To: Sensitech Customers and Distributors
From: Technical Support and Marketing
Date: March 22, 2004
Re: TempTale 3 and TempTale 4 monitors

This bulletin explains timekeeping in the TempTale monitors.

Affected Systems
TempTale 3 (TT3) and TempTale 4 (TT4) monitors.

Explanation

TempTale 3
There is an oscillator circuit in the TT3 that records elapsed time. This means that the TT3 does not know the current clock time. The TT3 simply keeps track of the number of measurement intervals that have elapsed since the TT3 was started. As a result, the measurement interval programmed into the TT3 (at the time the unit is configured) determines the resolution, with which one can determine when a specific reading was taken.

The figure below shows how the time associated with each reading is determined when a TT3 is downloaded by a PC (TTM or CCM software). In this illustration, the TT3 was started and the first reading R(1) is taken at time t1. Subsequent readings are taken at regular measurement intervals. A total of n readings were taken. The most recent reading R(n) was taken at time tn. The TT3 is downloaded some time after tn but before the next reading has been taken.

At download, all the times {start time, t1, t2, …, tn} are calculated from the local time in the PC, since the TT3 has no internal clock. The PC simply assumes that the most recent reading was taken just prior to the download and assigns tn to the current local time (obtained from the real time clock inside the PC). Times for all the other readings, {R(n-1), R(n-2), … R(1)}, are calculated by subtracting the measurement interval from the previous reading.

There are two possible sources of error in determining the time of each measurement. The first error is inaccuracy of the local PC clock. If the PC time is incorrect, all the reading times and the TT3 start time will also be incorrect.

The second source of error is caused by the delay between the most recent reading, R(n), and the actual download time. This error can be as large as the measurement interval of the TT3. In TT3s configured with large measurement intervals (e.g. one hour), this error may be significant. The overall accuracy of the time associated with each reading is therefore dependent on the measurement interval and the accuracy of the PC’s clock. Times recorded using TT3s could be reported as being taken later than they actually occurred (by as much as the TT3’s measurement interval).

TempTale 4
The TempTale 4 has improved timekeeping accuracy. It solves the issues of the TT3 by incorporating its own internal clock.

The figure below shows how the time associated with each reading is determined when a TT4 is downloaded.

The TT4’s internal clock is set when the TT4 is configured. This typically happens at the Sensitech factory where computers used to configure TT4s are synchronized with the Internet atomic clock at least once a day. The TT4’s internal clock operates in Greenwich Mean Time (GMT).

When an operator starts the TT4, the TT4 interrogates its internal clock and records the start time, tstart. The TT4 then waits for the start-up delay to pass before taking its first reading. Reading are then taken at regular measurement intervals.

When the TT4 is downloaded by the PC/TripStrip, times for all the readings, {t1, t2, …, tn}, are calculated from the start time, tstart, stored in the TT4. The PC/TripStrip is interrogated for its local time zone in order to convert all reading times into the local time zone.

Notice that neither the accuracy of the PC/TripStrip’s clock, nor the exact time of the download, has any impact on the time accuracy of the TT4 readings. This is because the frame of reference for time is the TT4’s internal clock.

The TT4’s internal clock is accurate to within 2.5 minutes per month. If a TT4 has “sat on a shelf” for eight months, then there may be up to a 20-minute discrepancy in the start time recorded by the TT4. However, this time accuracy tolerance is well defined and independent of the measurement interval.

TT4s are typically started within one to two months of when they are delivered to a customer. They are usually configured shortly before shipment to the customer. In this situation, the accuracy of the reading times is within five minutes. This accuracy is normally well within the thresholds necessary to determine where a shipment was when an out-of-range temperature event occurred.

HOME | ABOUT | PRODUCTS | APPLICATIONS | SERVICES | COLD CHAIN INFORMATION | SUPPORT
CUSTOMER LOGIN | QUALITY CERTIFICATIONS | CONTACT US | PRIVACY POLICY

Copyright © 2008 Sensitech. All Rights Reserved. Sensitech is a business unit of Carrier Corp.