| Electronics > Projects, Designs, and Technical Stuff |
| Noob plays with transformers |
| (1/8) > >> |
| rebelrider.mike:
Hi folks. It's been a while since I've been on this forum, but I've gotten back into playing with electricity lately, and to do some of the stuff I want to do, I'll need some transformers. For my very first transformer build, I found an autotransformer from a broken UPS. Being a noob, I cut the wire out rather than dismantling the core first and just unwinding it. I still have the wire, but it's now in 20 foot sections. At least I still have the intact spool and core. Now, when I started this, I knew nothing of transformers or the math behind them, so I did the best I could. To start with, I figured Watts, Volts, and Amps on both sides of the transformer. Watts being the same on both sides. I never did figure out exactly how to estimate the maximum power a core could handle. I found an equation, core area (sq cm) squared = Watts. I still don't think that's quite right, but it's all I could find, so I used it. I came up with about 122W. With 122W, at 122V, that's 1A on the primary side. I used 22 AWG wire for that side as it can handle 1.2A. With 122W, at 9.7V, that's 12.7A on the secondary side. That's way more than I want, so I chose 18 AWG wire, as it can handle 3.2A. More than enough to power what I need. While figuring all this out, I managed to get the core sheets apart so I could rewind the spool. The number of windings on the primary side was not straightforward to me at all. I finally found an equation for 50/60Hz transformers: 42 / core area (sq cm) = Turns per Volt. This seemed promising as 42 is the answer to life, the universe, and everything. I learned later that this was a very simplified version with several assumptions in it. But my understanding is that you have to have a minimum number of turns per Volt, or the magnetic field will become saturated. This would cause current to start flowing excessively between the Live and Neutral wires, and perhaps trip the circuit breaker. I think that's how it works, anyway. So, with the equation, I had 42 / 11.1 sq cm = 3.78 turns per Volt. At 122V, the primary would need 462 turns. I did my best to estimate the length of wire I'd need, and came up with 244 feet, so I bought 250 feet. Turns out, I was still about 2 feet short, so I need to come up with a better way to estimate wire length. But it all worked out, and I ended up with a nice neat primary coil. (Which I later wrapped in tape before adding the secondary coil.) For the secondary winding, I guess I could have used the ratio of Volts, but I already had turns per Volt so I used that. For 9.7V I would need 37 turns. I didn't take inefficiency into account (because I don't know how), and so I ended up with 8.7V once I got the transformer put back together. But it was easy enough to add 4 more turns, and I got 9.6V. Close enough! Also, I got to play with my new pocket oscilloscope. I've never had one before, and am still learning how it works, but I did get a nice sine wave, plus several numerical values that my multimeter can't do. After confirming the transformer would actually work, I decided to make some improvements. I got some nuts, bolts, and washers at the hardware store, and a few L brackets to tighten up the core, and give it a way to be mountable. Squeezing the sheets together with the bolts reduced the buzz to the point where I can barely hear it. I also found a really nice power cable with a ground to replace the old lamp cable I was using before. I had a lot of secondary wire left over, so I used the rest of it to make extra taps in order to have different Voltages to choose from. I wrapped the outside with tape, and squeezed a couple pieces of cardboard between the coil and the outside parts of the core. Just to make sure the wires will never touch the metal. In addition to the 9.7V I wanted, I also tried to get 12, and 24V. I ended up with 12.8V and 24.7V, which is fine. By using different combinations of taps, I also got 3.1V, 15.1V, and 12.1V. I don't know what I'll use these for, if anything, but there they are. :) These are all open Voltages, so they drop a little under load, but not by much. Since building this transformer, I've continued to search, read, and watch videos. I know more now than I did, but not as much as I'd like to. I found a more involved equation for turns per Volt (T/V = 1/(4.44 x 0.0001 x B x CA x f) which tells me I probably had more windings than I needed. I've also heard a rumor that the core doesn't actually have a limit for Watts. It's all dependent on the windings. I don't know if that's true or not. And there's still more websites I haven't read yet. I'll apply all this to my next transformer. |
| timh2870:
Without doing any math, there's a pretty simple way to find out if you have enough turns on the primary; put a lightbulb in series with your coil. If it lights, then add more turns. Reactive power should cancel out, and all you're left with is the resistive losses of your primary winding and core losses in the iron. |
| exe:
Just in case, for transformers output voltage is specified under nominal input voltage and full load. Without load output voltage is higher. Sometimes it specified as "regulation", that is ratio (in percents) saying how much voltage is higher when there is no load. Bigger transformers have better regulation. Small ones may be very bad at it. |
| rebelrider.mike:
For my next transformer, I decided I want to build a 1:1 isolation transformer as a safety device while I work on various electrical projects. I found this guy at the local Habitat Store: It's an autotransformer, so no good as is for isolation. I'll have to rewind it. As I was taking it apart, I found that it has previously been exposed to moisture. The paper wrap is fine, as is the winding, so it probably just sat out in a barn or something for a while. But the core sheets are a mess, and will have to be cleaned up. I was able to remove all the winding, and will be able to reuse it. The spool is in perfect shape too. Looking at the wire, I discovered that it is aluminum and not copper. It only looks copper because of the insulation. It is 1.3mm, which works out to about 16 AWG. That would normally support 5.2A current, but since it's aluminum, I'll figure it could handle only 3.2A. I'll do some math later to confirm this. Anyway, to find the number of primary windings, I'll use the more complicated equation I found. That is Turns/Volt = 1 / (4.44 x 0.0001 x CA x B x f. f is the frequency of the change in current. In this case 60Hz B is the maximum flux density measured in Webers/sq m. Or sometimes Gauss. In this case, I'm using the average value for iron/silicon cores. About 1.2 Weber/sq m. CA is the area of the core. In this case, 4.4cm x 5.1cm = 22.44 sq cm 0.0001 is a conversion factor to make Weber / sq m work with the sq cm of the core. 4.44 is a value derived from calculus that I don't understand. Let's just call it a constant that is good for a transformer using sine waves. 1 / (4.44 x 0.0001 x 22.44sq cm x 1.2 Weber/sq m x 60Hz) = 1.39 T/V My household Voltage seems to vary from 121V to 123V, so I'll go down the middle and pick 122V for the primary side of the transformer. 122V x 1.39 T/V = 170 Turns. Normally I'd be shopping for wire at this point and would have to estimate the length. For my last transformer I used the core circumference per turn rather than the spool circumference. I think that is why I came up short when I bought that wire. This time I'll use the spool circumference of 20.6cm. I'll also add 10% extra. 170 Turns x 20.6cm/Turn x 1.1 = 3854cm / 30.48 cm/ft = 126ft. Another mistake I made last time was not to take into account that the spool will be larger around with the primary winding on it. This would be difficult to figure if I were buying both winding wires at once, but in this case I'm cheating and reusing the old wire. And I've already got the primary wound, so I can measure the new spool circumference: 27.6cm. So for the secondary winding, also 170 Turns, and adding 10% extra: 170 Turns x 27.6cm/Turn x 1.1 = 5163cm / 30.48 cm/ft = 169ft. Again, I'm cheating with already having the old wire, so I'm not actually measuring the length. After the primary winding, I've still got lots of wire left over, so I'm not worried about having enough. It took a little more effort to get the used wire nice and tight and straight on the spool, but I think I did pretty good. The covers were dented and had some rust under flaking paint, so I got them cleaned up and did a little body work. I found some old appliance enamel paint and gave them a new coat. I found some new insulated connectors to use on both sides, and they fit the existing holes perfectly. I even painted one green for ground. The ground will be on the primary side. The secondary, in order to remain isolated, will not have a ground. |
| planet12:
Firstly, good on you for getting stuck into transformers, black magic to most. I've now taught myself to design and make transformers from standard 50Hz iron-core through to >100KHz ferrite ones for SMPS use - understanding the magic is awesome. A great resource, if you haven't seen them before: http://ludens.cl/Electron/Magnet.html and http://ludens.cl/Electron/trafos/trafos.html Additionally, a warning: be very careful reusing magnet wire; the insulation is very thin and has a habit of getting fragile with age, making shorted turns / isolation failures more likely. |
| Navigation |
| Message Index |
| Next page |