UltraLo-1800

Measuring insulators in the UltraLo-1800 does require some special care, for two reasons:

1. Insulators may build up large static charges that interfere with the electric field inside the instrument;

2. Insulators are much more likely than conducts to absorb moisture that will then slowly outgas inside the counting volume.

Fortunately alleviating these problems is typically straightforward. To reduce the change of accumulations of static charge, handle the material carefully. If it will accumulate charge despite careful handling, you can use a piezoelectric anti-static device (made for removing static from vinyl records) to neutralize it. For more details on static charge in the UltraLo-1800 see the attached white paper below.

There’s less that can be done to eliminate moisture. Some plastics such as PMMA will absorb a significant amount of moisture and takes weeks to fully outgas. Others such as HDPE absorb hardly any and can be measured right away. If the insulator in question does absorb moisture, it must be allowed to outgas before taking good data. This can be done inside the instrument, or it can be done in an external dry box under N₂ purge. Additionally, in most cases the outgassing process will proceed more quickly at higher temperature, so heating the sample while purging with dry gas can ready the sample for measurement very quickly.

The easiest way to tell if a low-emissivity sample is outgassing is checking the moisture meter data, viewable by going to Tools->Logs in CounterMeasure. The read values output to the log every 30 seconds. Typically values need to be <80 ppm for the counting environment to be adequate, but this varies per counter. Before measuring an insulator take a note of the steady-state values in your instrument, and then use that as a guide after inserting the sample.

For more on the effects of moisture on measurements, see page 63 of the User Manual.

If CounterMeasure is having trouble connecting to the instrument, there are a few potential causes. The most common is that the counter is either powered off (via the switch on the back of the support box) or an unplugged USB or power cable.

If the cables are connected and the power turned on, then check the Windows Device Manager’s USB section for “XIA UltraLo-1800”. If it’s present then everything should be working properly.

If it’s not, try clicking the “Scan for Hardware Changes” button. If this turns up nothing, or if the UltraLo-1800 only shows up with an error icon, try cycling the power switch on the support box off and then back on.

If you still cannot connect to the counter, please contact our Support Team.

If counting stops during a measurement the most likely causes are detector power issues or USB connection failures. Especially if the counter is located in a busy lab the USB cables can come disconnected and cause interruptions. Further, temporary power outages, even if they’re very fleeting, can also cause communication issues. (If a stable power line is available, using it will help prevent power dips from causing problems.) If CounterMeasure experiences either interruption while running, it will stop the run and display an error message (see below). Left unattended in this state, CounterMeasure will attempt to reconnect automatically, and if the interruption was temporary it will re-establish the connection. However, CounterMeasure won’t restart a run, and the error message will continue to be displayed to alert you to the issue. To verify that the counter is connected, check for the red “Disconnected” icon in the bottom-left corner of CounterMeasure.

If CounterMeasure isn’t able to reconnect automatically, see I’m having trouble connecting to the instrument. Finally, if you experience counting stoppages that cannot be explained by interruptions in the power supply or USB connection, or particularly frequent stoppages, please contact our Support Team. CounterMeasure will also stop during a measurement if it detects excessive amounts of noise that affects the integrity of the data being collected. In that case it will display the error shown below. For more information on debugging large amounts of noise, see How do I fix a “Data Corruption” error?.

This error occurs when there are too many incoming events for the processors to handle. The root cause is sometimes easy to diagnose (an extremely active sample), or sometimes tricky (electronic noise being injected onto the ground line from noisy equipment elsewhere in a facility).

Assuming the root cause isn’t an active sample, the first step is to follow UL-SB0001 Sidewall Cleaning Procedure. Small bits of dust can land between the fieldshaper rungs, causing intermittent sparking. Cleaning the fieldshapers often resolves this problem. If that doesn’t work, inspection of the noise data may yield insights, environmental noise often follows a particular pattern, increasing when an instrument is turned on as a shift starts and decreasing when the shift is over, etc. To perform this analysis yourself, with the data file in question loaded, export the analysis results (File->Export->Analysis Results) and histogram the trigger time column.

If you need assistance with this error please contact our Support Team and include the buffer file from the run in question. You can find this file in the Measurement History dialog or at C:\Users\UltraLo\AppData\Roaming\XIA LLC\CounterMeasure\Buffers. If the file is too large to attach to an e-mail, you can use our data upload site. Please contact our Support Team if you do not already have a login for the upload site.

CounterMeasure has a maximum number of events it can display before it runs out of memory and crashes (~200,000). Since most 48 hour runs see under 10,000 events, this is usually not an issue. However, with very long runs on very active samples or when there’s excessive numbers of events for any other reason, this limit can be breached, resulting in a crash. If this error occurs when counting a very active sample, you’ll need to perform a shorter run. If the sample is not very active and the excess events are noise or sidewall events, follow the debugging steps in How do I fix a “Data Corruption” error?, as these errors typically have similar underlying causes.

This is due to the fact that the UltraLo-1800 has the secondary veto channel. Because radioactive decay is isotropic, it’s always possible for an alpha from the sample near the anode/guard boundary to pass under the guard and be falsely rejected as a sidewall event. Additionally, the odds of this occurring will increase with increasing alpha energy, as more energetic alphas create longer tracks and are thus more likely to cross the boundary. To account for this effect, there’s an energy-dependent geometric efficiency correction applied to each alpha counted (a table of these values is listed on page 75 of the User Manual). That’s the number displayed by CounterMeasure, with the raw count displayed next to in parentheses.

Note that this argument assumes that the sample will completely cover the active area. If it doesn’t, for example if you’re measuring a small hot source, it may be more accurate to calculate the emissivity based on the raw counts. This is only true if the sample emissivity is much greater than the tray emissivity, as with a source, otherwise the tray emissivity will need to be taken into account and subtracted off, which should use the corrected value.

Every instrument is calibrated by XIA before shipment. The report from this calibration is included with the instrument. This calibration consists of 5 measurements of 3 different samples:

1. Measurement of a NIST-traceable ²³⁰Th source in Full mode.

2. Measurement of the ²³⁰Th source in Wafer mode. These two measurements are used to calculate several counter-specific parameter values.

3. Measurement of a Sn sheet whose emissivity is in the Low Alpha (LA) range, approximately 0.025 alphas/cm²/hr, in Wafer mode.

4. Measurement of the LA sample in Full mode.

5. Measurement of an Al sheet whose emissivity is in the ULA range, approximately 0.001 alphas/cm²/hr, in Full mode.

While only the Th source is NIST-traceable, the other samples have been measured on every single UltraLo-1800 instrument ever made prior to shipment from XIA, and they’ve also been measured by many of those instruments in the field, so their emissivities are very well determined.

For more on Calibration, see the flyer below.

During typical use, with count rates of ~1 event/minute, the deadtime is extremely small. During these conditions the only deadtime is from the 300 μs event acquisition window. (Which means that the only events missed are due to pile-up.) This will obviously lead to a very, very small amount of deadtime, so small as to be negligible.

However, if count rates increase to tens of events per second or more, deadtime from the processor electronics will start to be significant. Normally the event buffers are sufficient to capture a quick flurry of events during otherwise normal operation, but if the elevated count rate is sustained for several seconds or longer these buffers will be swamped and deadtime will accrue. We have not determined an analytic formula for estimating deadtime vs count rate because these rates are many orders of magnitude above the intended range for the UltraLo-1800.

While the deadtime is not directly reported by the software, the livetime is used in the calculation of emissivity. Thus, if you wish to calculate it from the reported measurement values in CounterMeasure, you’d need to do:

Where corrected_alphas is the efficiency-corrected alpha count, emissivity is the reported emissivity, and active_area is the area of the electrode (either 707 for Wafer or 1800 for Full).

Samples larger than a 12” circle can be counted trivially in the UltraLo-1800, but for those smaller special measures are required, depending on whether or not several of the smaller pieces are available.

Many small samples are available
In this case, you can tile the samples to completely cover the tray. With rectilinear samples this is easy, just abut the edges and tile until the tray is covered. For non-rectilinear samples, you may need to overlap them to get full coverage. In that case the overlap may act as an outgassing sample, requiring extra purge/counting time to eliminate the moisture and oxygen from the cracks. For more information on outgassing samples see Do I need to take special precautions to measure an insulating sample?. Otherwise there is no issue with overlapping small samples.

Only one piece is available
In this case you will need to perform background subtraction to remove the contribution of the exposed sample tray (or another low-emissivity substrate, at XIA we use a conductive PTFE tray liner). For more on background subtraction see page 22 of the User’s Manual.

There are many reasons to want to line the tray of the UltraLo. While the emissivity is acceptable, it being a standard grade of stainless steel means that it’s higher than you may like if you’re measuring under-sized samples and need to do background subtraction. Additionally, while the stainless is relatively easy to clean it’s not ideal when counting a paste or a powder.

For those reasons XIA has long used a conductive PTFE tray liner in those applications, and many of our customers have purchased the same liner. We used a 21” roll of DeWAL 105 skived to 0.002” thickness. This covers the complete area of the Full anode size, while being conductive (so static charge does not build up), and providing enough thickness to fully stop alphas. Our experience is that its emissivity is less than half that of a typical tray, usually measuring around 0.0006-0.0010 alphas/cm^2/hr. (Note that this material is not screened for radioactivity so we cannot guarantee these number, they are simply what we have observed from the sample.) If this sounds appealing to you, contact the manufacturer for a quote.

The advantages of using an uninterruptible power supply (UPS) with the UltraLo-1800 are obvious, in any environment with occasional, brief power outages it will allow the system to keep running. The disadvantage is much less obvious: by modifying the instrument’s path to ground, it may cause performance degradation. XIA has not done extensive tests for compatibility with any UPS models on the market, but in short tests using models purchased for other purposes we did not have satisfactory results, and the injected noise compromised performance too much to justify the gain in uptime. However, several customers have reported successes with running on a UPS, so it is possible to find units that do not affect performance. We do not currently have a list of those models but will update this page with any successes that are reported to us.