HPLC Detector Lamp Replacement and Intensity Test
Deuterium lamp in HPLC UV, DAD and VWD detectors. Lifetime, intensity test, ignition faults, noise, drift, lamp exchange and warm-up time.
The deuterium lamp provides the UV light in an HPLC detector. As it ages, intensity drops, noise rises and the baseline becomes unstable. It draws on manuals and service pages from Agilent, Shimadzu and Thermo.
Which lamp does what
The deuterium lamp covers the UV range. In the Agilent 1260 VWD it works from 190 to 600 nm. The Thermo UltiMate 3000 VWD uses the deuterium lamp below 345 nm and an optional tungsten lamp up to 670 nm. Diode array detectors and Thermo units with a second lamp use the tungsten lamp for the visible range. Shimadzu fits only a deuterium lamp in the SPD 20A and adds a tungsten lamp in the SPD 20AV and SPD M20A.
Lifetime according to the manufacturers
The Agilent manuals we read give no fixed hour figure. Lamp on time runs in a counter whose limit you set yourself. Agilent explains that useful burn time depends on the application, such as sensitivity and wavelength. Thermo and Shimadzu do state values.
| Manufacturer | Lamp | Stated value |
|---|---|---|
| Agilent | Deuterium, VWD and DAD | No hour figure. Counter with user defined limit, replace based on test and baseline |
| Thermo | Deuterium, UltiMate 3000 VWD | Typical lifetime about 2000 hours |
| Shimadzu | Deuterium for SPD 20A and SPD 20AV | Guaranteed life 2000 hours or one year from purchase date |
| Shimadzu | Deuterium for SPD M10A(V) | Guaranteed life 1000 hours or one year |
| Shimadzu | Tungsten for SPD 20AV | Average life 2000 hours, guaranteed 1200 hours or one year |
The Shimadzu figures are warranty limits, not predictions. The intensity test remains decisive.
Running the intensity test
The test measures the lamp spectrum. Fill the flow cell with water so that no solvent absorbs light. Shimadzu recommends degassed deionised water. On the Agilent VWD the test covers 190 to 600 nm in Lab Advisor or ChemStation and rates average, minimum and maximum against limits.
The manufacturers test differently.
- Thermo monitors relative intensity at 230 nm for the deuterium lamp and at 700 nm for the tungsten lamp. You set a lower limit and a message appears when intensity falls below it.
- For the SPD M20A, Shimadzu gives a deuterium maximum near 235 nm of at least 7 V. For the SPD 20A it gives a reference intensity above 400 at 220 nm, and for the SPD 20AV also above 500 at 540 nm for the tungsten lamp. The page states no unit for these two values.
- Agilent rates three values or four spectral ranges with fixed limits, depending on the model.
Record the values at every check, since a slow decline shows the replacement point better than one reading.
A failed test does not always point to the lamp. Agilent lists an empty or dirty flow cell and air bubbles. Repeat the test without the cell. If it then passes, the cell windows are the problem. On DAD units a silvered or discoloured UV lamp can also lower visible intensity.
Ignition problems and lamp errors
A gas discharge lamp ignites poorly while it is hot. After an ignition failure on the DAD, Agilent advises letting the lamp cool for at least 15 minutes. Thermo blocks re-ignition for 5 minutes after switch off and shows the remaining time.
| Message | Manufacturer | Possible cause | Action |
|---|---|---|---|
| UV lamp no current (7450) | Agilent VWD | Loose connector, defective or non-Agilent lamp | Check connector, replace lamp |
| UV lamp no voltage (7451) | Agilent VWD | Defective or non-Agilent lamp, power supply, main board | Replace lamp, otherwise service |
| Ignition failed (7452) | Agilent VWD | Current does not rise within 2 to 5 s, lamp disconnected or defective | Check connection, replace lamp |
| No heater current (7453) | Agilent VWD | Heater current missing, lamp disconnected or defective | Check connection, replace lamp |
| UV lamp ignition failure | Thermo VWD | Lamp does not ignite | Switch off and on, otherwise replace lamp |
| UV lamp intensity low | Thermo VWD | Intensity below reference | Check limit, replace lamp |
| Lamp cooling down | Thermo VWD | Restart before the 5 minutes passed | Wait |
Agilent notes that non-Agilent lamps can trigger these errors.
Classifying noise and drift
| Symptom | Possible cause | Action |
|---|---|---|
| Baseline drifting downward after switch on | Lamp not warmed up enough | Warm up 30 minutes (Shimadzu), up to one hour for Agilent |
| Periodic swings even without a cell | Lamp near end of life | Remove the cell, if the pattern stays, replace the lamp (Agilent) |
| Disturbance with flow stopped | Lamp or detector | Check D2 intensity, replace a weak lamp |
| Noise and drift with a healthy lamp | Dirty flow cell | Flush with 2-propanol, then mobile phase |
| Sawtooth or spikes | Gas bubbles in the cell | Degas mobile phase, do not shorten outlet tubing |
| High noise with a new lamp | Burn in phase | Let it run, up to 24 hours for Thermo |
For DAD units Agilent recommends switching lamps off overnight or when idle for more than 4 hours. A reference wavelength close to the sample wavelength, for example 254 nm sample and 350 nm reference, reduces lamp instability. The ASTM noise test of the Agilent 1200 DAD has a limit of plus or minus 0.02 mAU. If it fails, Agilent lists too short a warm-up or an old lamp as causes.
Replacing the lamp
Switch the detector off. Thermo also requires unplugging the power cord. Wait until the lamp has cooled. Agilent gives five minutes for the DAD, Thermo asks for sufficient time because lamp and surroundings stay very hot. The deuterium lamp emits UV radiation that harms eyes and skin. Never run it outside the instrument and follow the manufacturer instructions.
- Remove the front cover and lamp cover.
- Unplug the lamp connector.
- Loosen the two screws and take out the lamp.
- Fit the new lamp. Do not touch the glass bulb.
- Tighten the screws without overtightening, reconnect the connector, refit the covers.
Agilent specifies a Pozidriv size 1 screwdriver and advises against a standard Phillips driver. For lamps without an RFID chip, reset the lamp hour counter. Shimadzu requires resetting D2 LAMP USED on the SPD 20A and SPD 20AV, and lamp usage time plus exposure time setting on the SPD M20A.
Warm-up and follow-up checks
After the exchange Agilent recommends more than 10 minutes of warm-up before diagnostics and 60 minutes before samples. Then run wavelength verification, the holmium oxide test and finally the intensity test. Wavelength verification takes about 2.5 minutes and is disabled during the first 10 minutes after ignition. Shimadzu expects about 1.5 hours for a stable PDA baseline. Thermo requires a calibration test and at least 24 hours of running before the first analysis, because strong fluctuations and increased noise can occur.
Matching parts from the catalogue
Check fit against the rating plate and part number before ordering.
- According to Agilent, deuterium lamp G1314-60100 is the standard lamp for the VWD G1314B and G1314C. With an RFID chip, Agilent lists G1314-60101 for G1314D, E, F and G7114.
- For the Agilent DAD models G1315C and D, Agilent names 2140-0820. The 5182-1530 is listed as a deuterium lamp for 1100 and 1200 DAD.
- The tungsten lamp for the visible range in the Agilent DAD is G1103-60001.
- Shimadzu lists 228-34016-02 for SPD 20A and SPD 20AV, 228-34016-00 for the SPD M20A and the tungsten lamp 670-14602-00 for the SPD 20AV.
- For Thermo UltiMate 3000 the relevant parts are deuterium lamp 6074.1110 and tungsten lamp 6074.2000.
- For flow cell problems, consider the standard flow cell G1314-60086 for the Agilent VWD and the flow cell 6074.0250 for the Thermo VWD.
FAQ
How do I know the deuterium lamp needs replacing?
Replace it when intensity falls below the limit or the lamp does not ignite. Periodic baseline swings without a flow cell also point to an old lamp.
Do I have to replace it after a set number of hours?
Not necessarily. Thermo gives about 2000 hours as a typical value and Shimadzu guarantees 2000 hours or one year. Agilent gives no fixed value and points to the test and the baseline.
Why is the baseline still noisy after the exchange?
A new lamp needs time. Agilent and Shimadzu mention one to one and a half hours, Thermo up to 24 hours before the first analysis. Repeat the intensity test only after the warm-up period.
Do you need one of the parts mentioned? We deliver the matching part with a quote the same day, express shipping is possible.