AC electromagnet - eddy current heat, lamination orientation?

[peskywinnets]

So I made a guitar sustainer (basically an AC electromagnet, with a permanent magnet on the bottom top pull the strings down in the absence of any AC signal going through the coil) - it works ok but the driver coil/core gets extremely hot ...especially at high frequency (frequency range is up to 5khz). I'm presently using a mild steel core (which isn't optimum, but easy to source!), but really should be using electrical steel laminations for the coil core.

For the least eddy current induced heat, which way should the laminations be orientated, I've done this (bad) drawing to help....

Bearing in mind where most of the magnetic direction force/flow is, which option would be best for the lamination orientation? Option 1, option 2 or option 3?

Also all related tips welcome! (type of laminations, supply source (I'll be ripping upa transformer for my immediate source!)

 

[MagBen]

Use option 2 to minimize the path lenght of eddy current, so less eddy curent losses.

Decrease lam thickness. in some range of thickness and applied current, losses in inversely proportional to the square of thickness.

yes, FeSi should be better than mild steel, if no choice, make sure mild steel is annealed.

 

[peskywinnets]

Thanks...option 2 is going to be a killer to put together...the dimensions of the inner core is just 4mm high x 2mm wide x 54mm long.....so I'd be looking at about 100) strips of 0.5mm thick electrical steel in a tiny 4mm x 2mm (it ain't gonna happen!). I was hoping that option 3 would be viable...as at least them I'm dealing with workable strips (4mm high x 54mm long). Hey ho!

What does annealing do to the mild steel to help out here?

 

[dgallup]

Option 3 will still help and be a LOT easier to make. Just the change from mild steel to silicon iron is a big improvement due to higher electrical resistivity. You might want to consider 430FR stainless too, it has even higher resistivity than silicon iron and won't rust. The saturation flux density is lower but I doubt that is a problem in this device.

I don't think annealing the mild steel will help the heat at all, annealing has no effect on resistivity. Annealing will affect the BH curve and the coercive force.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[peskywinnets]

Re 430fr stainless....every bit of stainless I've tried wasn't magnetic?

The core of the coil needs to be magnetic to (for want of a better word) 'conduct' the permanent magnet's flux attached to the base of the core up through the coil & out the top which results in a quiescent magnetic pull/attraction on the ferrous guitar strings....which is either reinforced or part negated by the AC signal through the coil).... if I sourced some 430fr (btw where on earth do you go to get such stuff?!!), & placed a magnet on the bottom....would the 430fr even attract & allow the flux through to the top of the core?

thanks.

 

[MagBen]

The annealing could help a lot if your mild steel was cold rolled. It is true the resistivity will not change a whole lot, but decrease in area of BH loop will decrease hystreresis losses very effectively. Under a relatively low frequency of 5khz, the loss component from hysteresis losses can be still high.

Option 3 is viable, but never option 1 in which the lam structure will not help much.

4xx SS is magnetic, but is the corresion resistance a big issue in a Guitar applciaiton? Also, SS will have a higher coercivity with increasing hysteresis losses.

 

[peskywinnets]

Corrosion on a guitar could be a bit of an issue (sweat dripping onto the sustainer from the guitarist's hand, beer spillages, humid environments etc)....and a rusty looking core won't look nice, so actually 430 FR quite appeals! (I'd actually been fretting how to make silicon steel a bit more rust proof, but if they make a stainless that's magnetic, then that's got to be worthy of onvestigation. Re the coercivity & hysteresis losses ....any worse than mild steel? (you wouldn't believe just how hot the magnetic core is getting at the moment!)

 

[dgallup]

All your ferritic stainless steels are strongly magnetic and the martensitic are too but you don't want them for this device. Only the austenitic stainless steels are non-magnetic. Even they can become magnetic with enough cold work.

430FR is used in lots of solenoid and relay applications. Fuel injectors and the like. Carpenter and Schmolz-Bickenbach are two large mills. I don't know where to get small samples.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[MagBen]

I doubt Carpenter can offer strip form product. 430FR is mainly for DC or low frequency applicaitons, so the main forms are bar, wire and billet. this alloy was also desgned to have free machining capability.

Can you decrease current, or turns, such that both eddy current and hysteresis losses decrease? are you using AlNiCo as the permanent magnet? could using a strong magnet, say NdFeB, decrease your needed AC signal? otherwise, try high perm 49 (Fe-50Ni) or Hymu 80Ni which will give you a much more sensitive core (much less current) while a better corrosion than FeSi and carbon steel. strip form for High perm 49 and Hymu 80Ni are always readily available.

 

[peskywinnets]

the problem is, the thinnest strings on a guitar need quite a bit of current through the coil (& quite a few turns too!)...in fact I've been on a fair while getting as much magnetic drive into the core as possible (to assist in getting the thinnest strings to sustain well), so any thought of now backing off the magnetic intensity to reduce eddy current heat isn't that paleteable!

 

[waross]

We have been assuming that all the the coil is well designed and that all the heat is generated in the steel. If the wire is undersized a lot of heat may be coming from the winding.
As to the shape, a "U" shaped core will have a shorter air path and require less current. I don't know how that will affect the sustaining action.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[peskywinnets]

Hi Bill, You raise a good point about the coil being the potential source of the high heat, my plan of attack...

Substitute the present steel core for a temporary lash up made with electrical steel laminations (it'll look ugly, but it';s just for a test)...carry out the same sustain tests at high frequency - if still very hot, then yes, maybe it's the wire size (fwiw the IC driving the coil gets war (bordering on hot) to touch, so there is a bit of current flowing therough the output IC/coil, but the coil is much larger than the IC (DIL08 package) & I'd expect it to run cooler than the IC.....and like I say, the IC isn't outrageously hot...so I'm figuring that the coil is not the problem. Also when I sustain notes at lower frequency (eg 80-200Hz), the core doesn't get anywhere near as hot vs when I've sending a few kilohertz into the coil.

re the U shaped coil...not sure how that would be approached on a coil breaching six stings over 60mm length? (trying to visualize how that would be implemented)

 

[MagBen]

You would also want to check if the heat source was from the permanent magnet.

Pick up coil is a "receiver", its signal is determined by the "change" of magnetic flux (the current is created by against flux change). so you donot apply any current, right? The sustaining function is related to field strength of permanent and core, has nothing to do with the winding.

To sustain well, you can use a stronger permanent magnet, decrease air gap (if any) between the permanent magnet and core.

"U" shaped core does decrease air path, so higher effective permeability, problem is how to design the coil to "see" the change of flux?

 

[peskywinnets]

The permanent magnet on the bottom of the core acts akin to a 'bias' ...it attracts/pulls the strings down in the absence of an AC signal through the coil. When an AC current runs through the coil, then the core's magnetic 'pull' is both reinforced & (to an extent) negated ...this is what causes the the string to vibrate in both directions (up/down)...remember you can't push a ferrous string, so the trick is to pull it down with the permanent magnet, then 'releas'e the string from that pull by overiding it with an AC signal....a faked 'push' of the string (which is not a push at all...it's a release). The sustaining function is very much related to current through the windings ...because it's only if you have enough current & windings, that there's an impact on the core (which is magnetized by the permanent magnet)

 

[MagBen]

Isnot the guitarist who makes the string vibration, and the pickup coils "sense" the string movement? So your desgin is to use mainly the magnetic polarity switch of the core (string up to decrease flux, down to increase flux) to control the vibration? That way, if your ampere turns is too big such that the field of core override the field from permanent magnet. The vibration will be "dampen" very qucikly to stop. While, if the NI is small, you just need to amplify more of your signal, and your core will have less control of vibration, and the string vibrates more naturally.

The winding can be designed directly on the permamanet magnet without using any soft magnetic core. Since the permanent magnet is normally not laminated, should eddy current effect even worse?

 

[peskywinnets]

No, it doesn't wok as you say.

Sure, a guitar pickup senses the string...that signal is then buffered/amplified and fed to the driver coil. The AC signal (i.e. as sourced from the pickup) modulates the permanent magnet field acting on the string...which causes the string to move in sympathy with the original string vibrations.

 

[IRstuff]

This is what the OP is referring to:

http://www.fernandesguitars.com/sustainer/sustainer.html

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[MagBen]

ok, thanks for clarification.
so it is the drive coil that gets hot, NOT the pick up coil!

Any idea what is NI and diameter of the coil? if NI is big, and the magnetic circuit is fully open, you are actually not taking advantage of the soft magnetic properties of the core. Either mild steel, si steel, SS steel, or permealloy would not not make much difference. The eddy current component will dorminate the losses. Decreasing the lam thickness and increasing resistivity would be the directions to choose core materials.

If you design is a close circuit with zero air path, hysteresis losses must be taken into account since it is proportional to frequency.

 

[IRstuff]

I'm guessing that commercial sustainers do not have a constant magnetic bias, and probably, neither should the OP. Since the string is vibrating at its resonant frequency, there is actually no need to have bi-directional stimulus. Just like plucking is uni-directional, from a equilibirum perspective, you can simply pulse the magnets to "pluck" pluck them at the resonant frequency. You can shape the pulses to minimize the harmonics.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[peskywinnets]

As you say you are guessing!

A little bit more info (because it seems there's many trying to tell me how a sustainer works!)...I've actually got the sustainer working well (it has been a long journey & belive me, I've tried every permutation known to man!), the last 'challenge' is the magnetic core material. I was happy enough with a mild steel core until I used the sustainer heavily in harmonic mode...this particular mode sees much higher frequency going into the driver coil & it's then that the driver becomes too hot to touch, hence me posting abouut lamination orientation (I still believ that this is eddy current related as at lower frequency the core isn't anywhere near as hot)

In short, the sustainer *does* need a permanent magnet (the goal is to move the string both 'sides' of its quiescent rest (up/down)...but since you can't push a string with magnets, the only alternative is to 'release' a permanent magnet's hold on the string)