Introduction: Portable Bar Graph Magnetometer

About: Electronics Technologist (44th year)

Expert Level Project

Uses Honeywell SS49E Linear Hall Effect Detector, LM3914 Linear Bar Graph in dot mode, 10 LEDs in a Bargraph DIP package, 78L05 small power regulator, Switch, Battery clip and 9V DC (1604) Battery.

This project creates a handheld probe for testing Electromagnetic relays, finding shorts in cable runs, detecting magnets (and their poles), finds issues in Automotive Ignition packs, and tests Electronic Fuel Injector solenoids. The detection of Auto Engine fuel injector system failure was the motivator for this "pen style probe" operated by battery.

In that other project, only one comparator threshold driven LED is used, but in this project, the Hall device's linear output is displayed in both magnetic pole polarities, and can display (approximate) field strength.



1x 1K POT, mini, PCB mount, 1turn 1x 50K POT, mini, PCB, 1turn

1x SB10101K LED Bar Graph DIP display, 10 RED LEDS approx.median 10mA forward current, 20 pins

1x SS49E HALL EFFECT Detector (3 pins)

(I used DIPsockets and pin Dupont headers but this is optional, on Vero PC PAD PTH board)

1x 78L05 5V regulator , 3 pin (or LM309)

1x 9V Clip, Battery and SPST latching switch

1x Schottky diode on the 9V input side of the regulator (optional)

at least 1 DC capacitor on the 5V rail to ground return, 10uF @ 25Vdc

various cables and jumps, and creative probe ideas like repurpose of a dried out PEN tube as case

You may need a DMM to calibrate, and a Soldering Iron if Vero or perf board is used to mount

Step 1: Construction

refer to Schematic

Typical (ANSI 1604) Battery supplies 9V power to 78x05 three terminal linear regulator, that provides power with about 40-50mA current at 5V DC, to the rest of the circuit. In my design , I also incorporated a Schottky 0.2V 1A diode in series with a latching power switch on the 9V battery clip itself.

The 5V also powers the SS49E detector. The output of the detector at rest is 2.5V or about 1/2 of the Vdd +5V supply. The output varies above and below this median level when a magnet influences the fluxgate. The deviation of output voltage is from 2.5V up to near 4.5V when a magnetic pole is detected on the 49E face 'front'. Similarly, if the opposite pole is applied to that same face, the output voltage decreases from 2.5V median down to 0.9 or so. The device has a certain linear flat response which curves as the device saturate.

Refer to the Honeywell specification sheet at

The SS49E output is presented to the input of a linear LED Bar Graph driver LM3914 IC in the 'dot' mode. Two potentiometers set high and low references as well as step and LED current limitation. This instruction will describe Calibration in a latter step, using just a DMM measuring points and adjusting the two POTs to suit. This dual pot design is also capable of calibration with an optional stage in an upgrade of sensitivity using a simple LM358 stage, as an addendum to this instruction, also in the latter pages.

All drawings will show the pinouts of the various components, and you can use a DMM on DIODE range to figure out which way to orient the SB10101 Bar Graph LEDs.

It took about 3 hours for me to construct this on Vero perf, using Teflon-Wrap-Wire #30 or solder bridges to create PCB in long rectangular "pen probe" format. I plan to mount the whole thing into a transparent tube for use in my shop, and build another probe for Auto Mechanics to use in their Repair Shop.

If I have time, I will make a *breadboard* version with DuPont jumpers, ala Arduino "Fritzing" for instruction and tutorial Class purposes.

Step 2: ​Calibration and Tests

refer to the first drawing;

The objective is that when the SS49 is powered properly with clean 5V

the output is at mid-range 'at rest' with no magnetic field near the sensor.

So the 1K POT is adjusted for 0.7V at the pin 4 point, and the 50K POT

is adjusted for the mid-scale indication with LEDs 5 and 6 about the same intensity

(ranges overlap).

Detector theory and indications are shown in the attached drawing.

The third drawing depicts two *optional* ICs, an LM358 in 2X differential amplifier mode

and a LM393 comparator for use with two LEDs in a later upgrade so that the two extra

LEDs are marked NORTH and SOUTH to indicate which polarity is facing which side.

At some point, I will lead 3 wires down to the end of the PEN case to those two LEDs

on either side of the detector, but this is far too advanced for posting here (yet).

You will notice that I have not yet soldered 4 POTs and several 100K resistors on that Input board (pending); the designs worked on breadboard prototype. The DMRR was handled with a 50K offset POT and I will post schematic of the Differential Amplifier stage in future on this page.

Step 3: Usage

I will post a video with usage examples and indications, later, to this page.

(1) I constructed the probe based on the much simpler 1 LED comparator threshold types to detect FUEL INJECTORS. I discovered that I can detect FUEL PUMP at closer ranges, and a number of other Automotive Technician test requirements in a repair shop.

(2) I then discovered that some wires carrying heavy current develop an EM Field around the wire. With help of partner switching the circuit from the Cab, I could sense or "sniff" that wire in a bundle. (Examples are) Headlamps, Rear Defogger inside glass, Heat Fan, electric air Vanes in the heater box and so forth.

(3) Some EM fields were/are overwhelming; I just backed away on the probe for tracing. The user can sense Ignition from long distances in the Engine bay; the EM Field is very strong with Spark Plug discharges when Engine is running.

(4) The other objective was to trace RELAYS.

With Automotive DC Relays, I can sense the field but just barely, and I have to get the sensor right down to the base of the Relay case, sometimes near impossible with the close spacing in junction boxes under the hood or dashboard.

I will design and implement a probe with differential Amplification 2X+similar to the example in the Honeywell application guides. LM358 and LM393 are very common Arduino experimenters parts. (pending)

The second and third images are with the 'pen probe' rubber banded to the wood splint but the pen is on a 10-12cm of cable. I built a second SS49E probe on a 1.5 meter shielded cable and the probe worked fine in the shop, to the instrument.

It can be said that this tool is non-contact safe, but only the brain of the operator makes it so.

"Engage Number one."

Pricing of parts

Not Bad for Honeywell (probably cloned counterfeit) sensors that cost $4 CDn per quantity 10 from China.

One LM3914 is $2.40 CDn, and the 10 LED Bar Graph 20pin DIP was $4 CDn.

The 78L05 small TO92 sized regulator or LM309 in larger can TO125? is harder to source

but the user can substitute AMS1117 5V SMD component for the 78L05 (600mA vs. 100mA).

The Current draw is less than 50 mA , not really a lot and the TO220 larger 7805 would be overkill ,

but that 3terminal 1Amp will work as well. The idea is to get stable steady 5V from a 9V battery.

Step 4: U[dates and Comments

I found several posts of Arduino AVR connected to SS49 with OLED display showing Bargraph. Ultimately the sensitivity is 1000+ steps on Analog Input instead of 10 steps of LEDs. I do plan to use ATTINY85 based Digispark board with 0.91 OLED as a follow up and post the schema and INO files to this page. The comment from colleague is that LED bargraphs are a bit retro, and now its the costliest component of the bunch. DigiStump clone boards are $4 CDn and the OLED i2c is also about the same with the instrument running on one 18650.

Take away: sensing small 12VDC relay coils requires sensitivity of 1-10Gauss.