HP/Agilent 34401A AC V Fail, Error 621 due to short in CR305 Zener

[VK5RC]

I rarely measure AC volts to a great accuracy (typically use a hand held DMM) so hadn't noticed a failure in my 34401 until a complete self test revealed Error 621 Fail full scale rms AC V
On AC V the display was jumping in values every .5-1sec often between 2 values e.g. 4.0V to 0.3V and back repeatedly. DC volts was working well.
Schematics are available from Agilents website, schematics 9-5 (components locator) and 9-11 (AC Circuit) are needed.
AC V uses an input buffer amp U301-A (a LF356M) and the supply voltage input into pin 7 of U301A measured only 2.5-2.8 DC V relative to AC ground (should be 15V).
A previous descriptor had identified a problem ( a short across) with the large tantalum across the supply15A, C322, but removing it and testing it out of circuit revealed no errors in C322 but a large voltage drop of 16V across the 3.3V zener CR305 used to drop the voltage from 18V or so to the 15V needed for the supply line labeled 15A.
CR305 is a BZX84C3v3 and a replacement was easily purchased and replaced with the circuit board still in the machine. Error test after replacement  passed.
Hope this is of help to others
Robert

[sync]

Interesting. I got my 34401A with the same error. Error 621 and no AC measurements. It had also a blown CR305. For testing I replaced the zener with a resistor and there was excessive current draw. I suspected the tantalum C322 too. But it was ok. The culprit was C309. A 0.1uF ceramic cap which shorted. Replaced the zener and the bad cap and AC is working again. But the reading are too low.

Edit: The readings are about 0.24% too low.

[VK5RC]

Thanks for those comments,  the only in calibration bit of equipment I have is a fluke 87v, and on a quick test was reading within 1-2mV for a 1-2V AC signal. I have an accurate DC source,  the DC side seems fine.
My real understanding of the 34401 is quite limited and I don't know what impact that modest changes of supply voltage to the AC V input buffer amp would have on accuracy,  I always am suspicious that when one component goes west,  it can easily take several others with them!

[asp]

My 34401A had the same error 621 but it was caused by a relay gone bad. it was either K102 or K103. so just keep that in mind that you can get that 621 error from that as well ( I will update this post after I will look on the board and check which one was bad ). It was actually stuck on the ON position all the time.

I did switch k102 and k103 at the time and that caused the meter to pass all tests but it never got anything useful out of measured values and when switching to resistor check it would display overload.

edit: it was k102 that was the faulty one which gave 621 error and it seems that if k103 is faulty the DMM passes all the tests but it's not usable for most part and resistor measurement displays OVLD ( overload ). Hope this helps someone.

[free_electron]

i've seen those tantalums commit suicide before. the ones feeding the true rms convertor are notorious for frying themselves.

[VK5RC]

I am pretty sure the only problem was the zener in mine, since replacing the zener it continues to work 'well' ACV DCV ACI DCI Ohms (2w and 4w), to my 'collection' I have recently added a 34461A, some quick checks reveal the only out of spec is the ACV is ~0.5% out (high), I am not that keen to recalibrate as my potential to stuff it up may be more than that small error!