Re: HP 3465A offset when input shorted
|
Low leakage (< a few pA) given the input impedance for the low DC volt ranges.? A 2N4117A/PN4117A diode connected FET perhaps? ?40V breakdown and 50 mA max gate current.? Horowitz/Hill in the Art of ELectronics use them as clamps - better than the commonly available 'low leakage' diodes!
I'd check the other end of CR9 - it should be around -3.9 V as CR11 is a 3.92V zener.? Also check R66 is 91K.
Orin.
On Tue, May 24, 2022 at 12:49 PM Chris Smith < me@...> wrote: Ok well that was embarrassing. I found out what was wrong with it purely by accident.
Looks like CR9 was broken. I was inspecting the area around the high impedance after the 100K carbon comp input resistor and clamp and and noticed that CR9 looked a bit suspicious. I poked the leg with my DMM probe and it fell to pieces. Not sure of the failure mode exactly but now it's out of circuit it works absolutely fine. The original role was half of a voltage clamp which makes sense with respect to the offset measured. I assume it was damaged by over-voltage or was leaking.
Picture for reference. The blob of leg is actually dangling from the broken leg by a microscopic metal fibre.?
I checked the input scale was ok by attaching a voltage source to the input terminals and measuring Q16 gate and it was exactly what was expected from the divider arrangement. It appears to zero and work fine now!
Now I can't leave it like this as it's the voltage clamp on the input so it'll probably blow up something else without it if it gets overvoltage input.
The part is 1901-0586 which appears to be a TO18 canned diode. It's not in the usual parts cross references unfortunately. Vendor is HP, spec is derived from SM as DIODE - GEN PRP 30V 25MA.? As it's got en expensive package, it smells like it might be a low leakage type, possibly a diode connected BJT or FET. Any thoughts?
|
