transformer question

[soadrage7654]

Thanks for the input here. So now that we are sure that it will or can be made to work, I was wondering about putting it together. I have never messed with AC voltages before, and I was wondering how safe this is. Would it be ok to try out the circuit on a breadboard first? Should I be trying this in my dorm room?

 

[The Electrician]

You will have to get yourself a line cord for the primary. A good place to get one cheap is to find some piece of equipment that is being thrown away, and cut the line cord off. After you have connected the line cord to the two black wires (the secondary side has two yellow wires and one black wire; the primary side has only two black wires), you must insulate the connections with suitable electrical tape, or preferably heat shrink tubing.

Once you have done that, you will be safe from dangerous electrical shock. The 12 volt secondary won't harm you if you touch it, but good practice dictates that you don't touch (with your fingers; touching with voltmeter probes is ok, just be careful not short anything with the probes) parts of the circuit if it is energized. Unplug the line cord when you are making changes to the project.

The remaining danger is overheating, if you overload the secondary circuit. I checked my transformer and with a dead short across the 12 volt secondary, I measured a current of 3.3 amps, with a corresponding primary current of 4/10 amp (the transformer was dissipating 40 watts with the secondary shorted). This is much too small to trip the mains circuit breaker, but if such an overload is allowed to persist, the transformer will seriously overheat, and could even catch on fire. However, if you accidentally short or overload the transformer, no damage will be done to the transformer if you notice and remove the overload within, say, 10 seconds or so. However, other parts of your circuit may be damaged if the overload current passes through them. If an overload persists, the transformer will begin to smell before it catches on fire, and this will provide your final warning that something is wrong!

The first time you apply power and subsequently each time you test the circuit after making changes, monitor the secondary current when you power it up. Since you probably don't have two ammeters, put a 1 ohm, 2 watt resistor in series with each yellow wire (you can remove them when the project is complete). A sustained short could overheat the 1 ohm resistors, but you will have enough time to make a measurement and notice that you should shut it down. Use your voltmeter (on AC volts range) to measure the voltage across each resistor when you power up; the value of the voltage is the same as the current in amps when the resistor is 1 ohm. If either voltage is over .45 volts RMS (.45 amps is the rating of the transformer), the transformer is being overloaded, and this condition shouldn't be allowed to persist; troubleshooting is called for. Don't walk away while it's plugged in until you've completed it and are certain there's no overload under normal use.

There shouldn't be a problem doing this in your dorm room if you don't start a fire!

You really should have a voltmeter (DVM) while you're working on this. If you can't afford one, it might be better to work on the project in a lab at your school where a voltmeter would be available.

 

[soadrage7654]

Thanks for helping me get started on this. Yes, I have a digital multimeter (it was cheap, like $5), but it only has AC settings for 750V and 200V with a 4 digit display. That means I could only get accuracy to the nearest tenth. I guess that this would be accurate enough, right?

I also have heat shrink tubing and can easily get those 2W resistors. Getting the cable might be interesting, but I could probably find a way. I just need the two-pronged plug, since it's only 120v and not 240v right?

 

[The Electrician]

Thanks for helping me get started on this. Yes, I have a digital multimeter (it was cheap, like $5), but it only has AC settings for 750V and 200V with a 4 digit display. That means I could only get accuracy to the nearest tenth. I guess that this would be accurate enough, right?

Just barely good enough. You might want to take it to your school lab and see how well it does with voltages in the .1 to .5 range; compare to a reading on a meter with a lower range. If you have a serious overload with your project, the reading should be maybe 1 or 2 volts. You could also use 2 ohm resistors instead of 1 ohm; then a reading of .9 volts would be the overload threshold. At the expected maximum output you said you need, 100 mA, you will only lose .2 volts of the voltage from your rectifier; that should be tolerable if you have 120 VAC input to the transformer.

I also have heat shrink tubing and can easily get those 2W resistors. Getting the cable might be interesting, but I could probably find a way. I just need the two-pronged plug, since it's only 120v and not 240v right?

An ordinary 2 wire line cord will be ok. Maybe your local computer recycling place could give you a line cord.

 

[The Electrician]

Another consideration that occurs to me is the question of whether your DVM reads "true RMS AC+DC"; I suspect it probably doesn't. Usually, meters that do read true RMS, say so. If it isn't, then it will read low on the current from the yellow wires. For the expected rectifier "pulse" waveform, the reading you get on an "average responding" meter should be multiplied by about 1.5 to get the correct reading. This is only a rough approximation, and for any other waveform the correction factor will be different.

This means that you won't know accurately if you are overloading the transformer. You may want to work on this project in a school lab where you can use a suitable DVM, one that can read "true RMS AC+DC". Ask your instructor if they have such a meter. The Fluke model 189 is an example of a suitable meter.

 

[be80be]

Thanks for helping me get started on this. Yes, I have a digital multimeter (it was cheap, like $5), but it only has AC settings for 750V and 200V with a 4 digit display. That means I could only get accuracy to the nearest tenth. I guess that this would be accurate enough, right?

I also have heat shrink tubing and can easily get those 2W resistors. Getting the cable might be interesting, but I could probably find a way. I just need the two-pronged plug, since it's only 120v and not 240v right?

There a store off of western ave SHIELDS Electronics-Datacom Supply...The Distributor with the National EDGE! check it out

 

[RCinFLA]

One thing I don't think anyone mentioned. The transformer output voltage is usually specified at the rated load current. Especially for a Radio Shack transformer you will have an even higher voltage with the lighter 100 mA load.

I suggest you use it in a full wave, CT configuration. This takes two diodes, one on each of the two outside (12v) leads. CT is grounded. Cathodes of both tied together to filter cap. This is like doing a full wave bridge rectifier on half the transformer secondary but has half the rectifier voltage drop of a bridge config and only uses two diodes.

The output voltage is A.C. rms voltage at full rated load. (are you sure its not 12.6 vac xfmr?) When you rectify and put a filter cap on output you will be near peak of sinewave where the sqrt(2) comes in.

Assume you have about 6.4 vac *1.414 = 9.0 v
9.0 - 0.8vdc (for normal silicon rectifier, use 0.4 if Schottky) = 8.2 vdc peak on filter cap.

For 7805 regulator you should ensure at least 2.5 vdc overhead meaning filter cap voltage should not sink below 7.5 vdc.

To give some margin lets say we can allow 0.4v ripple on filter cap.

Caps size for:

500 mA output would be: I= CdV/dt, 500 mA = C * 0.4v / 8.333 msec. (time between pulses of full wave rectifier is 1/120Hz = 8.333 msec.)

C = 10,000 uF at least rated voltage of 10 vdc.

For 100 mA output would be C = 100 mA * 8.333 msec / 0.4 vdc = 2,000 uF @ 10 vdc.

7805 heat dissipation for 500 mA = 500 mA * 3.5v = 1.75 watts ( needs a small heat sink)

7805 heat dissipation for 100 mA = 100 mA * 3.5v = 0.35 watts ( no heat sink needed)