Portable Magnetic Detector

This design is based upon the content of this page [http://danyk.cz/hall_en.html] and a video by MRAMAKERs (youtube # 4Xvo60A-Kt0) describing a common component found inside older junked brush-less computer fans, and converting that component into a useful portable tool for Technicians and Hobbyists.

That component is the AH276 complementary Hall effect latching Fan commutation controller Integrated Circuit inside a dirty old brush-less computer Fan. The reader gets to break apart the non-functioning Fan.

Power Supply Computer Fans with two red and black wires (12V) would likely be a candidate for this hack. Millions of these were used in AC power supplies, inside desktop and tower cases, in the 1980's to this present day.

This constructable will add two resistors, a Bi-color LED, an SPST switch, a 9V battery and clip to an existing Fan controller board to convert that board into a portable magnetic detector.

There is a video by Easy One( youtube # _i0rNIoo5Zk) that shows the fan disassembly and the AH277 component, similar to this presentation, using two separate RED and GREEN LED's.

datasheet for the device: https://www.diodes.com/assets/Datasheets/AH276.pdf

The first image is a disgusting buildup of lint and dirt on a computer fan. I will deny that this is my image, but I admit finding similar conditions at my house and in my shop.

The second image is of a Fan, taken from a legacy 350W ACDC Power Supply from a Tower. This is the humble brush-less 12 Volt 0.18 mA 4" common Fan.

I pealed the model sticker off to reveal the shaft cover, and then pried that plastic cap off, to reveal the shaft "locking" nylon washer. Your discovery experience will vary.

The fourth image is the use of a small flat screwdriver to pry out of the locking washer.

The fifth image shows the coils, the driver board, and in the center, the detector IC.

The sixth image shows the actual AH276 Hall Effect IC . Pry that whole board away from the fan housing; mine was glued, and I broke that mount plastic.

Discard the housing, turbine fan blade, bushings, label and just retain the board.

Un-solder the center coil 3 contacts and remove that plastic or metal coil assembly. (7th image)

The last image jumps ahead of this step with Bi-LED and two resistors attached, *but the objective is to extract the driver board itself (as shown).

Your experience will vary, but the last 3 FAN's from different manufacturers that I have snapped and cracked open or apart, all look identical.

Referring to first image, the areas circled are existing contact points on the brush-less driver board.

Refer to the attached schematic for the wiring concept.

I used a Bi-LED 3mm BIVAR 3BC-F and two 470 Ohm 1/4W 5% resistors.

I mounted the 2-lead Bi-LED into orientation so that the GREEN cathode faces PIN 2 of the AH276 IC,

and the RED cathode side faces PIN 3 of that Hall Effect IC.

When PIN 2 goes LOW, the IC has sensed 'South' face of external magnet, and if that magnet position is reversed so that 'North' is facing that IC, PIN 3 goes LOW. Pins 2 and 3 are complementary and toggle in opposite polarity, ideal for use with a 2-lead Bi-color LED (old school style).

The AH276 is reverse polarity protected. The circuit will work with 3.5 Volts, and up to 15 Volts.

The AH276 can sink 300+ mA potentially driving some bigger turbine fan blades. The AH277 can sink 500 mA on larger fans, but the device pinout and function is the same as AH276.

I have confirmed operation with the attached schematic on three pulled boards (I need three detectors).

Apply 9 Vdc in correct polarity to the boards; I left the original red and black 12V Fan wires attached for this power. I added an SPST switch in series with the red (positive) wire. On two samples, I used MOMENTARY buttons.

At power up, the Bi-LED can be either RED or GREEN, but never both ON, and never both OFF.

Approach the face of the AH276 with a permanent magnet with either side; if the LED states do not change, withdraw the magnet, turn the magnet around 180 degrees, and approach the IC again, directly toward the outer face of the AH276.

Most Fridge magnets are 30-50 Gauss, and work well with these Hall Effect Detector IC's. Verify your test magnets poles with a compass; 'opposites attract' method.

Once verified, you can put the project into a case. You could mark Green as SOUTH, Red as NORTH.

In my designs I added 1 extra feature; a GROUND CLIP, and a 24" length of black or green wire to an alligator style clip, from the - battery side. This connection is used for DC Solenoid Relay coil testing.

First image shows parts to be discarded, and target parts to retain.

Second image is same as first without a markup.

One of the test objectives for this type of cheap tester is to indicate if a RELAY coil is energized.

In Automotive Industrial Vehicle Electrical systems, it would be advantageous to discover which relays are energized when testing those systems. The advantage of this tester is that it is visual, and can be used in a very noisy Mechanical Shop.

The ground clip attaches to the Vehicle Ground or Battery Negative (if that is the return terminal), and the Hall Effect "probe" or AH276 IC "face" is brought close to the RELAY under testing. That UUT (unit under test) relay is operated, and the detector should indicate magnetic state changes when the solenoid is activated. This tester will not test the relay contacts. The device is convenient with many relays in a large housing for an Industrial Truck, with loud shop noises to obscure the tell-tale relay 'click', and this tester is handy to indicate that the coil is energized. Many Relay failures are attributed to open coils caused by operational vibration.

Refer to images for operation examples showing South (GREEN) and North (RED) detection.

This modified circuit should draw around 40 mA from a 9 Vdc (type 1604) battery in any state.

AH276 sensitivity is barely usable for some smaller coil relay solenoid coil detection. On standard Form1A 12 Vdc Tyco relays, I can detect activation from 5-15mm away from the relay body. With a micro miniature 5 Vdc type, I could not detect state changes.

Share and add your comments for

• improvement of sensitivity (the newer Allegro devices?)
• ratiometric linear method in the Hall Effect devices, decoupled from Vcc
• experiences in disassembly of fans,
• cases, enclosures
• probe on the end of cable

I enclose images of that first sensor, completed. I have two other DIY HE sensors.

UPDATE 28JAN: Built 10 using the discarded "slot fans" with faulty bearings, from older Desktops.

Example of one is in last four images, using the Brush-less switch PCB itself, discarding the coils and guts, and

adding two LEDS and one resistor (connected to both LED ANODES and 9V switched).