The HP42329A measuring DC leakage of a capacitor
This is a resistance meter, but not the normal kind. It is made to measure very high resistance. Maybe is is better to say that it can measure the resistance from things that we call insulators. In theory theyare insulators, but if you need one in real life, you enter the zone of black magic. The area that is populated with nano-voltnuts, femtoampere hunters, micro-grabbers and electroncounters.
When you measure extreme resistance you are measuring very small currents. And for very, very small currents this is almost counting electrons. 1A is a lot of electrons. No problem if you lose a few of them while measuring. It equals 6,241,500,000,000,000,000, electrons/sec; 1 pA is still a lot of those little suckers, a quick headcount (you only have a second) tells us there are around 6,240,000 electrons passing by. And if we go deeper in the catacombs of metrology we meet 1 fA, that equals to 6240 electrons/sec.
Measuring the tiny current through a high resistor is not easy. You get into trouble with current leakages caused by contamination of the surface. A fingerprint can be enough to make the outcome of your measurement complete useless, noise sources, and stray capacitance are also hard to control. That asks not only for special gear but also for sound knowledge about how to use it.
The HP is a member of the family of high resistance meters that contains instruments like megohm meters, insulation meters, meggers and they also relate to things like coulomb meters and conductance meters. Often they are sourcemeters and electrometers. They are made for things like sheet-resistance measurement, leakage through isolators etc. It is a lab instrument. Not something you use to test the leakage of an electro motor or heater. The difference is not just the looks but also the performance and lack of protection against operator faults. Do not ask me how I found out. OK, you win, I used it to test the DC leakage of a capacitor. The inimini amount of charge that build-ed up between discharging and coupling it a few minutes later to the meter was enough to kill a switch contact. I was lucky to find out that the innerparts of a Philips switch in my junkbox, and some patience could solve the problem.
As said, it is a source meter. The test voltage is adjustable between 10 and 1000V. It measures the current that passes the resistor. So in base it is a current measuring instrument with a scale in ohm. The current must be within 50fA and 20uA. It can also be used to measure current and display it.
Most electrometers are based on a neon, mechanical or electronic choppers. But this one uses a selfoscillating parametric impedance converter made of 2 varicaps in a bridge. HP named them pico-diodes. They have a 10pA leakage current and that makes them usable as a very high resistance that can be controlled with a biascurrent to balance the bridge with the help of some capacitance. If a current enters the bridge, the balance is gone. The bridge amplifier starts to amplify this to restate the balance but it’s output is inverted, so it amplifies more and because there is some LCR and delay in the circuit it starts to oscillate. The magnitude of AC amplification is linear related to the current through the resistor. The AC signal gets rectified and fed to the analog meter.
The resistance range covers 500K upto 2×10¹⁶ Ohm or more impressive if you note it the “normal” way; 20,000,000,000,000,000 Ohm, or 20 peta-Ohm. To be short, a Lot.
Mine needed some TLC. It was used in a calibration lab that closed in the 80’s but all gear stayed there unused until it came to me thanks to someone I know.
Inside are two PCB extenders. Very nice and a huge contrast with the modern gear. Back then they promoted repairing your own gear. Today many brands try to make that as impossible as they can. Aligning is easy. You have to adjust two voltages and feed it a 200mV test voltage to set the current range. After that the start and end of the scale.
From the top, left and right the extenders
the bottom with the beautiful shielded switches
It needs a very special testlead and that was lost.
It is a coaxial cable with a graphite layer that makes it possible to bow the cable without to much tribo-electric effects. And a special “connector” that connects the outer shield to the guard circuit. This prevents leakage current over the outside of the cable. I use a piece of coax because I do not have the cable.
If you want to measure very high values you need that cable. If I use it and stay about a meter from the setup I see the meter moving in the rhythm of my chest while breading. That is why I use a shielded aluminum box. Inside is a piece of teflon to carry the DUT. Be careful, 1000V is a lot and not only can damage you but also the meter. This is a sensitive lab instrument, not am industrial Megger-like devise. (Megger is a brand that is famous for making insulation meters)
Very important !!
Meter and probeset
probe
specs