Maker Pro
Maker Pro

MOT Winding

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
Hi you all.
I was looking for some advice, finally got round to removing the secondary high voltage winding of the MOT Ive claimed from our old microwave.

What i was hopping to find some idea as to wire diameter verses current, maths is not my strong point, so all i can find here in the retail sector is 14 SWG approximately 12 AWG the diameter is a little over 2 mm, i opted not to grind apart the ( EI ) on the transformer, the cut was made both sides, then put an 6,8,10 mm drill bit through the old secondary winding, and knocked out the shunts.

My question was does any one have an idea of the current peak if i use 2mm enameled copper wire if wound to approximately 12 volts AC, then allowed for rectification and capacitance, about 16.90 volts unregulated DC.

Thats if i can with some extra help fight winding the wire as the new low voltage secondary, tough as its going to be, i was kind of thinking of a more supple high temperature other coated copper wire, but i think all the cables i can find wont cope with the heat.

So was looking for varnish suitable for coating windings, really as insurance in case i need to use a soft mallet for uniform encouragement of the new windings.

In summary can anyone advice me on the current, and possible source of a suitable high temp varnish, i should have measured the primary resistance, but i did apply power for 30 seconds with a meter in series, no secondary winding so no load, current was about 3 amps.
Any help appreciated.
Dave. :)
PS, if i could find a thicker enameled copper wire i doubt me or extra help from a friend could manipulate it and so wind it. So i cant go beyond 2mm, or 14 SWG. Only if i can find a high temp coated multi strand.
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
Hehe, a 3A no-load primary current is quite high, especially on 240V mains.. It sure could use some extra turns there. How thick is the primay wire?
A common transformer wire load (ampacity) is 3-4A per square mm. A 2mm wire is 3.14mm2 and is good for 10-12A continously.
Winding several strands simultaneously is the easiest way for more current capacity. You'll need to calculate & cut the lengths you need anyway.
You'll therefore need to establish how many turns per Volt the primary is set up for. Wind a 10-turn test secondary and measure the voltage.
While varnish is good for keeping the winding together and for long term reliability you can not expect it to re-insulate a damaged wire coating.
Any damage will be at the most stressed points and the wires will rub against each other right there so the varnish can't build a coat.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT Rewind

Hi Resqueline. Ok that answers the varnish question, but still ok to hold windings together, just have to line the ( E ) aperture carefully. enameled copper wire 2mm is going to be a git to wind, if i multi strand that i am not sure i will get enough turns for 12 volts. Yes 720 watts before you even start is a lot i know, the primary wire diameter at a guess is about 1.1 to 1.2 mm, probably could wind a few more turns to bring that current no load down to something more exceptable. Dave.
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
To nitpick, that's 720VA, not watts (big difference here) ;). Ok, so the primary is around 1 square mm, so it's good for only one Amp more.. Thinner wires are actually more efficient at filling out the winding area (and you want to fill it out as much as possible). There are calculators on the net to find the number of turns that will fit in a given area with a given wire size, and they'll even calculate the length of the wire needed. You might want to use 2 or 3 x 1.5mm instead, but you'll have to find the winding ratio first so you know the exact number of turns to use.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT rewind

Yes sorry transformer expressed as VA, ha whats the PF G knows, i will look for a site for the formula to calculate the winding length, i only have a few reference books at home, most of the literature ive been through and use is library stuff as ours has loads of data reference books.

I had another idea come to me about the primary, i think if ive got it right in some atx power supplys ive opened there is a laminated transformer on the high voltage input side of the circuit, i dont know as ive not tested its resistance, but i am thinking its a ballast transformer as its only got two leads from the transformer or ballast to the pcb.

I know its a long shot as it could be to much resistance, but i was thinking after testing its resistance putting it in series with the primary, as its from a 500 watt atx psu it should be ok in series with the primary, at this stage it only an idea, as the MOT primary winding and the ballast might be totally a bad idea, i will look round the net see if i can get more info on this what i believe is ballast, with only two leads what else could it be, and its on the atx's primary side of the circuit.
Dave. :)
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT rewind

Hi again.
Ive measured the MOT's primary its 3.3 ohms, and the what i think is ballast in the 500 watt atx is 2.2 ohms, not sure what difference it will make, would it be 2/3 greater resistance in the primary ? not sure until i test it.
Dave. :)
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
I've seen those atx "ballasts". They are power factor correction inductors. It might be able to survive in series with the MOT, but you won't get any power out of the MOT.
I believe the PF number is simply Watts divided by VA. A pure lossless capacitor/inductor will thus have zero PF, and a pure resistance will have a PF of 1.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT rewind

So the ballast i mentioned in series with the primary would take it up to about just a bit above 5 ohms for the total primary resistance will give no power output on the rewound secondary, how can that be ? i expected some resistive loss from the primary, as its a couple of ohms higher resistance, but no power out, i dont get that, its still mutual magnetic inductance from primary to secondary, pass, not sure.
Dave. :)
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
It's not about resistance but inductance. If the ballast doesn't saturate then it'll pass only a specific current and almost no voltage is left for the MOT as load increases.
It won't give no output but it'll severely limit the power obtainable, unless the ballast happens to go into saturation as a load is put on the MOT output.
I happen to have one such ballast around. It's 2 ohms. I don't quite want to put it straight on the mains to measure the current, but on Monday I can test it with a variac.
It'll essentially do the same job that those magnetic shunts did.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
Ok i am getting what your saying about its not the resistance but rather inductance in the ballast taking it away from the MOT's primary so no power left for the MOT.

So how does the atx power supply manage its rated output, well i should reword that, the power output the manufacturers claim there supplys put out.

I would not expect any one to be silly enough to connect a ballast unit like that strait to the mains incoming voltage 110 or 240 volts AC.

A variac started at a modest value volts AC i expect will be the same as the maths, from the resistance reading, meaning the outcome will be as you would expect it to be, but coupled to the MOT in series will either be a total flop, a no go, or it might take some of the heat and current from the MOT but still be more than ample for the secondary on the MOT's output.

In all honesty i dont know, all i do know is the ballast is part of a more complex circuit in the atx psu, and with regard to the extra circuitry the atx manages a reasonable output.

As you said it might go the ballast takes to much from the MOT, its a try it and see thing maybe, nothing lost nothing gained.
Dave. :)
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
I've not checked exactly how the "ballast" is wired in those ATX supplies, but I can check it out within a few days.
If the ballast has only a small inductance it could be wired in series with the PSU, if it has a larger inductance it could be connected straight across the mains.
Its purpose would be to lag about the same current level that the PSU main capacitor leads, thus cancelling each other out, only leaving a resistive current.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT windings

I googled it to death last night, it seems some kind of ballast is desirable as apposed to none, some say use another mot and short the windings, ahhh no that ones not for me, a did find a 20 amp mot psu project this guy had built, no ballast in his psu, air cooled, and BIG capacitor, and a lot of 2N3055's for the regulator circuit, i should have bookmarked it, i cant find it again now, it was built in a tidy diy wood case with one panel meter in it, best one ive seen, most are just slung together.
Dave.:)
PS, ive gone off the MOT idea take a look at this, WOW i will build one of these 600 meters WOW
Only joking feet back on the MOT i mean ground.

http://w5jgv.com/4-400A_600M_amplifier/
 
Last edited:

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT windings

Hi to all.
Decided winding 2mm diameter enameled copper wire on the transformer would be a bit of a nightmare, trying to wind neatly on such a tight space with thick wire, well 14 SWG and about the equivalent to 12 AWG, not that it will take many windings to get to about 15 + volts or so.

Any way got it split, primary came out nicely good and clean, no damage to it, could add some windings to the primary now, make up a bobbin for winding the turns.

Just got to work out now how to reassemble, not keen on welding, heats an issue, mechanical bracket bolts etc maybe, someone said glue it to me, that does not seem a good idea as its part of the magnetic circuit hmmm, i will think of something, any reassembly ideas, as in the ( E I ) welcomed, only split it so its easier to wind, tidy, and wont look bodged up like some slung together you see on the net.

If its worth doing then i believe its got to be done in a tidy neat finish, even if the final power supply no one see's it, but i will know it there, and if its bodged it would annoy me.

Some pictures of where i am at so far.
Dave. :)

ABCD0005-5.jpg


ABCD0004-6.jpg


ABCD0006-4.jpg
 
Last edited:

jackorocko

Apr 4, 2010
1,284
Joined
Apr 4, 2010
Messages
1,284
well they come welded, so why would that be a problem? At what temperature does the enamel melt? You could drill and tap it, but it is not gonna be as good as if you weld it.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
Yes its the original way it was assembled, but factory welding was shallow and quick i expect, i dont have a mig or ark welder, a mate does, he could do it for me, but hanging around on its not an option, it needs doing quick and shallow to avoid access heat. Winding insulation temp cant be much more than 120 Degree C or maybe less, depends on class of varnish used. might opt for mechanical, 6mm studding heated bent drilled bolted through the base plate. or a copper shield cover the frame rivet it.
Dave.
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
A weld ensures that no laminates are free to buzz, and that there's a contractive force on the gap. A clamp should do the trick for you I think, glue being an optional addition.

I've measured the ballast. It's labeled 19.21022.011 and is wound with 0.75mm wire, having a couple of ohms resistance. It measures 54mH on an RCL-meter (50Hz).
At 1A it drops 20V, at 2A 33V, at 3A 39V, and at 4A 43.5V. As you can see from the graph the inductance is not linear, as the iron saturates more at higher current.
With an even sharper non-linearity I guess it could be a used as a voltage dropper in front of a MOT, not having an adverse effect on the maximum output power.
But nothing beats adding to the primary winding itself. Can you measure the winding opening/area, and estimate the number of primary turns??
 

Attachments

  • 19-21022-011 at 50Hz.GIF
    19-21022-011 at 50Hz.GIF
    7.6 KB · Views: 260

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT windings

Hi Resqueline.
Thanks for looking into that for us, i will measure the aperture length width and height, and post a drawing of it.
I guess measurement of the core will give me an idea of length per turn, but this will increase with every turn, being just over 2mm diameter wire, each layer will increase by that figure ? i think.

So is that the ballast you tested in series with the mot ?, ive found a high temperature epoxy there are two, one i am not sure of its property's, the other is a titanium based epoxy with a safe temp in operation in excess of 150 degrees C , i was going to keep to just the original primary as space is tight, it wont take long to fill the space already there. Ive read i would need an extra 50 turns on the primary to be of any use in improving saturation of the core, i think the ballast and really good quality cooling characteristics will be ok.

I was going to clad the frame in flat heat sink materials, meant for high power apps on semiconductors, there about £ 7.00 each, then get them cut to the outer frame size, including the base, fixed with the metal conductive epoxy, i think cooler running will be more efficient, and extent the transformers winding life a bit. A constant flow of forced air through the supply should keep things cool, going to put a temp alarm sounder for safety in the circuit for over temp, and a max shut down temp as well.

As i am only needing it for linear RF use i am going down the 2N3055, or TIP2955 route, looking for a max peak of 12 amps, and constant 10 amps for this psu.

Will measure the aperture and post drawing, by hand and scan, its quicker.
Dave. :)
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
The radius will increase by 2mm for every layer, yes. It's common to use the midpoint radius as basis for wire length calculations I think.
The ballast I tested above was the PFC correction "circuit" out of a Chieftek HPC-360-202 PC PSU. I can't imagine it would stand 3-4A continously though.
Forced air cooling is about four times as important as cooling area. I don't see the point of heatsinks if they're entirely flat. Or did you mean flat on one side only?
The most efficient winding utilization would be when the primary occupies (at least) as much space as the secondary winding. They'll reach the same temp's then.
The measuremens of the core/winding window amounts to just 4 numbers, no need for a drawing for that, but the heatsinking idea could need a drawing.
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT windings

I was thinking of the flat one side, finned on the other, and clad the frame in that, forcing cooling air across the unit, this will dissipate masses of excess heat thats probably going to be created by the entire circuit.

I was hoping to find ballast that could take the excess all the time continuous.
Anyway here is the dimensions of the frame.The primary takes about 1/3 give or take a little.
Dave. :)

Scan0001_converted.jpg
 

davelectronic

Dec 13, 2010
1,087
Joined
Dec 13, 2010
Messages
1,087
MOT windings

So at this point i could use a bit of insight. I cant, or am not sure how to work out the primary number of turns, how many more turns would i need on the primary to improve the efficiency ? . Ive got 20 meters of 14 SWG enamelled copper wire, and other bits together. So in a word how many more turns on the primary would i need, ideas apprecated. I thought the ballast would share the primary current, but if more primary turns is best then i will go that way.
Dave.
 
Top