Agilent 34401A 1% error on AC volts?

[mzzj]

I'm not sure if this belongs more to repair or metrology but anyways..

34401A failed on AC volts calibration at 50hz with really big error (1% low on all voltage ranges)
Adjustment routine runs fine until adjustment point at 10Vac 10hz give error 733 "AC low frequency convergence failed"
Manual wont tell more than that about error 733.

Any ideas or has someone actually stumbled on the same problem?

Oddly the VAC range measures ok-ish  at 10Hz with less than 0.1% error and 1kHz is spot on.
There is a dip around 50Hz on all ranges and it is about 1% of reading. 

[strawberry]

seems high pass filter
oscilloscope and ac amplifier up to RMS converter

[Gyro]

I don't know the internals of the 34401A in any detail but is it possible that (a section of) a DC mode line frequency rejection filter is wrongly switched in?

[Kleinstein]

The 34401 does not have an extra line rejection filter. The CMRR is just from using multiples of the power line period for the integration.

The rather large error looks a bit like a defect of some kind, though a slightly odd one.
The lower frequency end could be effected by AC coupling capacitors and the filter capacitor at the RMS converter.
There filter at the RMS converter has 2 settings and the switching there could be wrong or defective.

The amplifier / divider part could be checked with the scope and a low frequency (e.g. 20 Hz) square wave.
Depending on the range different parts of the amplifier are used or not. So some defects may only effect some ranges.

[mzzj]

Measured at 50hz and 1khz the voltage response up to rms converter seemed good (measured over R315)
Things got bit weird after the rms converter: error seemed to do something with Q301, turns out it gets 0v or 2.7v gate drive depending if <200hz filter is selected or not.
2.7v gate drive didn't make sense, U304 (DG4011) was driving the output pin from 0 to 15 volts. -->desolder the Q301 MTD3055VL mosfet.
G-S resistance measures 2-3kOhm--> toasted-->replace with  physically larger  TO-220 3055VL mosfet.

Seem to work fine now, now I'll just have to take this back to cal lab and run full calibration/adjustment cycle 

[strawberry]

that TH part will definitely fly off by first chance including PCB traces

[mzzj]

that TH part will definitely fly off by first chance including PCB traces

Fairly large pads for the original TO252 and I'm not planning to use this in cement mixer. 

[strawberry]

been there