TMR Based Current Probe
Using a Tunneling Magnetoresistance sensor to create a DC-5 MHz current probe
Why?
Current probes are a very useful piece of test equipment. Usually, I would use a current transformer when measuring AC currents, but CTs don't work for DC currents. The TMR sensors can work with DC into the high-frequency range, and a DIY option would make them accessible to hobbyists. Professional options cost thousands of dollars, so a low-cost alternative would be incredible!
I first became curious about TMR sensors after reading this page on the TDK website: https://product.tdk.com/en/techlibrary/productoverview/tmr-angle-sensors.html. There are some clear benefits of these chips over Hall-effect or AMR devices, and since I was looking for a new project, I thought making a useful device out of these relatively new devices would be a perfect fit!
It Works!
The probe works! After finishing soldering and cleaning the boards, I did some preliminary tests and everything seemed to work as expected. I was easily able to measure transient pulses by shorting a DC power supply, and DC currents as well. Eventually, I even tested the probe with my AWG and found the bandwidth to be around 5 MHz. I am proud that the flip-flop circuit I used for the relay and 1x/10x gain setting works exactly as expected. I currently do not have access to a high-frequency current standard, but when I do, I will update the page!
Version 2
I would like to improve upon this project. As outlined below, there are many problems preventing it from achieving the performance that the sensor technology is capable of. I need better test and measurement tools to characterize the probe, and I also need to better design the analog section. This is going to require a full redesign!
Topology
I used TMR-based sensors from Multi-Dimension. While they are a company that specializes in this kind of sensor, it is difficult, as an American, to get them to ship me samples. However, after a long email chain, I succeeded. The chip works by integrating two current-sensitive resistors into a classic Wheatstone bridge, outputting a differential signal depending on current fields. I picked the TMR2583, which is their mid-spec chip sensitive in the Z-direction. The data sheet is a little sparse, but from some research, I was able to create a circuit that works with these chips.
Schematic
The circuit is a simple instrumentation amplifier using the AD8421. Something of note is the isolated DC-to-DC converter I used to get my 12 V rails. These isolated, dual-output devices are very convenient for split power rails used by many op-amp circuits. Unfortunately, due to tariffs and shifting industry needs, they are becoming expensive and hard to find. The circuit I use to control the relay is a D-type flip-flop that acts as a latching circuit for a push button. The relay controls the gain of the IA for hopefully 10x and 1x modes in practice. For revision 2, I would like to add a microcontroller for auto-zero and to simplify my design.
Circuit and PCB Design
I created a quick 4-layer board design using KiCad. I chose four layers so I could keep the layout clean and compact. The boards were ordered on JLCPCB and populated by hand using a regular iron and hot air. The KiCad-to-JLC-to-home pipeline really does make these projects a breeze.
PCB
Boards Arrive
The boards arrived from JLCPCB and were cleaned, populated, then cleaned again. I made two small footprint errors, flipping one of the regulator pinouts and picking the wrong package size for the DC-to-DC converter. Both ended up being fixable issues; the regulator was mounted upright, and the DC-to-DC converter was soldered floating a few mm above the rest of the board.
