Teardown: The Tektronix P6042 current probe is a classic
The architecture of the probe is clever. The powdered-iron core clamps around the wire-under-test. This closes a magnetic circuit. There is a Hall-effect sensor installed into a gap in the core. The Hall sensor will work at dc. However, it is not very fast, in the range of a few kHz. The differential output of this Hall sensor is amplified by a discrete transistor circuit, and that output goes to one side of a coil on the iron core that responds to fast ac signals. The output of that coil, now a combination of the ac and dc components of the measured current, goes to a second discrete transistor amplifier. The 50-ohm output of that amp goes to the BNC connector on the front panel. The dc output to the coil bucks the dc magnetic field it is measuring. This feedback improves the dc range of the probe and linearizes the output. The range attenuator circuit is not shown, nor is the degaussing oscillator and probe-open circuits. The complete P6042 manual is available online, including calibration, schematics, and theory of operation.
There were no IEC ac cords back in 1969. The ac cord stows in the back of the unit. The gauge of the wire is much smaller that the 18ga IEC cords on newer equipment. One downside is that the cord is not easily replaceable.
With the ac cord unspooled, you can see the fuse holder that you can switch to change the input voltage from 115 to 230VAC, or accommodate variances in these voltages. At the upper right of the case cutout is one of two 2N3767 T0-66 transistors used in the power supply circuit.
The probe unit has a slider to allow you to clamp it around a wire. The probe does not have a connector to the main unit, its hard-wired. Each probe is matched to resistors inside the unit that the assemblers selected to match the particular probe. The little white arrow at the tip shows the direction of current that will result in a positive-going output on the BNC connector.
When you unscrew the thumbscrew on the bottom left of the front panel, you can slide the entire chassis out of the case, with the ac line cord still attached.
The rear of the unit has the power transformer on the right, with a sheet-metal electrostatic shield bent up to protect the sensitive circuitry from interference. In the middle of the panel is the fuse holder and wiring block that lets you change between 115 and 230V input. On the left side of the back panel are the two T0-66 metal can transistors that provide final regulation for the +16V and −16V power supplies. Below those are part of the main circuit board with the large “garbage can” aluminum electrolytic capacitors for the power supply. These are the worst reliability components; they dry out over time. Always look for bad capacitors when you are troubleshooting a broken unit.
The backside of the front panel has the metal can for stowing the probe on the left. The probe wire comes in above it, and is routed to a small interface PCB with factory-selected components that match the particular probe. Underneath it is the dc-offset potentiometer. In the center is the large rotary switch attenuator you use to change range. The switches and pots can often be renewed with a spray or droplet of DeoxIT contact cleaner. It’s common for the wires inside the probe cable to break right where it enters the unit. If you can get the output to jump or start working by flexing the cable at this point, you need to cut that last 8 inches off the cable and re-solder it to the unit. Note the two small coax cables inside the probe main coax cable have markings. One has a tracer line on it so you know which is which.
This is what $4100 of mechanical and electrical engineering looks like. Jim Williams loved Tek equipment. It’s easy to see why. Not only did Tek have very high-performance circuitry, but the mechanical design was equally breathtaking. The main PCB has the power supply circuit at the lower right. The rest of the PCB has the two amplifier circuits and the degaussing oscillator. There is a front-panel momentary switch you depress to send an exponentially-decreasing oscillation to the same coil in the probe used to measure ac current. This reduces any residual magnetism in the core, especially if a large current pulse in the test wire has driven the core into saturation.
A plan view of the unit shows three gray dual-can heat-sinks used to keep two transistors isothermal. The Hall-effect amplifier is a five-transistor hybrid in a 12-pin metal can. It is located just above and left of the two blue power resistors towards the bottom right. The main PCB mounting screws at its corners connect the metal chassis to a ring of copper around the periphery of the PCB. This gives an RF-quality ground to the PCB circuits.
The P6042 compares well to the AM503A current probe, also old Tek equipment. Don’t laugh at my fire extinguisher. You should always have one handy when you work on power circuits, along with a pair of safety glasses. Note the extinguisher is not a dry-chemical type, it’s chlorofluoro bromomethane. You don’t want white powder ruining your circuits. This extinguisher might kill you from the fumes, but at least your circuit will look good when they put it in your coffin. A Kikusui PLZ303W electronic load provides a staircase current draw from my lab supply. A LeCroy LC684DXL 1.5GHz oscilloscope displays the output of the P6042 (green) as well as a pair of AM503A current probes (blue, orange). The AM503 series has removable probes, but otherwise very similar to the probe in the P6042.
Note the green trace at the top from the P6042 has much less noise than the blue and orange traces from the pair of AM503A current probes. This is a good reason to have a P6042, as Jim Williams attested. I thought the P6042 had a gross measurement error until I read the front panel and realized it wants you to set the scope to 50mV/div, while the AM503A units work with a scope sensitivity of 10mV/div. The current pulse is 2A, 1A, and 0A. The time-base is 0.5ms/div. The noise difference was reduced when I limited the scope bandwidth to 200MHz, but it was still quite noticeable.
For now, I am just delighted to have three working current probes. I also have a Tek A6303 200-ampere probe for the AM503A. That way I can measure the cranking current on one of my Harleys. Decades ago, I told a senior engineer that I had a hard time understanding switching power supplies. He told me that I needed to put current probes on every node of the circuit, since it is impossible to understand a reactive circuit unless you know both the voltage and the current and the phase relationship between them. That is a lot easier now that you can simulate power circuits in Spice, but you really have to measure real circuits on the bench to see how a real circuit behaves. This is why I treasure my Tek P6042 current probe.
- Using current probes to estimate E-fields
- The home lab of Paul Rako
- Be it ever so high-tech, there's no place like home