PCSGU250 Ch1 and Ch2 have unequal voltages, calibration?

I have been using my PCSGU250 about 3 times a week since i bought it in DEC09.

It has been an excellent addition to my lab, but I have some concerns about results I have seen during an experiment I have been running.

I attempted to perform a measurement on a RL circuit with total impedance of 1kOhm at DC. I ran the function generator at 1.00Vrms but noticed that as my frequency increased, the source voltage did not stay constant, and would experience considerable droop over the AC sweep I performed. The amplitude at 1Mhz did not equal the amplitude at 100Hz. I removed all circuit elements, and measured only the function generator with CH1 and Ch2. I have noticed two things:

1 At 10Vpp with dc offset set to 0 on the FuncGen, Ch1 max reads 4.56/min is -4.66 , Ch2 max reads 4.31min is -4.50
Shouldn’t this read -5/+5 for both?

  1. As frequency increases, droop occurs and the disparity increases between the source voltage and what it should be according to settings. The drop can be anywhere from a few hundredths of a volt to 2 or 3 volts when attempting to use 10Vpp with a sine source and a 1kOhm load.

Calibration was attempted to fix the issue, but did not produce successful results.

What are the frequency characteristics of the function generator under load? Am I overloading the system by using too small a system impedance?

This is the only specification for the accuracy:
bandwidth: two channels DC to 12 MHz ±3dB

I think your results are within this specification.

Do you have the latest v1.06 PcLab2000LT software installed?

Here are two example plots.
Generator output connected to the oscilloscope input with a direct cable.

Generator output voltage set to 1Vrms.
Bode plotter 100Hz to 1MHz:

Generator:
Sweep 100Hz to 1MHz.
Sweep time 50sec.
Output 5Vpp
Spectrum analyzer:
FFT Window: Flat Top
FFT Mode: Maximum