So yeah, I'm kind of a newbie but I'm working on rewinding a transformer I had lying around. It's an EI transformer with a plastic split bobbin and it's original output was stated at 12-14.4v at 1.8A. So it's around 20 - 25VA I guess? I took it apart and from measuring the laminations, calculated the area of the core to be 27.28 cm square.
When I started I didn't know you weren't supposed to unwind the primary. I figured to it working ideally you should unwind both windings which is what I did and I didn't count the number of turns (stupid, I know).
Anyways so my target voltages are 120v @ 60hz (mains voltage) on the primary and 200v on the secondary. It shouldn't have to deliver a lot of current. Less than 100ma at that voltage, I figure.
Using the information from this page (Transformers Part 1 - Beginners' Guide to Electronics) I calculated a turns ratio of 1.82 and used 220 windings on the primary and 366 on the secondary (Yeah, I rounded to the nearest whole number). I wound it by hand but I wound it as tightly as I could. I used new magnet wire from the source (used to be Radio shack here in Canada). I then varnished it (windings and laminations) as well as I was able and let it set a couple days.
So obviously I did something very wrong because when I tested it, it overheated and melted the insulation on the primary. The secondary looks fine though. It was fairly quiet though and the windings are still tight. I'm thinking my turns per volt was too low? If you use more turns it should generate less heat right? Transformers are complex and there's a lot about them I don't understand for sure.
Any help would be much appreciated! Thanks in advance.
When I started I didn't know you weren't supposed to unwind the primary. I figured to it working ideally you should unwind both windings which is what I did and I didn't count the number of turns (stupid, I know).
Anyways so my target voltages are 120v @ 60hz (mains voltage) on the primary and 200v on the secondary. It shouldn't have to deliver a lot of current. Less than 100ma at that voltage, I figure.
Using the information from this page (Transformers Part 1 - Beginners' Guide to Electronics) I calculated a turns ratio of 1.82 and used 220 windings on the primary and 366 on the secondary (Yeah, I rounded to the nearest whole number). I wound it by hand but I wound it as tightly as I could. I used new magnet wire from the source (used to be Radio shack here in Canada). I then varnished it (windings and laminations) as well as I was able and let it set a couple days.
So obviously I did something very wrong because when I tested it, it overheated and melted the insulation on the primary. The secondary looks fine though. It was fairly quiet though and the windings are still tight. I'm thinking my turns per volt was too low? If you use more turns it should generate less heat right? Transformers are complex and there's a lot about them I don't understand for sure.
Any help would be much appreciated! Thanks in advance.
Before you power it up next time, measure the primary and secondary inductances. Your Primary Inductance needs to be fairly high of at least 500 mh if not much higher. Otherwise your primary impediance (Z = L*2*Pi*Freq) will be too low allowing too much current to flow into the primary, causing a meltdown. Typically an small 120VAC@60hz transformer will have thousands of turns on the primary. The only way to have a low turn count is to use a much higher frequency (measured in Kilohertz) used by Switch Mode power Supplies. Perhaps some low power transformers may have fewer turns because the use very thin gauge wire that has a fairly large resistance per meter, which can be added to to impediance. R = Z + DC-resistance. I (current) = V (voltage)/R(Resistance). Ideally you want the transformer to draw current measured in 10's of Millamps with no load on the secondary.
Note that when measuring primary and secondary inductances, you leave the windings open. Shorting one winding while measuring the other winding will provide you the inducance leakage. Generally, Split Bobbin transformers will have a fairly high leakage inductance.
I suspect that you make have trouble winding enough turns on that transform to produce the 200Volts you need. You'll need a lot more enamel wire to wind it then supplied on the typical Radio Shack Magnetic wire spools. In my opinion, you better off building a flyback or half bridge unregulated power supply. All you need is a power transistor, an oscillator (such as the LM555 timer IC), and a ferrite core transformer, and a DC input source between 12V and 24V. Google for "flyback 555". This should turn up a few good reference designs. I would go with 2 turns per volt on the primary (ie 24 turns @ 12V and 48 turns @ 24V), and use the standard turn ratio calc to get your target output voltage. I recommend add 5% extra turns on the secondary and adjusting the duty cycle on the 555 to fine tune the output voltage. Its a lot easier to mess with the switching duty cycle than it is to rewind the secondary. I would also add a series output inductor (yellow -powder iron) between the voltage input and the transformer primary (ie Vdd--Inductor---Primary--Transistor--Ground). The inductor will help limit the current draw into incase you supply too much duty cycle or excess output current that causes the transformer to saturate (causing impediance to drop to zero). Typically when the transformer saturates the switching transistor will blow or get excessively hot. Consider the extra inductor as bicycle training wheels. You can always remove the inductor if your satified with your design, or just leave it in for a safety margin. Make sure that the 555 duty cycle never exceeds 50% (transistor is switched on for more than 50% of the switching cycle) or the transformer will saturate.
best of Luck to you.
Note that when measuring primary and secondary inductances, you leave the windings open. Shorting one winding while measuring the other winding will provide you the inducance leakage. Generally, Split Bobbin transformers will have a fairly high leakage inductance.
I suspect that you make have trouble winding enough turns on that transform to produce the 200Volts you need. You'll need a lot more enamel wire to wind it then supplied on the typical Radio Shack Magnetic wire spools. In my opinion, you better off building a flyback or half bridge unregulated power supply. All you need is a power transistor, an oscillator (such as the LM555 timer IC), and a ferrite core transformer, and a DC input source between 12V and 24V. Google for "flyback 555". This should turn up a few good reference designs. I would go with 2 turns per volt on the primary (ie 24 turns @ 12V and 48 turns @ 24V), and use the standard turn ratio calc to get your target output voltage. I recommend add 5% extra turns on the secondary and adjusting the duty cycle on the 555 to fine tune the output voltage. Its a lot easier to mess with the switching duty cycle than it is to rewind the secondary. I would also add a series output inductor (yellow -powder iron) between the voltage input and the transformer primary (ie Vdd--Inductor---Primary--Transistor--Ground). The inductor will help limit the current draw into incase you supply too much duty cycle or excess output current that causes the transformer to saturate (causing impediance to drop to zero). Typically when the transformer saturates the switching transistor will blow or get excessively hot. Consider the extra inductor as bicycle training wheels. You can always remove the inductor if your satified with your design, or just leave it in for a safety margin. Make sure that the 555 duty cycle never exceeds 50% (transistor is switched on for more than 50% of the switching cycle) or the transformer will saturate.
best of Luck to you.
Last edited:
Actually the inductance is part of it. If the flux density is high enough, the core will saturate and you will have meltdown. You must know the flux density; the core material data sheet will show you where saturation occurs, and you need to make sure the inductance is high enough to keep the Volts per turn low enough to stay out of saturation.
In other words, transformer design is not a simple thing, such as a resistor or capacitor. You need to know the core losses and the copper losses and the flux density so that they are within reason. The primary number of turns will control the flux density at a given frequency and voltage; the primary wire size is determined by the window area - usually the primary will occupy about half the window area.
The power handling of a given transformer core is determined by the window area, and little else, given a particular conductor material.
Copper and core losses should be about the same at full load. Core loss is a combination of hysteresis and eddy current losses. Eddy currents are minimized by many thin laminations but that reduces iron area.
And so on.
In other words, transformer design is not a simple thing, such as a resistor or capacitor. You need to know the core losses and the copper losses and the flux density so that they are within reason. The primary number of turns will control the flux density at a given frequency and voltage; the primary wire size is determined by the window area - usually the primary will occupy about half the window area.
The power handling of a given transformer core is determined by the window area, and little else, given a particular conductor material.
Copper and core losses should be about the same at full load. Core loss is a combination of hysteresis and eddy current losses. Eddy currents are minimized by many thin laminations but that reduces iron area.
And so on.
I would not recommend you wind your first transformer with these voltages. You need to know a lot more about insulation in particular. Its less dangerous had you left the primary alone, and rewound a low voltage secondary, Even then you would have needed adequate insulation between primary and secondary, and of course between coils and laminations. The danger is both shock. which at these voltages can easily be be fatal, and fire. In fact, unless you are talking about quite large low voltage transformers, the saving is not worth either the risk, or the effort.
No, the power handling of a given core is dependent on window area. I was designing transformers for a while and ran an analysis to determine required core size. The amount of power a given core can handle is NOT dependent on the core size but on the window area. Basically, you have to cram as much copper into the window as possible to get as much power as possible. This in turn is limited by the resistivity of the wire. If you use silver wire you can get more power.
It's a complex relationship.
It's a complex relationship.
Possible Layer Shorting
If everything else was right, perhaps the varnish dissolved the insulation on
the magnet wire and caused the windings to layer-short. Compatibility
between the varnish and enamel magnet wire insulation is critical. I've
rewound many small transformers in my lifetime and can only remember
of one that primary-secondary shorted. "God looks out for drunks and
fools." The equipment touched a grounded metal outlet box, sparked,
then internally blew the short in the transformer open. I could have just
as easily been electrocuted, had my body made that connection. I,
personally, advise against rewinding a power transformer unless 1) the
primary winding goes directly to solder lugs on a plastic bobbin(windings
spliced to lead wires have a much higher risk of primary-secondary
shorting or shorting to frame.) or 2) The rebuilder is very knowledgeable
on insulation technique and insulation safety!(A Local Agency-Approved
transformer is usually High Potential Tested at about 20X normal line
voltage, Pri-Sec and Pri-Frame for at least a minute!)
If everything else was right, perhaps the varnish dissolved the insulation on
the magnet wire and caused the windings to layer-short. Compatibility
between the varnish and enamel magnet wire insulation is critical. I've
rewound many small transformers in my lifetime and can only remember
of one that primary-secondary shorted. "God looks out for drunks and
fools." The equipment touched a grounded metal outlet box, sparked,
then internally blew the short in the transformer open. I could have just
as easily been electrocuted, had my body made that connection. I,
personally, advise against rewinding a power transformer unless 1) the
primary winding goes directly to solder lugs on a plastic bobbin(windings
spliced to lead wires have a much higher risk of primary-secondary
shorting or shorting to frame.) or 2) The rebuilder is very knowledgeable
on insulation technique and insulation safety!(A Local Agency-Approved
transformer is usually High Potential Tested at about 20X normal line
voltage, Pri-Sec and Pri-Frame for at least a minute!)
EI winding
It's hard to wind square window coils without making a shorted turn. I know, I did it. I measured my transformer resistance wire to iron before I plugged it in- less than 1 ohms, means I had a possiblity of a shorted turn. End of project. It was a 120 VAC pri, 36 VAC sec, 70 amp output golf cart charger transformer. Would made quite a good arc welder. Bottom line, you can't hipot test at home, but measure resistance coil to iron before you plug it in. Easy to knick that fine magnet wire insulation.
It's hard to wind square window coils without making a shorted turn. I know, I did it. I measured my transformer resistance wire to iron before I plugged it in- less than 1 ohms, means I had a possiblity of a shorted turn. End of project. It was a 120 VAC pri, 36 VAC sec, 70 amp output golf cart charger transformer. Would made quite a good arc welder. Bottom line, you can't hipot test at home, but measure resistance coil to iron before you plug it in. Easy to knick that fine magnet wire insulation.
Excellent Reference Info!
The information in the link was perfect and I bookmarked it for my future
use, thanks for the link! Your transformer, if stated V and A are correct,
is about 25va. Typical EI cores of that size do need about 10 turns/volt.
The author's technique of overwinding 10 turns, measuring the voltage,
then dividing by 10 to get turns/volt works every time. I tend to be
somewhat ADD, so I check my numbers a few times to be sure. A great
safety procedure is to use highest ohms setting on a multimeter to verify
pri/sec and pri/frame isolation, then apply power to the unloaded
transformer through an incandescent light bulb of equal or slightly lower
watt rating than the va rating of the transformer. The bulb should glow
dimly, if at all. A well-built transformer with a low-loss core will not cause
the bulb to glow visibly, but will almost always warm the bulb's outer
surface.
The information in the link was perfect and I bookmarked it for my future
use, thanks for the link! Your transformer, if stated V and A are correct,
is about 25va. Typical EI cores of that size do need about 10 turns/volt.
The author's technique of overwinding 10 turns, measuring the voltage,
then dividing by 10 to get turns/volt works every time. I tend to be
somewhat ADD, so I check my numbers a few times to be sure. A great
safety procedure is to use highest ohms setting on a multimeter to verify
pri/sec and pri/frame isolation, then apply power to the unloaded
transformer through an incandescent light bulb of equal or slightly lower
watt rating than the va rating of the transformer. The bulb should glow
dimly, if at all. A well-built transformer with a low-loss core will not cause
the bulb to glow visibly, but will almost always warm the bulb's outer
surface.
measuring the resistance from primary to secondary to core should be in the gigaohm range... or hundreds of megs for larger stuff.
@bob91343
transformer ratings are dependant on both core area and window area.
today it is best to throw most of those old rules of thumb out.
toss a test coil on it, measure the iron loss with a good wattmeter (kill-a-watt meters need to be calibrated, don't take them at face value below 40 watts!) , subtract the copper loss from the readings with a dc test.
core to copper loss ratio should reflect the average power to peak power ratio to minimise cost and and your electric bill, there is no reason to run it 50-50 unless its running full load 24-7
@bob91343
transformer ratings are dependant on both core area and window area.
today it is best to throw most of those old rules of thumb out.
toss a test coil on it, measure the iron loss with a good wattmeter (kill-a-watt meters need to be calibrated, don't take them at face value below 40 watts!) , subtract the copper loss from the readings with a dc test.
core to copper loss ratio should reflect the average power to peak power ratio to minimise cost and and your electric bill, there is no reason to run it 50-50 unless its running full load 24-7
wind on 10 turns.
Power up and measure the Vac of those 10 turns.
Then calculate how many turns to get your 30Vac, 14Vac & 12Vac.
Remember that you will be measuring open circuit voltages. The actual voltage under load on each of the windings will vary with the loading.
You decide how many extra turns on each winding to give the required loaded voltages for your circuit loadings.
Power up and measure the Vac of those 10 turns.
Then calculate how many turns to get your 30Vac, 14Vac & 12Vac.
Remember that you will be measuring open circuit voltages. The actual voltage under load on each of the windings will vary with the loading.
You decide how many extra turns on each winding to give the required loaded voltages for your circuit loadings.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.