DIY high voltage & high power transformer / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

DIY high voltage & high power transformer

1 2 3 4 next

Move Thread

LAN_403

Kizmo

Fri Jul 20 2007, 08:41PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Things have changed, look latest posts

Sulaiman

Fri Jul 20 2007, 08:48PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

If at all possible you should not introduce an air gap into the magnetic circuit
it will (dramatically) increase the current that the transformer draws with no load connected
and it will give poorer regulation (change in output voltage vs output current.

Can you take a photo' from the side ?
i.e. turn the transformer by 90 degrees.

Unless the laminations are welded together it may be possible to extract them.

Kizmo

Fri Jul 20 2007, 08:57PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Laminations are just glued together. This transformer has quite thin laminations, about 0.3 or 0.4mm, that makes about 230 layers of metal. Friend told me that heating the core to 250-300 degrees C would make glue so weak that it would be possible to disassemble whole thing into hundreds of U and I shaped metal strips. Im just wondering how on earth that can be assembled backwards :D

EDIT:

Air gap could be minimized by carefully sanding (and then lapping with water sandpaper) both pieces so that rough bandsaw finish disappears?

Finn Hammer

Fri Jul 20 2007, 09:30PM

Registered Member #205 Joined: Sat Feb 18 2006, 11:59AM
Location: SkÃ¸rping, Denmark
Posts: 741

Kizmo wrote ...

Laminations are just glued together. This transformer has quite thin laminations, about 0.3 or 0.4mm, that makes about 230 layers of metal. Friend told me that heating the core to 250-300 degrees C would make glue so weak that it would be possible to disassemble whole thing into hundreds of U and I shaped metal strips. Im just wondering how on earth that can be assembled backwards :D

The glue is lacquer, really. It doesn`t penetrate far in between the laminations. It is mostly on the outside. And it cracks easily-
Use a wood chisel to separate the lams, and pull them out with a heavy duty plier. The first 5 or ten will get bent badly. after that, when there is place, with a little care, you can easily get them out. Takes a lot of cleaning, though.

Kizmo wrote ...

EDIT:

Air gap could be minimized by carefully sanding (and then lapping with water sandpaper) both pieces so that rough bandsaw finish disappears?

Cut like you showed, and you will have to wind the coil directly on the core. A machine shop can surface grind the mating faces, but you cannot do it with the method you described. Being a toolmaker myself, I can assure you, that it takes a skilled metalworker to get the level of accuracy required, to reduce leakage flux to a reasonable level.
Go for it, nothing beats the feeling of winding your own!

Cheers, Finn Hammer

Kizmo

Fri Jul 20 2007, 10:20PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Its true that nothing beats DIY HV things :)

I just ordered some copper wire for secondary (0.20mm, 2.6km) and some polyester based transformer insulation tape to insulate secondary winding layers (Dielectric strength: 7500 V) Now i just need to start building some sort of rev counter to my lathe :)

Sulaiman

Sat Jul 21 2007, 02:46AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I really would go with Finn's method,
I've dis-assembled a few laminated steel transformers
the loss of one or two laminations in the extraction process
is far better than introducing an air gap.

Use a solvent to clean the laminations prior to re-assembly

Once re-assembled tap (HIT) the laminations together at the top and bottom
to make the E and I sections of each layer have as small an air gap as possible
to 'discourage' the magnetic flux from 'jumping' sideways from lamination to lamination
this will minimise losses due to eddy currents.
Then apply thin varnish and allow it to seep between laminations to hold them in place
and reduce hum/buzz.

Winding with 0.2mm wire and insulating with tape each layer will be quite tedious
do some research on winding techniques first
and make sure that you do not wind to the ends of the insulation or overall insulation will be difficult.

For the voltage that you are aiming for I would;
- varnish each layer before applying tape
or preferably
- use craft/brown paper instead of tape and immerse the whole transformer in oil.

Just because your transformer insulation seems ok for a few minutes (or days)
doesn't mean it will last for months or years.
If not under oil, the corona at 20 kV + will slowly 'eat' the insulation.

Hazmatt_(The Underdog)

Sat Jul 21 2007, 03:57AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

You'll want to insulate for coil to coil breakdown too by placing several layers of an insulating material between the two, Cambric paper or Fish paper are good for this. Then you want to also insulate the HV coil from the core with pieces of the same paper on top and on bottom of the coil. This will reduce the risk of corona from the ends of the coil. I would go a step further and use some kind of silicone insulating material between the paper end cap and the coil as well.

Finn Hammer

Sat Jul 21 2007, 10:40AM

Registered Member #205 Joined: Sat Feb 18 2006, 11:59AM
Location: SkÃ¸rping, Denmark
Posts: 741

Insulation btwn. layers is one thing, and certainly a factor. with around 50 layers in a 20kV coil, there are 400 volts between layers.

Another, more urgent need, is acheived by the interleaving material: maintaining a winding without overlapping turns. Everytime you make an overlapping turn, you waste radial space for a whole layer!
There is limited space in a transformer, so each layer has to be filled with the maximum possible turns.

You _must_ put it under oil. You will realize this the moment you turn it on. there will be blue corona, buzzing and hissing like a beeswarm.

Hazmatt_(The Underdog) wrote ...

Then you want to also insulate the HV coil from the core with pieces of the same paper on top and on bottom of the coil. This will reduce the risk of corona from the ends of the coil.

Hazmat:
If you have corona in an airgap, you gain nothing by adding solid dielectric. The field strength in the remaining (now shorter) airgap will increase.
The voltage drop across individual layes, in a sandwich of dielectric materials, is determined by the dielectric constant of the individual material.
Higher dielectric constant means lower drop.
Let`s say we have a 10mm air gap between 100mm balls. Breakdown voltage is 29kV in air.
So with 25kV across this gap there is no corona.
Now we introduce a 9mm thich slab of dielectric with a dielectric constant of 2,2.
The voltage drop across the solid dielectric becomes ~20kV leaving 5kV across the remaining millimeter of air.
50kV/cm in air is way into corona territory.
You cannot insulate your way out of a corona problem, unless you exclude the air completely. Oil is the easiest way, solid potting cured under heavy vacuum is another method

Cheers, Finn Hammer

Dr. Dark Current

Sat Jul 21 2007, 11:10AM

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

Finn Hammer wrote ...

Insulation btwn. layers is one thing, and certainly a factor. with around 50 layers in a 20kV coil, there are 400 volts between layers.

Actually, for 20kv ac and 50 layers, there are 800 volts ac (or 1130 vpeak) between the two "ends" of 2 layers.

Finn Hammer

Sat Jul 21 2007, 01:50PM

Registered Member #205 Joined: Sat Feb 18 2006, 11:59AM
Location: SkÃ¸rping, Denmark
Posts: 741

jmartis wrote ...

Finn Hammer wrote ...

Insulation btwn. layers is one thing, and certainly a factor. with around 50 layers in a 20kV coil, there are 400 volts between layers.

Actually, for 20kv ac and 50 layers, there are 800 volts ac (or 1130 vpeak) between the two "ends" of 2 layers.

Whoops, my bad.
Forgot the back and forth.
Thanks for reminding.

Cheers, Finn hammer

1 2 3 4 next

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