Voltage control of capillary electrophoresis high-voltage power supplies in biomarker assays and traditional medicine fingerprinting
A capillary electrophoresis instrument uses a high-voltage supply ranging from 0 to 30 kV, with a current limit of 200 μA. The fused silica capillary has an inner diameter of 50 μm and a total length of 60 cm. The separation voltage is set at 20 kV for protein biomarker assays, and at 25 kV for traditional medicine fingerprinting runs. The electric field along the capillary reaches 333 V per cm at 20 kV, driving the electroosmotic flow at 2 mm per second. The supply must hold the voltage within 0.1% during a 10 minute run, because a drift of 20 V shifts the migration time by 0.5%, enough to misidentify a peak in the fingerprint.
The ramp-up from zero to 20 kV takes 5 seconds, at a rate of 4 kV per second. The ramp rate was set slow to avoid a sudden electroosmotic flow that would disturb the sample plug at the capillary inlet. A faster ramp of 10 kV per second once caused a 2 mm wide sample band, doubling the peak width in the chromatogram. The 5 second ramp is now the default, and the operator accepts the extra 3 seconds per run. The ramp-down at the end of a run is also slow, taking 3 seconds to reach zero, to prevent backflow in the capillary.
The current through the capillary is monitored at 10 Hz. A typical run draws 30 μA at 20 kV, corresponding to a buffer conductivity of 0.1 S per m. A current spike above 100 μA triggers an immediate shutdown, because the spike indicates a bubble in the capillary or a leakage path across the buffer vial. The shutdown response time is 50 μs, and the voltage drops to zero within 100 ms. The supply then attempts a restart after 10 seconds, and if the spike repeats, the supply locks out and requires manual reset.
The buffer vials are 2 mL polypropylene, held in a carousel that positions the inlet and outlet ends of the capillary. The vials contain 1 mL of borate buffer at pH 9.0 for protein assays, or 20 mM phosphate buffer at pH 3.0 for traditional medicine fingerprinting. The vials are changed every 20 runs, and the capillary is rinsed with 0.1 M NaOH for 2 minutes between runs. The supply has no direct contact with the buffer, but the electrodes in the vials connect to the supply output through shielded leads.
The high-voltage electrode in the outlet vial is a platinum wire, 0.5 mm in diameter, immersed 10 mm into the buffer. The wire corrodes slowly over 6 months of continuous use, and the surface turns dull. The corrosion raises the contact resistance from 10 mΩ to 100 mΩ, negligible at 30 μA, but the local heating at 3 μW warms the wire by 2 degrees. The wire is replaced at the 6 month mark, preventively. The inlet vial has no high-voltage electrode, because the capillary inlet dips into the buffer and the supply ground is connected to the outlet electrode.
The supply cabinet sits on the bench next to the electrophoresis instrument, connected by a 2 m shielded high-voltage cable. The cable is a 30 kV silicone type, with the outer jacket grounded at both ends. The cable once developed a hairline crack in the jacket, and the leakage current rose from 1 μA to 20 μA. The leakage did not trip the 100 μA overcurrent threshold, but the extra 19 μA added noise to the current measurement, and the peak integration became unstable. The cable was replaced at the next scheduled maintenance, and the leakage returned to 1 μA.
The control board uses a 16-bit ADC, sampling the output voltage at 1 kHz. The ADC reference is a buried-zener type, with a drift of 2 ppm per degree Celsius. The cabinet temperature is held at 25 plus or minus 1 degree by a small fan and heater, and the reference drift over a 10 hour run stays below 5 ppm, equivalent to 0.15 V at 30 kV. The drift was measured during commissioning, and the supply was accepted for the 0.1% stability requirement. The reference is recalibrated annually against a 0.01% standard divider.
The instrument runs two shifts per day, 6 days per week. The first shift runs protein biomarker assays, about 40 samples per shift. The second shift runs traditional medicine fingerprinting, about 30 samples per shift. Each sample requires a 10 minute separation, plus 2 minutes for rinsing and vial change. The supply is energized for about 12 hours per day, and idle for 12 hours. The idle state keeps the output at zero, with the control board still powered. The standby power draw is 15 W, and the daily energy use is about 0.5 kWh.
The safety interlock on the buffer vial carousel was tested during commissioning. A light curtain across the carousel access area triggers a shutdown when an operator reaches in. The shutdown response time is 20 ms, and the residual charge on the output capacitor bleeds through a 10 MΩ resistor within 100 ms. The light curtain was tripped by accident once when a technician leaned over the carousel, and the supply shut down mid-run. The run was lost, and the sample was discarded. The light curtain sensitivity was adjusted to ignore a slow lean-over, and the false trip has not recurred.
The maintenance log lists routine tasks. Each shift, the operator records the output voltage, current, and cabinet temperature. Weekly, the technician inspects the platinum electrode and the cable jacket. Monthly, the insulation resistance is measured with a 2500 V megohmmeter, and the reading should exceed 1000 MΩ. Quarterly, the divider is calibrated against a 0.1% standard. Annually, the ADC reference is recalibrated. The log is signed at each step, and the record is filed in the plant archive.
The high-voltage transformer inside the supply cabinet is oil insulated, with a 2 L oil volume. The oil temperature reaches 38 degrees after 8 hours of continuous operation at 20 kV and 30 μA, against an ambient of 25 degrees. The silica gel breather on the transformer tank changes color after 12 months, and the gel is replaced before the color reaches half saturation. The oil breakdown voltage is tested annually, and the last measurement showed 45 kV per 2.5 mm gap, well above the 25 kV replacement threshold. The transformer sound level is 45 dB at 1 m, measured with a sound level meter during commissioning.
The input side of the supply draws 230 V single phase at 1 A. A 2 kVA isolation transformer sits between the wall outlet and the supply, to break the ground loop between the instrument and the supply. The isolation transformer has a 1500 V insulation rating between primary and secondary. The input cable is copper, 1.5 mm², rated for 10 A, routed along the bench edge. The input breaker is a 10 A miniature type, with a thermal-magnetic trip. The input power factor is 0.95, measured with a power analyzer during commissioning.
The output divider is a series chain of 100 MΩ high-voltage glass glaze resistors, totaling 1 GΩ. The divider tap at 1:1000 feeds the control board ADC. The divider is calibrated quarterly against a 0.1% standard divider. The calibration drift over 12 months was 0.3%, equivalent to 60 V at 20 kV. The drift is within the 1% acceptance limit, but the divider is scheduled for resistor replacement at the 24 month mark, when the drift is projected to cross 0.5%.
The instrument data system logs each run with the sample ID, voltage, current, migration time, and peak areas. The data is stored on the instrument hard drive, backed up weekly to a network server. The supply voltage trace is recorded at 1 Hz, and the trace is superimposed on the electropherogram. A voltage dip during a run is visible as a flat spot in the current trace, and the run is flagged for review. The data retention policy is 5 years, and the records are available for audit.
The customer acceptance test was run over two days. Day one covered the protein assay at 20 kV, measuring voltage stability, current noise, and peak repeatability. Day two covered the traditional medicine fingerprint at 25 kV, measuring peak resolution and migration time reproducibility. The test report was signed by the customer representative, and the supply entered the 12 month warranty period. The warranty terms cover parts and labor, with the customer paying for field travel.
The capillary is a fused silica tube, 50 μm inner diameter, 360 μm outer diameter, 60 cm long. The capillary is held in a cassette, with the inlet and outlet ends aligned to the buffer vials. The cassette is loaded into the carousel, and the supply electrodes make contact with the vial electrodes through spring-loaded pins. The capillary window, a 5 mm long section of polyimide coating burned off, sits in front of the UV detector. The detector measures absorbance at 214 nm for proteins and at 254 nm for traditional medicine compounds.
The buffer solution preparation is done in a separate lab room. The borate buffer is prepared from boric acid and sodium hydroxide, adjusted to pH 9.0 with a pH meter. The phosphate buffer is prepared from sodium phosphate monobasic, adjusted to pH 3.0 with phosphoric acid. The buffers are filtered through a 0.22 μm membrane filter before use, to remove particles that could clog the capillary. The filtered buffer is stored in a sealed bottle, and discarded after one week. The buffer preparation area is 10 m from the instrument, and the buffer bottles are carried to the instrument on a tray.
The sample introduction is done by electrokinetic injection. The capillary inlet is dipped into a sample vial, and a 5 kV injection voltage is applied for 10 seconds. The sample ions migrate into the capillary under the electric field. The injection time and voltage are optimized for each sample type. Protein samples use 5 kV for 10 seconds, while traditional medicine samples use 10 kV for 5 seconds. The injection voltage is supplied by the same high-voltage supply, with a separate output channel. The supply has two independent outputs, one for separation and one for injection, each rated at 30 kV and 200 μA.
The instrument room environment is held at 22 plus or minus 2 degrees Celsius, with 40 to 60 percent relative humidity. The room has a dedicated air conditioning unit, with the airflow directed away from the instrument to avoid drafts on the capillary. The room illumination is 500 lux, with dimmable LED fixtures, because the UV detector is sensitive to stray light. The room has a sign on the door, warning of high voltage and UV radiation. Entry requires a lab coat and safety glasses, and non-authorized personnel are not permitted.
The instrument data system records each run, with the sample ID, the voltage, the current, and the electropherogram. The data is stored on the instrument hard drive, backed up weekly to a network server. The quality department reviews the data monthly, and trends the migration time and the peak area. The data system generates a monthly report, showing the assay throughput and the failure rate. The report is distributed to the lab manager and the customer.
The lab safety audit is conducted annually, by an internal safety team. The audit covers the electrical safety, the chemical safety, and the radiation safety. The audit report lists the findings, and the corrective actions are scheduled within 30 days. The last audit found one minor finding: a worn gasket on the buffer vial carousel. The finding was corrected within the scheduled time. The audit record is filed in the quality system.
The lab spare parts inventory includes: platinum electrodes, high-voltage cables, fuses, and buffer vials. The inventory is stocked at 20 percent of the annual consumption, and is reviewed monthly. The inventory is stored in a locked cabinet, and the key is held by the lab supervisor. The parts are logged out when used, and the log is reviewed by the quality department. The inventory was once depleted when a batch of electrodes oxidized, and the stock was replenished within 2 days.
The analytical lab operates two shifts per day, 5 days per week. The day shift runs the biomarker assays, and the evening shift runs the traditional medicine fingerprinting. The sample changeover happens between runs, and the buffer vials are swapped. The shift handover includes a status report, listing the completed runs, the migration time, and the peak areas. The handover log is reviewed by the lab supervisor, who allocates resources for the next shift.
The supply has run for 18 months. The platinum electrode has been replaced once. The high-voltage cable has been replaced once. The output capacitor shows a 2% capacitance loss, still within the 10% threshold. The unresolved issue is the ramp time. The operator asked for a faster 2 second ramp, but the faster ramp disturbs the sample plug. A software update is under evaluation, but the update requires a hardware change to the control board, and the change is not scheduled until the next instrument upgrade.
