Dear Sean, At 23:52 2017-03-25, Sean Breheny wrote: >Hi Mario, > >I was trying to determine what level of fidelity you needed between realit= y >and simulation. There are a few possible complications which can arise wit= h >modeling this in simulation if you are really trying to model all of the >possible effects (most of which are ignored when these type of sensors are >used): Yes, I know that accurate simulation is practically impossibile. Just to na= me one, the sensor will effectively "drag" the wheel, and the higher the elect= ric load on the sensor, the higher the drag. This was just one example. There are others effects as well I'm aware of, s= ome are less subtle than this, others are more. But I'm not after great accuracy in this case, I'd rather want to understan= d roughly what's going on. I also know that inductance varies depending on wh= eel tooth proximity, it's not the fixed value that I measured with my meter.. b= ut I just wanted to know if, imprecisions apart (which at low speeds will not = be too dramatic), the basic idea is valid. >I will try to list them from most possibly relevant to least possibly >relevant: > >1) The real part of the impedance (a.k.a AC resistance) will be frequency >dependent due to skin effect, proximity effect, hysteresis, and eddy >current losses. It will not be the same as the DC resistance you measure >with a multimeter (except maybe below 30Hz or so). It'll be used at low frequences, so it's (approximately) ok. >2) The inductance will also show frequency-variable behavior due to >parasitic capacitance, eddy currents, and frequency-dependent core >permeability. Yes, clear. >3) The inductance will vary a bit depending on whether there is ferrous >metal near the sensor or not (i.e., whether there is a gear tooth or other >salient piece of metal which is being detected). >4) Depending on how strong the magnet inside the sensor is and how well >coupled the sensor is to external ferrous metal, and also whether you are >driving current through the sensor (this would be unusual), the inductance >could vary due to saturation All right. >So, as you can see, it's not really practical to just make a general model >for this sensor (or any component) which just captures ALL behavior - you >have to have some idea of what the model will be used for. Since most >people use these sensors just to produce pulses which can be used to count >rotations of a gear or wheel, the typical model would not include details >of the inductance and resistance. Even if it did, it would likely just be >an equivalent circuit which produced the roughly-correct frequency roll-of= f >assuming a particular load impedance, rather than a high fidelity >distributed RLC model showing full-spectrum impedance versus frequency. It's all right. But, on a side note (as I will later work on this other and someway similar thing too), the pickup of an electric guitar.. can the resistance and inductance be used to model the characteristics of it (without extreme precision of course)? The string doesn't perturb the magnetic field a lot, so to get an approxima= te (only approximate, not precise) frequency response can I model it as a volt= age source + measured resistance + measured inductance, and then add a potentio= meter, a cap or whatever, and expect approximate equivalent circuit and thus frequ= ency response? Even +/- 30% error for me would be ok, it's just to get a rough indication = of components to use, it's more to get the theory than for real practical use,= then I'd hook my oscilloscope and signal generator anyway. But I'd like, for int= uitive reasons, to know if - more complex physical aspects apart (inductance that = varies with flux, parasitic capacitance, etc..) - it's correct to assume that the = equivalent circuit of a pickup is a voltage source with in serie the measured resistan= ce and measured inductance of it. I know reality is much more complex than theory, with a lot of subtle effec= ts than in theory aren't apparent, but I still like to have an intuitive, simp= le view of things, and then add complications later. ;) With kind regards, Mario > >Sean > > >On Sat, Mar 25, 2017 at 2:22 PM, Electron wrote: > >> >> Hi Sean, >> I really just wanted to know if the "electronic" part is equivalent to a= n >> inductor* in serie with a resistor** and a voltage source. >> >> * =3D as I can measure on the real pickup with a ohmeter. >> ** =3D as I can measure on the real pickup with a inductance meter. >> >> I am very well aware of the other problems, I just aim to experiment wit= h >> impedance, filters and such, to know if I e.g. can expect the same volta= ge >> drop or loss of high frequences on the real pickup. >> >> Say I have a real pickup which measures 200 ohm resistance and 50 mH of >> inductance. At 2000 RPM I get a 10V peak to peak sinusoidal waveform >> without >> any load. >> >> So I get into the simulation software, create a 10Vpp sin oscillator, pu= t >> in >> serie a 200 ohm resistor and a 50 mH inductor, I still get that 10Vpp >> simulated >> signal. >> >> Now I add e.g. a resistor to the simulated circuit, and I measure a cert= ain >> voltage drop. >> If I add the same resistor, but this time for real, to the pickup, will = I >> get >> the same voltage drop? >> >> I just want to experiment with the design of a interface circuit without >> possibily doing it on the real thing, as it's in certain aspects >> impossible. >> >> With kind regards, >> Mario >> >> >> At 16:52 2017-03-25, Sean Breheny wrote: >> >Hi Mario, >> > >> >It depends on what you are using the output of the sensor for. If you j= ust >> >want to use it to count gear teeth then the voltage source alone would >> >probably be enough. If you care about the shape of the pulse it produce= s >> >then you would care about the inductance but you would probably care ab= out >> >other things, too, like the shape of the piece of metal it is sensing. >> > >> >Sean >> > >> > >> >On Sat, Mar 25, 2017 at 10:58 AM, Electron >> wrote: >> > >> >> >> >> Hello, >> >> >> >> is it OK to assume that I can model a variable reluctance sensor >> (pickup) >> >> in Spice and similar as an inductor with in serie a resistor and a >> voltage >> >> source? >> >> >> >> And is it OK to assume that the resistor value for my model must be t= he >> >> resistance I measure at the pins of the pickup, and the inductance va= lue >> >> I have to use for my model is the one I measured with an inductance >> meter? >> >> >> >> Voltage of course will be proportional to simulated frequency / >> rotational >> >> rate. >> >> >> >> Thank you. >> >> >> >> Cheers, >> >> Mario >> >> >> >> -- >> >> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> >> View/change your membership options at >> >> http://mailman.mit.edu/mailman/listinfo/piclist >> >> >> >-- >> >http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> >View/change your membership options at >> >http://mailman.mit.edu/mailman/listinfo/piclist >> >> -- >> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist >> >--=20 >http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >View/change your membership options at >http://mailman.mit.edu/mailman/listinfo/piclist --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .