EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: OiD on June 09, 2016, 03:38:31 pm
-
Hello again,
So I'd like to make a high voltage high frequency transformer as getting hold of them here in Spain is a real pain.
The idea is to use a ferrite core of a significant size to house the high voltage coil with it's insulation (it will be submerged in oil for extra insulation), now I have a rather large collection of large square ferrite beads for cables and they consist of two U sections and I'm wondering if I can stack them for make the core for the transformer...
Problems; Gaps. Every extra surface will mean a break in the field and the transformer will quickly end up in flyback territory (which is still good if it works) but with lots of the ferrites the gaps may be too big and I'd end up with nothing much.
I've also got some rectangular ferrite rods that I'm thinking of trying as well.
A few tests show me that it does work, but I have yet to quantify any values. I've tried searching online but I've yet to find anything except radio antennas made from stacked ferrite rods.
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=231590;image)
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=231592;image)
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=231594;image)
How crazy is this idea?
Cheers!
-
Try this, should me more then suitable for high voltage:
http://www.ebay.co.uk/itm/E8020-E80-Ferroxcube-3C90-E-EE-Ferrite-Cores-transformer-AL-5000-1set-/370683164776?hash=item564e703868:m:mSNeMgErZsk5z9QKqb5CAgA (http://www.ebay.co.uk/itm/E8020-E80-Ferroxcube-3C90-E-EE-Ferrite-Cores-transformer-AL-5000-1set-/370683164776?hash=item564e703868:m:mSNeMgErZsk5z9QKqb5CAgA)
http://www.ebay.com/itm/Bobbin-25x20mm-25mm-1x0-8-1-for-E80-E-EE-ferrite-core-transformer-coil-former-/251526231063?refid=store&ssPageName=STORE:HTMLBUILDER:SIMPLEITEM (http://www.ebay.com/itm/Bobbin-25x20mm-25mm-1x0-8-1-for-E80-E-EE-ferrite-core-transformer-coil-former-/251526231063?refid=store&ssPageName=STORE:HTMLBUILDER:SIMPLEITEM)
fail that bobbin, have one machined on a lathe with loads of little sub-bobbins.
-
Hi TCWilliamson,
Thanks for the link, with the way eBay's search engine is working lately I'm have a horrible time finding things there now. It's quite a bit cheaper than the other ones I've founde on eBay, but it's still a bit on the small side, looking at E110, UU160/125/20.
So far the closest thing i've found that looks remotly similar to what I want to do is a service from MH&W;
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=231846;image)
And here is a mockup that I need to find a method for glueing them together or something that can fill the small gaps between the pieces;
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=231848;image)
-
And here is a mockup that I need to find a method for glueing them together or something that can fill the small gaps between the pieces;
pic
I'd be more inclined to overlap the blocks, brick wall style, to distribute / minimise the gaps. It would be mechanically much stronger too.
-
Hi Gyro,
True, overlapping them sounds like a good idea, but I'm unsure of how to handle the 90 degree corners as the blocks are rectangular and the contact area is only on the edges as the centre is hollow... Maybe if I could precisely cut 45º angles and had all mating surfaces ground nice and flat I could get something to work! But I don't have a grinder...yet. :phew:
-
Really, you don't want to be messing around with this kind of set up, it will make a huge mess and be rather inefficient.
I strongly suggest a pre-designed core or buy a load of blocks of the suitable material and have them ground to shape to fit together, try finding a surface grinding company that deals with graphite or similar, they should be able to assist you.
-
What are the specs of the supply you need? voltage, current capability, and frequency? Judging by the size of the cores you're using I'm guessing this is >100W
If its just for arc's sparks or HV cap charging, flybacks (coupled inductors) are king. For regulated supplies, a current fed royer converter would be better. Split bobbins can be fairly straight forward to make, check out the transformers used in CCFL supplies.
-
Hi TCWilliamson, yeah, it's will be inefficiant, but if it works enough... it would be worth it as the ferrite is free. Finding a local bissunes with a grinder sound like a good idea too. For now it seems to at least work.
https://www.youtube.com/watch?v=XdGlefH2ILc (https://www.youtube.com/watch?v=XdGlefH2ILc)
Hi Buriedcode, specs are more than 100W, if it can handle 500VA o more then better, 25-75KV at high frequency, 22-100khz. I will have a look at the ideas you have suggested. I never remember about ccfl transformers as I've never seen any "big" ones just little ones.
-
There is no such a thing as one single "ferrite".
Filter ferrite (like in ferrite beads) material is designed to be lossy, so the chances are that your efficiency at higher frequencies gets so low that it does not even work at all. OTOH, we'll see how it goes.
Switcher transformer and coil cores need to be the exact right material for that purpose because the core losses are very important. I have made many switchers using "random" cores and the real total efficiencies can vary between like 20 and 80% only depending on the material!
-
What about a Tesla coil?
-
Hi Siwastaja,
Thanks for the input, I'm trying to get the datasheet for these ferrite beads. Thanks for the pointer about the frequency, I will plot some data at different frequencies and see how much it's affected by the changes. I've noticed that I lose lots of power the further away I get from the primary winding, especially after the 90º bends where the magnetic flux is limited to just a small portion of the ferrite. I' must admit, I'm surprised it even worked!
I'm constantly on the lookout for large ferrite cores, any idea what kind of appliances would have such large cores?
Hi NiHaoMike,
I'd like to be able to rectify it, and my only tesla coil puts out 75cm streamers. I must admit that I have never tried running the 6kw beast of a variac but rectifying such high voltages and frequencies sounds expensive.
My goal is to have a 20-75kv supply of a suitable high frequency and run it through a voltage multiplier to achieve around 100-120kv DC.
I've tried a 10 stage voltage multiplier under oil and i'm only getting 30-40KV instead of the 110 I should be getting from it. The final goal is to have a nice high voltage source for various experiments.
Thanks for all you ideas, please keep them coming!
-
Actually, the ferrite used in beads usually has modest losses at power-switching frequencies.
The worst I've seen, is the Laird/Steward #28, e.g.
http://cdn.lairdtech.com/home/brandworld/files/28B0375-300.pdf (http://cdn.lairdtech.com/home/brandworld/files/28B0375-300.pdf)
The inductance drops rather quickly over 1MHz (notice the Z(F) slope is not a straight 1 decade to 1 decade as an ideal inductance would be, but it falls considerably short!), while the impedance remains relatively flat, and largely resistive (as you'd expect). They make rather crappy transformers, as the core's resistance acts in parallel with the windings, and you need rather a lot of turns (compared to lower loss materials), which impacts the high frequency response of your transformer (for RF purposes).
So, naturally, they make pretty good ferrite beads! :P
#43 is a common "medium frequency" material, and has pretty reasonable performance below 10MHz. You can use it in pulse and power transformers at those frequencies, without much loss.
#31 is another low frequency material, but it's pretty much a standard "power" ferrite, for frequencies under 1MHz.
The main things to avoid are the lossy low frequency materials (that will noticeably affect your circuit at power-switching frequencies), and the low permeability, high frequency materials (#61, #67) which won't give you enough inductance to get good transformer action.
Speaking of, air gaps do the same thing as reduce your permeability, so ground faces, overlapping core pieces, and all that, are the way to go. As others have already noted.
You also want to put the primary and secondary on top of each other, in case you're wondering. :)
Tim
-
Hi T3sl4co1l,
Thanks for the info about the materials, guess I've struck gold with the #28's that I've used!
http://cdn.lairdtech.com/home/brandworld/files/28A2024-0A2.pdf (http://cdn.lairdtech.com/home/brandworld/files/28A2024-0A2.pdf)
Well, I guess that they are the way they are to do their intended function... RF attenuation!
Putting the coils on top of each other is ideal, but the main reason I wanted a reasonable sized coil is to keep distance between the high voltage side and low voltage side.
I'm guessing that giant flybacks would be suitable for their cores, but their pretty hard to find now days :-//
-
Hi again,
Regarding the "succes" of the ferrite bead core I made another one with some russian ferrites from ebay. I believe the material is n67. It's made from 16 pieces of ferrite, there are less gaps and it's better material. The core gets pretty warm running around 100W at approximatly 25kHz. This time I wound a better secondary, about 1500 turns giving me a bit over 5KV output. As expected I destroyed it but as it's just a quick coil... :popcorn: You can see the failire at the end of the video.
Here is a photo of the ferrite core;
(https://www.eevblog.com/forum/projects/ferrite-cores-and-diy-alternatives-for-a-transformer/?action=dlattach;attach=236226;image)
https://youtu.be/2s5QEsNoqg4 (https://youtu.be/2s5QEsNoqg4)