EEVblog 1476 - Keithley 515A Wheatstone Bridge TEARDOWN & TUTORIAL

• EEVblog 1476 - Keithley 515A Wheatstone Bridge TEARDOWN & TUTORIAL
  Posted by EEVblog on 03 Jun, 2022 12:39
• Teardown of the Keithley 515A Megohm Wheatstone bridge, plus a tutorial on how Wheatsone bridges work and their applications.
  Manual & Schematic: https://download.tek.com/manual/515A(Model515A).pdf
  1960's Daven precision resistor and attenuator catalog: http://www.technicalaudio.com/pdf/Daven/Daven_attenuator_switch_lab_Catalog.pdf
  Keithley electrometer teardown:
  
  00:00 Keithley 515A Megohm Wheatstone bridge
  03:50 - The Magic smoke test
  04:00 - Zero check test
  04:40 - Standardise calibration step
  05:30 - Range calibration
  06:21 - Measuring a 200M resistor
  09:41 - Confirmation with a Keithley DMM7510 7 1/2 digit multimeter
  10:37 - Weatstrong Bridge Tutorial
  14:55 - Gaurd traces for high impedance measurement
  16:22 - Teardown
  28:25 - OUCH!
  
  
• #1 Reply
  Posted by PartialDischarge on 03 Jun, 2022 13:16
• I find it surprising that that resistor changes that much with only a 90V increase. The voltage coefficient is usually very low in that range, maybe some ppm per volt
• #2 Reply
  Posted by free_electron on 03 Jun, 2022 16:06
• That input fet is made by https://www.linearsystems.com/
  i (probably) have the last 3 in existence   I repaired two of these years ago and they only had 5 of these fets in stock and i bought them all. 2N5904 or 2N5907. there is a whole range of these fets. depending on what model electrometer you have they use one or the other.
  
• #3 Reply
  Posted by EEVblog on 04 Jun, 2022 00:51
• Quote from: PartialDischarge on 03 Jun, 2022 13:16

  I find it surprising that that resistor changes that much with only a 90V increase. The voltage coefficient is usually very low in that range, maybe some ppm per volt
  

  
  Yeah, I was a bit surprised by that. Maybe some non-linearity in the old 515.
• #4 Reply
  Posted by Kleinstein on 04 Jun, 2022 07:44
• For the bridge in storrage for a long time in a possibly slightly damp basement it may take some time to dry up. So nonlineartiy with the 515 is a possible issue. This would effect the reference resistor(s) not the switchable decade resistors.
  
  Another point could be some bias current from the null detector. This could be visible when using an external 0 V bridge voltage.
• #5 Reply
  Posted by free_electron on 05 Jun, 2022 21:19
• -cringe- triaxial bnc ... it's threaded so it should be a TNC  (B = bayonet , T = Threaded )