
6. PROBE CALIBRATION
6.6 PRACTICAL EXAMPLE OF CALIBRATION FOR COMPARISON
It is supposed to know, by means of data coming from previous analysis, that the duct contains around 10 mg/m3 with an output signal of
18 mA.
Suppose that the maximum value allowed by the law is of 20 mg/m3.
In order to obtain the best calibration, you should adapt the two values in the way that about 70% of the STC signal (15÷16 mA)
corresponds to the maximum dustiness value allowed (20 mg/m3).
On a scale of 4÷20 mA. The value at 50% is of 12 mA
Having previously noticed an output signal from the probe of 18 mA for 10 mg/m3, it is evident that the amplification is too high.
An increase of 30 % of dustiness can be enough to bring the signal to the full range around 20mA by rendering illegible all the emission
values above 13 mg/m3.
Therefore, it is necessary to reduce the sensitivity through SW21 placed inside the probe.
If you consider a higher margin of reading of 30% of STC over the maximum limit of emissions allowed ( 20mA = 20 mg/m3+ 30% = 26
mg/m3 ), a simple calculation shows that at 10 mg/m3 the signal must be around 10÷11 mA.
Therefore, it is sufficient to regulate the SW21 rotary switch in the way to bring the STC output signal around this range of current and set
the activation of the alarm of high emission at about16 mA.
6.7 PRACTICAL EXAMPLE OF CALIBRATION FOR SIMULATION
If the data obtained by the analysis are not available, by following this most approximate method it is possible to make the same
calibrations as previously described. You must simulate an emission of a certain quantity of dust.
This method consists in putting the pre-weighed quantities through the hole mentioned in point 5.2 into the sampling duct in a pre-fixed
time and measure the STC signal generated by the TC probe.
For this calibration it is necessary to know the range of the duct.
The calibration for simulation cannot be done in the absence of dust.
The following example indicates how to proceed:
Range of the duct = 100000 m3/h
Emission that you wish to simulate = 10 mg/m3
Time of insertion of dust in the duct = 60sec.
Quantity of dust to out in 1 minute = 10x100000x60/3600 = 16,66 g.
So, putting in the duct 16,66g in 1 minute, the correspondent emission is 10 mg/m3
Proceed as indicated in point 6.3 CALIBRATION OF THE 4÷20mA SIGNAL
6.8 CALIBRATION OF THE ALARM FOR BROKEN BAG OR BROKEN CARTRIDGE
This type of calibration is relatively simpler than the previous example 6.6 and 6.7 because it consists in checking that the STC signal
detects an anomalous passage of dust in the duct..
As the broken bag or the broken cartridge causes emissions of considerable quantities of dust which are visible at the exit of the chimney,
the calibration of the probe becomes simpler.
This type of calibration is done by throwing in the duct a handful of dust through the hole (see point 5.2) and checking the increase of the
STC signal which is generally considerable.
This peak of the STC signal is used to detect the broken bag by means of the connected equipments or the internal relay in case of
TR30R probe.
WARRANTY
The warranty lasts 4 years. The company will replace any defective electronic component, exclusively at our
laboratory, unless otherwise agreed, upon the Company's prior consent.
WARRANTY EXCLUSION
The warranty is not valid in case of:
1) Tampering or unauthorized repairs.
2) Wrong use of the device, not in compliance with technical data.
3) Wrong electrical wiring.
4) Inobservance of the installation rules.
5) Use of the device, not in compliance with CE rules.
6) Atmospheric events (Lightning, electrostatic discharge), Overvoltage.
Pg. 5