What is the 6890 GC analog output conversion to millivolts when connected to eSATIN? - WKB6557

ENVIRONMENT

• Empower
• Empower 2
• Empower 3
• Millennium, all versions
• 6890 GC Analog data collected via eSATIN

ANSWER

• Technical Service Note TECN1852238 - Interpreting the signal output of Flame Ionization Detectors (FID) in Agilent gas chromatographs
• The 6890 analog output provides signal in the ranges of 0 to 1 V, 0 to 10 V, or 0 to 1 mV.  The output 0 to 1 V provides 1 mV per detector display unit at Range 0.
• This article applies only when the data is collected via eSATIN when GC is controlled with inbuilt driver in Emower (RS232 Communication). With respect to driver / ICF based communication this article do not apply.

1 mV translates to

• 1 pA of FID/NPD
• 25 uV of TCD
• 5 Hz of ECD
• 1 Hz of uECD

ADDITIONAL INFORMATION

Electrically, there is Ohm's Law, which is V=IR (Voltage = Current x Resistance), so there should be a simple conversion between milliVolts and picoAmps, which would be R, but this is not the case. The reasons for this are:

• We cannot easily measure the R for a 6890;
• The 6890 can amplify or reduce its analog output as part of the method settings, so the "constant" that I would determine for my instrument could be different from the constant that you determine for your instrument. Depending on the method, the constant that is determined for method A could be different from the constant for method B.

The best solution is to do a qualification exercise using the "old" way of doing the analysis (taking the analog data in mV) and comparing it to the "new" way (digital data in pA). If you collect the data both ways on the same injection, then you can get a good idea of the conversion factor between the two signals FOR THAT METHOD. If other methods use the same settings for analog range and attenuation, then their conversion factor should be the same within normal variation.

Also, sometimes a lab uses other instruments to collect the analog signal, such as an integrator. In this case, the integrator could add its own amplification or also apply its own constant to the signal. I worked with one lab who found that they had to take the pA signal and multiply it by 3600 in order to get a value that matched their integrator signal, so the conversion is not always just a factor of 10. Of course, it is easy to create a custom field (such as "Integrator Area) right in the software to help make the comparison. To go further, a derived channel could also be created that would multiply the digital data points by some factor in order to obtain a channel of data that you should be able to integrate with the existing method.