Ferrite Pole Pig / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Ferrite Pole Pig

previous 1 2 3 next

Move Thread

LAN_403

Patrick

Fri Jul 01 2011, 03:49AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Klugesmith wrote ...

Saturation is a voltage thing, and has little to do with load current.

Suppose that with no load, you measure the primary current and infer
the B-H operating point and/or curve.
Now when you connect a load and current flows in the secondary,
there is a corresponding increase in primary current, such that their
magnetizing effects cancel. So the magnetic flux swing does not change.

More precisely, it actually goes down to the extent that IR losses
use up some of the applied primary voltage. That leaves less to be
bucked by the induced voltage in primary.

Your right klugesmith, i failed to make clear that the 8040 would have higher H than the ALL50-- Im a little tipsy right now, so ill chack my math when sober. Also, how can he and i calculate the max power output of a given ferrite and related geometry?

dude_500

Mon Jul 04 2011, 06:22AM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

I've been playing with secondaries, and having all sorts of corona issues. Basically, only way to make it work with a traditional simple TL494 driver would be to oil pot the thing, which is not what I want to do, too messy! With a fairly modest 10:125 step up, I was getting arcs that form at a 3/4" gap with only 75VDC on the bus. This is way too much ring-up to live with, that would be several hundred kV ringing at 300VDC bus.

So, I put together a new drive circuit that I will etch and test hopefully tomorrow (if I'm not too busy playing with plasma on the 4th). The idea is that when there is no current detected flowing through the grounded end of the secondary coil (eg there is no arc being drawn), the driver will only pulse the transformer, at what I'm thinking will be a good rate of hitting one of every 10 cycles from the TL494 (of course this will be adjustable). The moment any current flow from that strike pulse is detected from the current transformer on the ground return, the driver keeps going, and within a cycle or two of the current going away (eg the arc disappears), the driver goes back to only pulsing one of every 10 cycles awaiting the next arc.

My one concern is that perhaps the disappearance of the arc isn't so much an event as it is a gradient going from arc to no arc, traveling for a significant time between low voltage and high voltage as the plasma disappears, during which point a new arc will have time to form at the ferrite core before the current hits sufficient zero to shut off the driver back to pulse mode. Anyone with more knowledge in arc-physics know how this will affect the concept?

Schematic attached.

Dr. Dark Current

Mon Jul 04 2011, 06:58AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hi,
I was also thinking about the ringing problem.
These are the possible solutions I came up with:

-Wind a "snubber" winding under the secondary and rectify it back to the DC bus, the idea here is that this winding has much less leakage to the secondary than the primary and should cut the spikes much better

-Use a LC filter before the primary (C parallel to primary, L in series). To avoid resonant rise on the primary, use two diodes to draw excessive energy from the LC circuit back to the dc bus (easy with a half-bridge). There will be a kind of distorted sine wave on the primary, but much better than spikey square.

-Use a pure sine wave converter, like a current fed parallel resonant inverter (which was simplified and popularised as the "zvs" driver) or a class-C oscillator with vacuum tubes.

-Jan

Edit: Maybe you hit the transformer's resonant frequency or some of its sub-harmonics, try to tune the frequency away a little.

dude_500

Wed Jul 06 2011, 07:20AM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

The board has been assembled and tested at around 1kW, the concept seems to work great, assuming you can put up with the awful audio-range "strike voltage" generator, which is unfortunately pretty fundamental to the operation. I'm hoping the noise gets better when I tighten the ferrite down real hard.

Here is a video of it running (Note the current on the scope, didn't even realize that was up during recording, but it is neat to see what it is doing):

I'm not sure quite why it is having the strike hum oscillating at about 10hz, it did that every now and then and at the end of the day for filming, I didn't feel like tracking down the cause, probably just picking up noise due to the sketchy test setup.

Going to make a new secondary tomorrow and firm up some of the wiring so it's more reliable and noise immune, and see how that works, maybe at a couple kW.

I have it detuned from resonance by nearly 10khz, since right at resonance it draws >50A at about 35VDC bus, and puts out huge vicious arcs that melt my 1/4" brass electrodes in under a second, and subsequently trip my variac breaker.

Attached pictures of the controller, and the current detector torroid, which is about 10 turns with a burden of 10 ohms. I think I may try a hall-effect current sensor to see if that can pin down some of the slight noise issues I'm having at higher running powers.

teravolt

Wed Jul 06 2011, 04:48PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

hi dude 500 well done are you using resonant caps with the primary? what is the resonant frequency of the secondary?

dude_500

Wed Jul 06 2011, 05:55PM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

Yes, it is a capacitor series load resonant which is resonant at 40khz (running around 32-33khz).

jpsmith123

Sat Jul 30 2011, 11:50PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

In that first picture I see what looks like a piece of PVC pipe enclosing the 4 legs of the 4 core assemblies. I'm wondering, did you wind a single layer coil on that? If so, can I ask, what size wire and how many turns did you use?

Thomas W

Sun Jul 31 2011, 11:03AM

Registered Member #3324 Joined: Sun Oct 17 2010, 06:57PM
Location:
Posts: 1276

666 views this transformer is cirtainly a beast :D

jpsmith123

Mon Aug 01 2011, 01:27PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Ok, from the video you made, I think I see a single layer secondary wound on the PVC pipe.

So if each primary has ten turns, and they are all in parallel, and the secondary has 125 turns, then your step-up ratio will be 4 * (125/10) = 50/1.

dude_500 wrote ...

I've been playing with secondaries, and having all sorts of corona issues. Basically, only way to make it work with a traditional simple TL494 driver would be to oil pot the thing, which is not what I want to do, too messy! With a fairly modest 10:125 step up, I was getting arcs that form at a 3/4" gap with only 75VDC on the bus. This is way too much ring-up to live with, that would be several hundred kV ringing at 300VDC bus.

So, I put together a new drive circuit that I will etch and test hopefully tomorrow (if I'm not too busy playing with plasma on the 4th). The idea is that when there is no current detected flowing through the grounded end of the secondary coil (eg there is no arc being drawn), the driver will only pulse the transformer, at what I'm thinking will be a good rate of hitting one of every 10 cycles from the TL494 (of course this will be adjustable). The moment any current flow from that strike pulse is detected from the current transformer on the ground return, the driver keeps going, and within a cycle or two of the current going away (eg the arc disappears), the driver goes back to only pulsing one of every 10 cycles awaiting the next arc.

My one concern is that perhaps the disappearance of the arc isn't so much an event as it is a gradient going from arc to no arc, traveling for a significant time between low voltage and high voltage as the plasma disappears, during which point a new arc will have time to form at the ferrite core before the current hits sufficient zero to shut off the driver back to pulse mode. Anyone with more knowledge in arc-physics know how this will affect the concept?

Schematic attached.

dude_500

Mon Aug 01 2011, 04:54PM

Registered Member #2288 Joined: Wed Aug 12 2009, 10:42PM
Location: Cambridge, MA
Posts: 179

jpsmith123 wrote ...

Ok, from the video you made, I think I see a single layer secondary wound on the PVC pipe.

So if each primary has ten turns, and they are all in parallel, and the secondary has 125 turns, then your step-up ratio will be 4 * (125/10) = 50/1.

I think you're right that the effective winding ratio is times 4, however I don't think it really matters so much for a series resonant driven ferrite transformer. I don't think it's a winding relationship relating the primary to secondary voltages at all (although yes, twice as many secondary windings is twice the voltage on it, but 1:1 isn't necessarily the same output voltage as input). Reason being, the series resonant capacitor allows the impedence of the primary to drop well below its actual inductive impedence, so an abnormal current can flow through the primary as compared to if it were being driven by a plain sine wave. And since induced EMF is -dFlux/dT, and Flux is the surface integral of B dA, and B is derived from the primary current, the secondary voltage can quickly get out of control with very small voltages on the primary in a series resonant circuit. This is why I made the new controller design, to try to reduce this problem.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob