Keysight B2980A series electrometer/picoammeter review/teardown

[bsw_m]

Repeated a similar test with my DIY picoammeter. I tried to keep the conditions as close as possible. I used a NK4-1 calibrator as a current source.

[Alex Nikitin]

The GPIB chip has arrived earlier today, so now the K617 is talking and listening again (though the plot below is still from the 2V analogue output measured by K2015).

The warm-up practically from cold (if +28C air temperature can be considered cold, especially in NW England  ) . 1fA/div vertical scale.

Cheers

Alex

[RLiangCN]

Hi, I noticed the range-selection relay in B2980A is Panasonic AGN2004H. My confusion is that won't this device insert too much leakage current when operating in picoammeter mode? In datasheet, it only has insulation resistance >1000MΩ.

I'm looking for a low-leakage relay to apply to my own design (for breaking the input connection and doing auto-zeroing operation). However I found the relay from COTO (9814-05-00) can leak a few pA when my input is open. COTO say this device has IR more than 1TΩ


Some datasheets:
AGN2004H datasheet https://api.pim.na.industrial.panasonic.com/file_stream/main/fileversion/477
9814-05-00 datasheet https://cotorelay.com/wp-content/uploads/2014/09/9814-9852_series_reed_relay_datasheet.pdf

[bsw_m]

Hi, I noticed the range-selection relay in B2980A is Panasonic AGN2004H. My confusion is that won't this device insert too much leakage current when operating in picoammeter mode? In datasheet, it only has insulation resistance >1000MΩ.

Some time ago I'm wrote about these relays: https://www.eevblog.com/forum/metrology/plagiarize-keysight-b2983/msg4368160/#msg4368160

[Kleinstein]

With the leakage specs there is often a large difference between the actual leakage and the specs. The specs are kind of limits on how far the parts are checked. Even the typical specs are no really typical, but more a limit for batch checks and including plenty of margin to avoid recalls / rejects.

[MK]

The automated testing also has to consider the length of time it takes to settle with very low currents to be measured, they do not have the time in production test.