow cell, bubbles, a weak LED or a wrong lyse volume all show in the blank and sample readings, which the status display shows as an HGB convert value.

Reagents and the analyser's view of them

The XP is designed for three reagents: CELLPACK as diluent, STROMATOLYSER-WH as lyse for WBC and HGB, and CELLCLEAN as a strong detergent used at shutdown and in cleaning procedures. Control blood is EIGHTCHECK-3WP, the calibrator is SCS-1000, and the gain adjustments of lesson 5 use CELLCHECK and a latex calibrator for PLT. The analyser monitors reagents in several ways: float switches in the diluent, waste and lyse chambers, filling-time limits, cycle counts since the last container change, expiry dates read from the reagent barcode, and two plausibility checks. One watches the conductivity of the diluent; the other watches whether the WBC histogram can be divided into its three populations. When either check fails, the analyser shows an analysis error and suppresses or marks the affected results. The manual states openly that these checks exist to detect reagents that do not behave like the specified ones. That is part of how the analyser protects patient results; it is not an obstacle to be worked around.

Many laboratories run third-party reagents. That is the laboratory's decision and its validation responsibility. Your job is to record which reagents are on the machine, to know that settings and calibration on such a machine may have been adapted to them, and never to change detection limits or protective settings to make an error go away.

Electronics, software, printer and LIS

The electronics are a small set of boards. The CPU board, PCB No. 6374, runs the software on an embedded processor, holds the program in flash memory, keeps settings, results and QC in battery-backed RAM with a lithium cell, has a CompactFlash slot used for backups and program updates, and carries the serial, parallel and Ethernet ports. The analogue/driver board, PCB No. 30012, digitises the pulse heights, converts the HGB signal, and drives the solenoid valves and the stepper motor; it also reads the float switches and the start switch. The analogue main board, PCB No. 20034 or No. 2135 (the manual lists both; check the parts list and the board fitted in your unit), amplifies the detector signals, supplies the aperture current, runs the HGB LED and photodiode, and carries the pressure, vacuum, temperature and clog sensing, together with the trimmers used in the adjustments of lesson 5. An LCD relay board, No. 9301, carries the touch controller, buzzer and backlight supply. An optional board, No. 90043, sends SNCS reports over the network. The power supply unit gives low voltages for logic, analogue circuits, valves and motor, and also a 100 V DC supply for the aperture current and an AC supply for clog removal and the compressor.

Data leaves the analyser in three ways. The built-in thermal printer prints results with histograms. A serial RS-232C port and an Ethernet port send results to a host computer or LIS; a handheld barcode reader reads sample IDs, reagent barcodes and QC target values. SNCS, where the distributor uses it, sends online QC, error logs and maintenance data to a remote server. Each of these depends on settings in the analyser, and a remarkable share of 'printer broken' and 'LIS not receiving' calls are settings that changed, not hardware.

On-site checklist

• Model, serial number and software version recorded
• Reagents on the machine recorded: diluent, lyse, cleaner, lot and expiry, original or third-party
• All four pressure and vacuum gauges read and noted
• Workload estimated: samples per day versus the cycle counts on the status display
• Host connection type noted (serial, Ethernet, SNCS) and output settings printed
• Board revisions visible on the parts you may need noted, for ordering

Check yourself