Toroid eddy current losses
Tesladownunder, Sun Apr 15 2007, 11:08PMHas anyone ever quantitated this? I still believe it must affect the Q at least and cause some losses. People (? Bill Wysock) have sectioned the toroid before with no discernible effect, but that wont really stop eddy currents to my mind.
Eddy currents are not always intuitive. Take can crushing for example of eddy current effect. You would expect a longitudinal slit to prevent crushing because the "ring" is now interrupted. Not at all. Crushing is almost unaffected. I discuss this in detail on my site with various examples.
Photos below show a slit can crushed readily.
It is possible to make a toroid structure that prevents all eddy currents. I will put this as number 1002 on my "to do" list.
TDU
Re: Toroid eddy current losses
Marko, Sun Apr 15 2007, 11:16PM
I was also somewhat worried about those thing, but effect is absolutely minor.
Conductive stuff, copper, aluminium and etc. will limit the eddy current primaryly by inductive reactance and ohmic loss will be completely neglible.
Marko, Sun Apr 15 2007, 11:16PM
I was also somewhat worried about those thing, but effect is absolutely minor.
Conductive stuff, copper, aluminium and etc. will limit the eddy current primaryly by inductive reactance and ohmic loss will be completely neglible.
Re: Toroid eddy current losses
Steve Ward, Mon Apr 16 2007, 01:52AM
Sure, just make the toroid out of rings linked up in a loop. But, this main loop would still have eddy currents induced, so you would have to move it away infinitely to have no eddy currents.
Ive had toroids get "hot" but the power it would take to even do that is small compared to the actual power of the tesla coil. Seems you could spend your time doing better things.
BTW, your crushing of the "slit can" makes perfect sense when you consider the fringing effects of the B field at the ends of the coil. I would expect the effect to be lessened if the can's length was < coils length.
Steve Ward, Mon Apr 16 2007, 01:52AM
It is possible to make a toroid structure that prevents all eddy currents.
Sure, just make the toroid out of rings linked up in a loop. But, this main loop would still have eddy currents induced, so you would have to move it away infinitely to have no eddy currents.
Ive had toroids get "hot" but the power it would take to even do that is small compared to the actual power of the tesla coil. Seems you could spend your time doing better things.
BTW, your crushing of the "slit can" makes perfect sense when you consider the fringing effects of the B field at the ends of the coil. I would expect the effect to be lessened if the can's length was < coils length.
Re: Toroid eddy current losses
Sulaiman, Mon Apr 16 2007, 05:30AM
Here's my version of a 'low-loss' toroid
The two ends of the drier-ducting are insulated from each other
and the ducting is tied to a bicycle wheel
the spokes conduct radially but reduce eddy-currents.
The main reason I did this was because a local bicycle shop gave me the wheel for free.
Sorry for the poor photo', it's 06:30 am here.
Sulaiman, Mon Apr 16 2007, 05:30AM
Here's my version of a 'low-loss' toroid
The two ends of the drier-ducting are insulated from each other
and the ducting is tied to a bicycle wheel
the spokes conduct radially but reduce eddy-currents.
The main reason I did this was because a local bicycle shop gave me the wheel for free.
Sorry for the poor photo', it's 06:30 am here.
Re: Toroid eddy current losses
Terry Fritz, Mon Apr 16 2007, 09:09PM
Hi,
MandK, the old DOS coupling program, can be used to figure out the coupling to the toroid from the coil and even the primary if you fudge the input number a bit:
Just assume the toroid is a single turn fat primary coil. I did that a very long time ago and the coupling was so low I never worried about it again.
Of course, you could make one of these toploads and not worry about it
Cheers,
Terry
Terry Fritz, Mon Apr 16 2007, 09:09PM
Hi,
MandK, the old DOS coupling program, can be used to figure out the coupling to the toroid from the coil and even the primary if you fudge the input number a bit:
Just assume the toroid is a single turn fat primary coil. I did that a very long time ago and the coupling was so low I never worried about it again.
Of course, you could make one of these toploads and not worry about it
Cheers,
Terry
Re: Toroid eddy current losses
Tesladownunder, Tue Apr 17 2007, 05:49AM
Your topload will reduce eddy current but they are large metal surfaces so must have some effect, compared to an eddy current free wire frame toroid.
It seems like experimentally it will be hard to show the difference except with a stable and reproducible DRSSTC coil.
TDU
Tesladownunder, Tue Apr 17 2007, 05:49AM
Firkragg wrote ...
Conductive stuff, copper, aluminium and etc. will limit the eddy current primaryly by inductive reactance and ohmic loss will be completely neglible.
Then why does a can crush?Conductive stuff, copper, aluminium and etc. will limit the eddy current primaryly by inductive reactance and ohmic loss will be completely neglible.
Steve Ward wrote ...
The rings have to come from a point not a loop to have no eddy currents. Easy to do but no-one has done it that I know.It is possible to make a toroid structure that prevents all eddy currents.Sure, just make the toroid out of rings linked up in a loop. But, this main loop would still have eddy currents induced, so you would have to move it away infinitely to have no eddy currents.
Steve Ward wrote ...
BTW, your crushing of the "slit can" makes perfect sense when you consider the fringing effects of the B field at the ends of the coil. I would expect the effect to be lessened if the can's length was < coils length.
Quite right and a fine slice of the can slit does not crush at all of course and I have pics of that and other variations on my site.BTW, your crushing of the "slit can" makes perfect sense when you consider the fringing effects of the B field at the ends of the coil. I would expect the effect to be lessened if the can's length was < coils length.
Sulaiman wrote ...
Here's my version of a 'low-loss' toroid......
Lower loss but still the metal surface of the toroid can have eddy currents in it or any one of a number of combinations of loops even if the toroid is interrupted in one place.Here's my version of a 'low-loss' toroid......
Terry Fritz wrote ...
...Just assume the toroid is a single turn fat primary coil. I did that a very long time ago and the coupling was so low I never worried about it again.
.. Of course, you could make one of these toploads and not worry about it
I understand that the modelling shows low coupling and hence of little significance. This should also apply to short circuiting the last 10 turns of your secondary which just seems wrong. I cant get my head around that. ...Just assume the toroid is a single turn fat primary coil. I did that a very long time ago and the coupling was so low I never worried about it again.
.. Of course, you could make one of these toploads and not worry about it
Your topload will reduce eddy current but they are large metal surfaces so must have some effect, compared to an eddy current free wire frame toroid.
It seems like experimentally it will be hard to show the difference except with a stable and reproducible DRSSTC coil.
TDU
Re: Toroid eddy current losses
Marko, Tue Apr 17 2007, 08:17AM
I was as well paranoid about those things but found it to be unjustified.
Crushing is effect dissimilar from ohmic heating; repulsive force will always occur although your current may still be limited mostly by inductance.
Considering your tank energy and coupling your coil is unlikely to crush it's toroid.
Actually, if you make your toroid out of closed wire mesh, losses may actually increase because you increased resistance in path of current.
People put big copper plates under their SSTC primaries and it only boosts performance because it acts as low-impedance ground, and doesn't seem to get warmer than room at all because it's resistance is very low. (Even in MHZ range!).
I personally doubt that there will be measurable difference on coils with and without a toroid since other losses like secondary and primary copper and spark gap loss are bigger by many orders of magnitude.
Only thing I would like to avoid around coils is excessive use of iron and stell, because it does and will get hot due to remanence loss. Remanence doesn't care about loop area, and represents much better 'impedance match' to the coil so it will dissipate large amounts of power if you allow it to.
Marko, Tue Apr 17 2007, 08:17AM
Then why does a can crush?
I was as well paranoid about those things but found it to be unjustified.
Crushing is effect dissimilar from ohmic heating; repulsive force will always occur although your current may still be limited mostly by inductance.
Considering your tank energy and coupling your coil is unlikely to crush it's toroid.
Actually, if you make your toroid out of closed wire mesh, losses may actually increase because you increased resistance in path of current.
People put big copper plates under their SSTC primaries and it only boosts performance because it acts as low-impedance ground, and doesn't seem to get warmer than room at all because it's resistance is very low. (Even in MHZ range!).
I personally doubt that there will be measurable difference on coils with and without a toroid since other losses like secondary and primary copper and spark gap loss are bigger by many orders of magnitude.
Only thing I would like to avoid around coils is excessive use of iron and stell, because it does and will get hot due to remanence loss. Remanence doesn't care about loop area, and represents much better 'impedance match' to the coil so it will dissipate large amounts of power if you allow it to.
Re: Toroid eddy current losses
Dr. Slack, Wed Apr 18 2007, 12:46PM
What happens if the topload is made out of wire, close-wound in a spiral on a former, essentially as a continuation of the secondary. A substantially heavier gauge than the secondary could be used to minimise winding effort, perhaps some insulated hook-up or mains wire. A rough 35 seconds'worth on paint shows the general idea. The former is not shown. The picture is actually made using discrete ellipses, but it doesn't take much imagination to believe them to be a continuous spiral with changing radius.
Choosing the shape of the former could give a smooth flare from the secondary, essentially acting as the corona control that you get with lower, smaller toroids below the main one. The use of a wound former, and the freedom to choose its shape, add another degree of freedom to the job of building a different-looking coil, or a good-looking coil if you can't source a decent toroid.
As far as I can see, the corona suppression of the whole should be fairly similar to a not-perfectly-smooth toroid, as adjacent wires shield each other electostatically.
The two main differences compared to a solid toroid would obviously come from the fact that it is actually a longish piece of wire. This will affect a) the effective capacitance and b) the discharge dynamics. A) The effective capacitance, in its effect on the resonant frequency, ought to be negligable, as long as an effective length is used for the secondary which includes some of the turns in the extended topload. B) The discharge dynamics may change as the inductance behind the arc has increased. This should be negligable for streamer formation, but an arc along the subsequent channel might be expected to be weaker.
thoughts?
Dr. Slack, Wed Apr 18 2007, 12:46PM
What happens if the topload is made out of wire, close-wound in a spiral on a former, essentially as a continuation of the secondary. A substantially heavier gauge than the secondary could be used to minimise winding effort, perhaps some insulated hook-up or mains wire. A rough 35 seconds'worth on paint shows the general idea. The former is not shown. The picture is actually made using discrete ellipses, but it doesn't take much imagination to believe them to be a continuous spiral with changing radius.
Choosing the shape of the former could give a smooth flare from the secondary, essentially acting as the corona control that you get with lower, smaller toroids below the main one. The use of a wound former, and the freedom to choose its shape, add another degree of freedom to the job of building a different-looking coil, or a good-looking coil if you can't source a decent toroid.
As far as I can see, the corona suppression of the whole should be fairly similar to a not-perfectly-smooth toroid, as adjacent wires shield each other electostatically.
The two main differences compared to a solid toroid would obviously come from the fact that it is actually a longish piece of wire. This will affect a) the effective capacitance and b) the discharge dynamics. A) The effective capacitance, in its effect on the resonant frequency, ought to be negligable, as long as an effective length is used for the secondary which includes some of the turns in the extended topload. B) The discharge dynamics may change as the inductance behind the arc has increased. This should be negligable for streamer formation, but an arc along the subsequent channel might be expected to be weaker.
thoughts?
Print this page