Post by thread:Fluxgate Compass

Hello all, I'm trying to make a simple fluxgate compass using a PIC. I constructed a fluxgate sensor by taking a 1 inch OD toroid and placing two windings of abt 150 turns each around the toroid in the normal manner ( wound through the center ). These two windings are wound in phase and wired in parallel. I did this to simplify the task of winding with such a length of wire. Then, I wound two windings around the toroid at 90 deg to each other, wound around the outside of the toroid. These contain about 75 turns each. I have tried many different means of exciting the larger, main winding and nothing seems to get it to act like a second-harmonic fluxgate should! I tried using the PIC to generate a 10KHz square wave. I then buffered this with a power transistor and ran it through the main winding at a peak current of abt 1 amp. This caused me to get a continuous 300mv 10KHz signal out of the outside windings. I tried lessening the current to abt 200 mA peak and I just got less of a signal out. This output signal stayed the same amplitude regardless of the strength of the external static magnetic field. ( Which I changed from ambient to extreamly strong by varying the distance and orientation of a super strong permanent magnet from the sensor ). I then tried putting 60Hz AC from a step-down transformer thru it. This cured the problem of continuously receiving a signal out the windings. However, I now needed to bring my very strong magnet right up to the winding to see any signal out at all! In fact, with abt 100,000 times amplification on the output, I had an ambient 80mV output unless I brought the magnet within 2 inches, in which case the signal shot up to 200mV and would peak at 2.5 V when the magnet toutched it. This signal, however, was at 60Hz, not 120Hz as the output of a 2nd Harmonic fluxgate should be. I also tried this with 400Hz AC at several hundred mA and got a continuous signal out, similar to the 10KHz attempt. Please help! I'm sorry in advance for writing so long a message. Thanks, Sean

On Sat, 5 Jul 1997 23:22:00 -0400 Sean Breheny <spam_OUTsbrehenyTakeThisOuTCSRLINK.NET> writes: >Hello all, > >I'm trying to make a simple fluxgate compass using a PIC. I >constructed a >fluxgate sensor by taking a 1 inch OD toroid and placing two windings >of abt >150 turns each around the toroid in the normal manner ( wound through >the >center ). These two windings are wound in phase and wired in parallel. Assuming enough drive voltage is available, wire them in series. You want as much magnetic flux as possible to saturate the core. Wiring the primary in parallel only increases the current capacity, which you don't need. {Quote hidden}>I did >this to simplify the task of winding with such a length of wire. Then, >I >wound two windings around the toroid at 90 deg to each other, wound >around >the outside of the toroid. These contain about 75 turns each. I have >tried >many different means of exciting the larger, main winding and nothing >seems >to get it to act like a second-harmonic fluxgate should! > I tried using the PIC to generate a 10KHz square wave. I then >buffered this with a power transistor and ran it through the main >winding at >a peak current of abt 1 amp. This caused me to get a continuous 300mv >10KHz >signal out of the outside windings. I tried lessening the current to >abt 200 >mA peak and I just got less of a signal out. This output signal stayed >the >same amplitude regardless of the strength of the external static >magnetic >field. ( Which I changed from ambient to extreamly strong by varying >the >distance and orientation of a super strong permanent magnet from the >sensor ). The purpose of the primary winding is to periodically saturate the core. If you aren't applying enough drive to do this then there won't be any nonlinear effect to detect DC magnetic fields with the other coils. Monitor the primary current and make sure it sharply increases at some point during the drive cycle. If it is a nice linear triangle wave (with square wave drive) it isn't enough. Obviously the choice of core material is important. The signal from the secondary coils should be processed with a synchronous detector. This could be done with PIC software most likely. {Quote hidden}> I then tried putting 60Hz AC from a step-down transformer thru >it. >This cured the problem of continuously receiving a signal out the >windings. >However, I now needed to bring my very strong magnet right up to the >winding >to see any signal out at all! In fact, with abt 100,000 times >amplification >on the output, I had an ambient 80mV output unless I brought the >magnet >within 2 inches, in which case the signal shot up to 200mV and would >peak at >2.5 V when the magnet toutched it. This signal, however, was at 60Hz, >not >120Hz as the output of a 2nd Harmonic fluxgate should be. > I also tried this with 400Hz AC at several hundred mA and got >a >continuous signal out, similar to the 10KHz attempt. > Continuous signal out probably means capacitive coupling between the windings. This will likely balance out in the synchronous demodulator. Building the unit with a grounded foil shielding layer could also help. The sudden increase with the magnet really close is consistent with not enough drive, needing help from the strong permanent magnet to saturate the core. With the same peak current, the lower the frequency the more likely the core is to saturate. Thus the better results at 60 Hz than higher.

At 12:53 AM 7/6/97 -0400, you wrote: Thank you for your response. I do have a couple of questions, though: >Assuming enough drive voltage is available, wire them in series. You >want as much magnetic flux as possible to saturate the core. Wiring the >primary in parallel only increases the current capacity, which you don't >need. > I don't understand why two windings, in phase and wired in parallel would not increase the flux. In fact, I doubt that I want to wire them in series because they are wound starting and ending at the same location on the toroid and both go around the whole circumference of the toroid. {Quote hidden}> >The purpose of the primary winding is to periodically saturate the core. >If you aren't applying enough drive to do this then there won't be any >nonlinear effect to detect DC magnetic fields with the other coils. >Monitor the primary current and make sure it sharply increases at some >point during the drive cycle. If it is a nice linear triangle wave (with >square wave drive) it isn't enough. Obviously the choice of core >material is important. The signal from the secondary coils should be >processed with a synchronous detector. This could be done with PIC >software most likely. > I have about 150 turns in each winding. The color code on the toroid is green. (I don't have specs on the toroidal core, so I don't know permeability or core material ), but shouldn't an amp of current produce enough flux to saturate it? At one point, when I had abt 2.5 amps going thru it, It was producing an audible sound at the excitation frequency and was getting hot ( since the winding has a resistance of only abt .1 ohm, I assume that this was hysteresis heating, not resistive heating ).

Sean Breheny wrote (and in response to Mike Keitz' answer): > I don't understand why two windings, in phase and wired in parallel would > not increase the flux. Because you can visualise them as winding of one piece of wire of double the thickness. Nice to pass MORE current through if you can afford the drive current. Usually however, you want less current but can afford a higher voltage as we are talking in the few ohms range, if that. > In fact, I doubt that I want to wire them in series because they are > wound starting and ending at the same location on the toroid and both go > around the whole circumference of the toroid. So they should. Your aim is as many TURNS as possible is it not? Series connection = more turns. >>The purpose of the primary winding is to periodically saturate the core. >>If you aren't applying enough drive to do this then there won't be any >>nonlinear effect to detect DC magnetic fields with the other coils. .. etc. Hint: if you want to saturate the core, use as narrow (minimal cross section) as possible. If you want to expose it to as much external field as possible, it should be as large as possible. Think of a ladies wedding ring in shape. Now the sense coils: They should be wound not around the ring of the toroid, but flat over it, and there are two wound at right angles. A coil with its axis coincident with the toroidal axis will sense virtually nothing. (It might just be possible to use three sense coils with a really weird shaped toroid!) >>Monitor the primary current and make sure it sharply increases at some >>point during the drive cycle. If it is a nice linear triangle wave (with >>square wave drive) it isn't enough. And if it's mostly flat-topped, the frequency is too low for the winding inductance. >>Obviously the choice of core material is important. > The color code on the toroid is green. It's arguable whether it should be a ferrite. Probably should be ferrous laminations, strap wound or even coiled wire itself. > (I don't have specs on the toroidal core, so I don't know permeability of > core material ), but shouldn't an amp of current produce enough flux to > saturate it? Without specs, including material, size and cross-section, it is quite impossible to surmise. > At one point, when I had abt 2.5 amps going thru it, It was producing an > audible sound at the excitation frequency and was getting hot You mean the WINDING was getting hot? > (since the winding has a resistance of only abt .1 ohm, I assume that > this was hysteresis heating, not resistive heating). I^2*R = 2.5 x 2.5 x .1 = 0.625W is a perceptible amount of heat. If the winding resistance was only .1 ohm, I fancy you were using wire thicker than necessary and should have opted for more turns, thinner wire. All of this equally "rule of thumb" of course. This is NOT my speciality! Cheers, Paul B.

Hi There, Take a look at Popular Electronics Magazine (August 97). It uses a single chip (well maybe two) to buld a compass. Based on Hall Effect, it looks interesting. Let me know what you think. Bye for now Eric At 11:22 PM 7/5/97 -0400, you wrote: {Quote hidden}>Hello all, > >I'm trying to make a simple fluxgate compass using a PIC. I constructed a >fluxgate sensor by taking a 1 inch OD toroid and placing two windings of abt >150 turns each around the toroid in the normal manner ( wound through the >center ). These two windings are wound in phase and wired in parallel. I did >this to simplify the task of winding with such a length of wire. Then, I >wound two windings around the toroid at 90 deg to each other, wound around >the outside of the toroid. These contain about 75 turns each. I have tried >many different means of exciting the larger, main winding and nothing seems >to get it to act like a second-harmonic fluxgate should! > I tried using the PIC to generate a 10KHz square wave. I then >buffered this with a power transistor and ran it through the main winding at >a peak current of abt 1 amp. This caused me to get a continuous 300mv 10KHz >signal out of the outside windings. I tried lessening the current to abt 200 >mA peak and I just got less of a signal out. This output signal stayed the >same amplitude regardless of the strength of the external static magnetic >field. ( Which I changed from ambient to extreamly strong by varying the >distance and orientation of a super strong permanent magnet from the sensor ). > I then tried putting 60Hz AC from a step-down transformer thru it. >This cured the problem of continuously receiving a signal out the windings. >However, I now needed to bring my very strong magnet right up to the winding >to see any signal out at all! In fact, with abt 100,000 times amplification >on the output, I had an ambient 80mV output unless I brought the magnet >within 2 inches, in which case the signal shot up to 200mV and would peak at >2.5 V when the magnet toutched it. This signal, however, was at 60Hz, not >120Hz as the output of a 2nd Harmonic fluxgate should be. > I also tried this with 400Hz AC at several hundred mA and got a >continuous signal out, similar to the 10KHz attempt. > >Please help! I'm sorry in advance for writing so long a message. >Thanks, >Sean >

At 06:48 PM 7/7/97 -0400, eric naus wrote: >Hi There, > >Take a look at Popular Electronics Magazine (August 97). >It uses a single chip (well maybe two) to buld a compass. >Based on Hall Effect, it looks interesting. >Let me know what you think. > If it's using the sensor with a small magnet, it's a re-re-recycled project. A couple years ago, they featured a compass (which I built), then the featured a talking compass that used a nw-unavailable ISD chip. This sensor is not super-reliable. They -- or Electronics Now -- had a project that featured hacking into Radio Shack's flux gate compass (~$50?). Chris ====================================================================== Christopher Zguris - .....czgurisKILLspam@spam@interport.net HSTA, HRCA, AMA 2. When you're flying through the air, remember to roll when you hit the ground. - Safety Tips for the Post-Nuclear Existence ======================================================================

Hi there again, The article uses the 1490 digital compass IC. Further it states that it contains a miniature rotor on jewel bearings surrounded by 4 solid state Hall effect IC's. Also the device must be held in its verticle position. Cost of the IC is $15.00. I'd be interested in hearing if you get your circuit working. Bye for now Eric At 07:00 PM 7/7/97 -0400, you wrote: {Quote hidden}>At 06:48 PM 7/7/97 -0400, eric naus wrote: >>Hi There, >> >>Take a look at Popular Electronics Magazine (August 97). >>It uses a single chip (well maybe two) to buld a compass. >>Based on Hall Effect, it looks interesting. >>Let me know what you think. >> > >If it's using the sensor with a small magnet, it's a re-re-recycled project. >A couple years ago, they featured a compass (which I built), then the >featured a talking compass that used a nw-unavailable ISD chip. This sensor >is not super-reliable. They -- or Electronics Now -- had a project that >featured hacking into Radio Shack's flux gate compass (~$50?). > > >Chris > > ====================================================================== > Christopher Zguris - czgurisKILLspaminterport.net > HSTA, HRCA, AMA > > 2. When you're flying through the air, remember to > roll when you hit the ground. > - Safety Tips for the Post-Nuclear Existence > > ====================================================================== >

> Tim Kerby wrote: > > > There aer some details in the picstart design contest book about > > this - > > they use a specific chip. ETI (magazine) also used a chip in a > > recent > > implentation but I never got a copy of the mag as they sold out so > > quick. > > A few people mentioned that I look in the Picstart Design Contest Book. > There is nothing about it a http://www.microchip.com nor in http://www.altavista.com. > Who printed or distributed it and is it still available? > > Thanx > Louis Marquette > Hi Louis, The Picstart design Contest book was published by Microchip in 1993. One article for a possible design contest entry is a digital compass using a PIC 16C71, one 16 X 1 LCD Display, and a dual analog Hall Effect Compass Sensor (Dinsmore 1525) which fed into AIN1 and AIN2 on the PIC 16C71-16. Hope this is off some help, but I,ve no idea where to obtain the book now. -Sandy.

Hi Its a microchip publication. Just email .....literatureKILLspam.....microchip.com and ask for document DS30229B Picstart Design Contest Application Brief Notebook and give them a name address and phone book and they should send you a copy for free. You will want to look at the Digital Compass on page 3-3 which uses a Dinsmore 1525 DualAnalogue Hall Effect Sensor and also page 3-4 for a direction seeking toy car which uses another dinsmore hall effect chip. Dinsmore should have a webpage somewhere. Address is as follows Dinsmore Instrument Company Electronic compass 1525 ($35) 1814 Remell St Flint, MI 48503 (313) 744-1330 Tim At 20:12 15/07/96 -0400, you wrote: {Quote hidden}>Tim Kerby wrote: > >> There aer some details in the picstart design contest book about >> this - >> they use a specific chip. ETI (magazine) also used a chip in a >> recent >> implentation but I never got a copy of the mag as they sold out so >> quick. > >A few people mentioned that I look in the Picstart Design Contest Book. >There is nothing about it a http://www.microchip.com nor in http://www.altavista.com. >Who printed or distributed it and is it still available? > >Thanx >Louis Marquette > > ------------------------------------------------------------------ Personal Web Pages: http://web.ukonline.co.uk/members/tim.kerby/ PIC Site: web.ukonline.co.uk/members/tim.kerby/pic/ The PIC Pages are under construction and I am looking for projects ------------------------------------------------------------------