Step up Transformer, simple error

[rwbohring]

No shame, I need troubleshooting assistance.

 

I bought some 17kv wire, an extension cord and a toroidal core with the intention of stepping up 120 to 6kv.

 

Turns: 100:2. (Fyi, Pic demonstrates 1 primary turn not 2)

 

I made two 50:1 voltage dividers (together 2500:1) to measure this voltage.

 

I should have measured about 3-6 volts after the divider.

 

And I did.

 

Then I disconnected one divider, I should get "overload" but instead I get an identical measurement as before.

 

So I disconnect all resistors and get the same voltage. ~4 volts.

 

So I get frustrated and touch both secondary leads together, no sparks. So I measure again, same low voltage. No divider.

 

I made a mistake with this transformer but can't figure it out.

So I rewound the primary to one turn, and moved it closer in proximity to the core.

 

Similar disappointing results.

 

Continuity of both windings checks out.

 

I figured a transformer would be easy.

What did I do wrong?

 

http://postimg.org/image/44pw6rfy9/

[EdEarl]

Well, you have discovered there is more to it than turns-1:turns-2. It sounds like you intend to attach 120vac to a one turn primary, which should be a short and trip your circuit breaker. You need enough reactance in your primary to limit the input current. BTW current steps down as voltage steps up. 6kvac at 50-60 Hz can be dangerous. Tesla coils operate at high frequency, which makes transformers with less metal more efficient, and makes the HV safer. You can buy a 6kv neon sign transformer for < $20. You could dissassemble one to discover how it is made. I'm afraid your attempt is extremely lossy, at best.

[rwbohring]

I need full current pass through.

Also, I'm not tripping a breaker since I built my fuse/rheostat console. I haven't even burned a fuse yet. 20amp.

 

But do I need more primary windings?

Here I have my phase 1 set up.

http://postimg.org/image/3x87azoht/

 

Current is being limited by the bulb and a dimmer. Dedicated fuse, and breaker.

 

 

How would I "reduce loss"?

[studiot]

 

Also, I'm not tripping a breaker since I built my fuse/rheostat console. I haven't even burned a fuse yet. 20amp.

 

 

 

I suspect your 'problem' lies here (underlined).

 

If you do not understand IMHO you should be heeding Ed's sensible safety advice.

Edited May 7, 2016 by studiot

[EdEarl]

I need full current pass through.

Also, I'm not tripping a breaker since I built my fuse/rheostat console. I haven't even burned a fuse yet. 20amp.

 

But do I need more primary windings?

Here I have my phase 1 set up.

http://postimg.org/image/3x87azoht/

 

Current is being limited by the bulb and a dimmer. Dedicated fuse, and breaker.

 

 

How would I "reduce loss"?

Forgive me, you don't have the right stuff.

 

I'm too ignorant to design the transformer you describe. In school about forty years ago, I might have done it. But I know enough to know you can't get there with your collection of components. You need to know the information in this Wikipedia article before you can begin to design and build the transformer you desire.

[rwbohring]

So because I've limited the current to levels that significantly reduce the magnetic field strength, my turns are no longer close enough together to be considered a single inductance, and each turn is actually its own inductor in series? Making it one big resistor?

 

I'll remove the bulb and the knob, add a switch and retry.

Thanks for the safety tips. =)

[EdEarl]

 

one big resistor?

Your insulation is so thick on the secondary your metal core isn't effective (efficiency terrible), the rheostat reduces voltage across the transformer to <<120v making output v <<2500, one turn primary is a short at 120v, .... You have many things that will prevent success.

 

Let's suppose you need 100 turns on the primary to avoid a short circuit; you need to know how much current you will pull to be able to buy the right size wire Is it 22 gauge or 6 gauge. The same is true for the secondary winding. Moreover, you can use thinner insulation, because winding-to-winding voltage is less than 6K, which allows you to put the wire closer to the iron core, key to having more double digit efficiency.

Edited May 7, 2016 by EdEarl

[studiot]

 

So because I've limited the current to levels that significantly reduce the magnetic field strength, my turns are no longer close enough together to be considered a single inductance, and each turn is actually its own inductor in series? Making it one big resistor?

 

I'll remove the bulb and the knob, add a switch and retry.

 

 

I didn't say that.

 

I said that you should know that the external impedences are much greater than the winding impedences. So Ohm's Law tells you that

most of the applied voltage will be dropped across the these, not the transformer primary.

 

For this to be the other way round your primary impedance has to be much greater than the external ones.

 

This is basic electrical circuit theory.

 

Yes I also thought that your transformer would be very inefficient because

 

 

1) For one or two primary turns most of the core material will be external to the primary field.

 

Edit - That was not very well put since the core ring concentrates the field and draws it into itself.

What I meant was that the volume of a primary of a few turns is very small compared to the ring core volume.

 

 

2) The heavy insulation of the windings increase spacing and reduce flux density.

Edited May 7, 2016 by studiot

[Enthalpy]

Such a core at 60Hz can't achieve 6kV reasonably. Imagine it has 15mm*15mm iron laminations saturating at 1,7T peak: the maximum voltage per turn is 0.1V rms. So even the 100 turns can't fit 120V, let alone 6kV. If connected to 120V without the current limiter, something will go bang, hopefully the fuse.

 

This means also that for <100W power, transformers for 6kV or 20kV use a higher frequency as far as possible. Ignition at gasoline engines uses a sudden current drop, so did CRT television sets. If not, it needs many thousand turns, costly with a winding machine, impossible by hand.

 

The parallel stray inductance doesn't drop at low current; it's independent of the current until a stronger current lets saturate the core. But this inductance is too small for the desired transformer. Imagine a relative permeability of 3000 at the 250mm long laminations: 1A*t induces then 15mT and 3,4µWb so the inductance per square turn is 3,4µH. Even 100 turns (let alone 2 turns) make only 34mH or j13ohm at 60Hz which is too little for proper operation at 120V - where in addition the core would saturate.

 

So:

• You can't achieve 6kV with such means.
• Even 120V would be seriously tedious. A ring core is asking for uncomfort.
• Please keep off the mains as you clearly don't have the necessary knowledge.

[rwbohring]

Lots of cool knowledge. I'm left unphased by your personal attacks.

 

I'll probably just get a utility pole transformer and backfeed it.

Thanks.

[studiot]

No one was getting at you personally, but folks who are disappointed when the laws of electricity or other physics don't confrm to their guesswork are going to be doubly disappointed.

[rwbohring]

If I use a NST, say 60mA, my load will require charging for over a minute. Not acceptable.

 

I imagine I might could have 2 or 3 transformers in series, each with a fraction of the total voltage amplification.

There's probably a catch there too. Enlighten me.

 

I made a crude prediction. I tested it failed. I post to ask questions and learn why. While the safety concerns are sound, saying "you're too X to do Y" is not something I heed.

[EdEarl]

Neon sign transformers come in many sizes, but 3.6A, 6kv? You want to charge something in a second that would take 60ma for 60 seconds, that's 3.6A for a second. That's serious electricity. The primary would pull the following:

3.6A*6000 v= xA/120v

xA=3.6A*6000v/120v

xA=180A

180A @ 120v is way more than you have in your house, which is probably limited to 30A @ 120. You might have an electrician install a 240V line big enough. However, your torus transformer is too small by at least a factor of 10, maybe 100. The transformer you need is closer to the ones you see on utility poles, two or three feet tall and 16" diameter.

 

I can think of nothing you can charge using AC at 6kv.

[rwbohring]

I had a lineman come out and explain what the utilities in the area were equipped with. Based on our conversation (there's a 40amp fuse 500 yards away, on the 7200 transmission lines. I can't draw more than about 120 amps on my side before I risk blowing the neighborhood switch, should they be running high draw as well. I have 2 neighbors on this circuit. I like them.

 

So I got a 13kw generator. No prob. I needed one any how.

 

But you're right. My charge rate ideally would be appx 0.1-0.2 hz.

 

Of course I could resize my components based upon a known input like a HST. That's an easy thought.

 

And it's probably my most realistic option it seems.

[John Cuthber]

Just at thought- not worthy of it's own topic- perhaps those with a better understanding could help me with this

If I wind a single turn "coil" of wire it will have an inductance.

If I remember rightly, the inductance is proportional to the area enclosed.

 

How big a loop would I need for my single turn primary to draw a viable (say 1 amp) current from the mains (call it 120 volts)?

Feel free to make any assumptions you like about the size of the wire and the material of the core (we may need to pretend that the copper wire is stupidly cold so it's electrical resistance is smaller than the usual value in order to keep the Q out of the range of "stupid").

[Enthalpy]

Low-voltage (understand 120V, 240V, 400V... up to 500V in this job) transformers have less and less turns as their core gets bigger with the power. At some point readily attained, the LV coil is indeed a single turn of copper sheet or plate (which doesn't mean "thin" in mechanical engineering) instead of wire, and to increase the power further, the transformer's shape departs from the optimum aspect to keep exceedingly sleek cores, but longer and spaced wider to give room to the windings.

 

At 120Vrms (in star connection) and 60Hz, it takes peak 0.45Wb. Using the laminations at 1.7T peak, a single turn suffices with 0.26m² core, or 0.51m*0.51m. That's a bit bigger than 1MVA, maybe 3MVA.

 

It is a good reason to have ~1MVA MV/LV transformers. It's also said to be an optimum between transformer cost and LV wiring cost and losses.

 

So even at usual 1MVA and 50Hz 240V, a MV/LV transformer has very few turns. As square wires would be impossible to bend, these transformers have sheet LV windings in a spiral. Excellent for the filling factor, but less for the eddy currents in the LV coil: flux leaking outside the iron core would cut all edges on its path around the currents hence flow through both windings near the core's edges, but the sheet LV winding prevents it (as it's thicker than 8.9mm skin depth at 50Hz) by flowing extra "eddy" currents that mean losses.

 

An answer to this is to split the HV winding, half within the LV and half around, so that the A*t observed from the core outwards look like +- -+ instead of +-. I haven't seen interleaving exaggerated beyond splitting the HV, but this is done in switched mode power supplies where the LV winding is of sheet for the same reasons, though at lower power and higher frequency.

 

No complete solution exists, because the sheet winding is much wider than the skin depth, and also thicker so it really prevents the variable flux pass through, by running currents as needed and especially near the sheet's edges. Though, as the A*t of the primary and secondary nearly cancel out, the losses are reduced by circulating the HV A*t as uniformly as possible, which most A*t in the sheets mimick. Ideally, only the difference of primary and secondary A*t, the magnetizing current, still induces eddy current in the sheets.

[rwbohring]

Would using a flat wire work to increase the area? Is that what you mean by 'sheets'?

[EdEarl]

 

But you're right. My charge rate ideally would be appx 0.1-0.2 hz.

Do this mean charging every three to six seconds?

[rwbohring]

Do this mean charging every three to six seconds?

6-12 seconds.

 

At 60mA, I would be charging at a rate much closer to 2 minutes given natural waste I'll incur. Requiring me to scale down my load to meet expectations. I'd prefer not to think in that direction yet.

 

I've already 'started small' and proved concept on bread board. But as it turns out, the transformer doesn't scale up nearly as easily as the other components. Very disappointing. =(