EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: dentaku on March 08, 2014, 10:52:51 pm
-
Last night was measuring the AC voltages of various taps of a transformer I found in a stereo and want to build a power supply with.
This morning I measured it again and noticed it was about 1V lower. The dryer was running.
Tonight I see that it's slightly different again.
Also, I was measuring the power from the outlet and it was about 120V AC then when the water pump switched on, my meter read 122V.
If I build a power supply with LM317/337, are they fine dealing with fluctuations in power and is this just normal for most houses?
-
As long as the input voltage is always at least 2v higher than the desired output, then the LM317/337 will have no problem giving you a stable output voltage.
So you have to rectify the ac output of your transformer (and account for the voltage drop on the 2 diodes that are active at any time in a bridge rectifier, at both low and high currents) and use enough capacitance to minimize the variation of voltage at the input. If capacitance is unrelistically high for your needs, it will be easier to pick a transformer with higher ac output in the first place.
-
if you measure around 2PM, it will likely be a good 5V higher than at 9PM at night, that is the contribution of neighbors solar panels pumping power in locally and the power grid running a bit high in expectation of large industrial loads,
As for the line voltage becoming higher when switching on your pump, if you have a large amount of switch-mode devices without PFC then an inductive load could be improving your power factor and thus lowering the voltage drop to your house,
as its a 317/337 pair the highest ideal center-tap transformer you could get away with is ~26VAC (+-35V DC when rectified), which means for every volt your mains changes, it only changes about 0.2V on the un-rectified AC, which is going to be less than the ripple from your smoothing capacitor unless you use a very big value or drive a very small load,
when building your supply, aim to run it 5V below the rectified mains DC voltage to give yourself some head room, as if the input voltage drops to less than 2V below your desired output it will then drop, (drop-out voltage) if you do this you wont even notice your mains shifting,
-
If I build a power supply with LM317/337, are they fine dealing with fluctuations in power and is this just normal for most houses?
Yes, this is exactly what they are designed to do. They keep the output voltage the same even if the input voltage varies or the output load varies.
Fluctuations in mains voltage are completely normal. Your mains supply comes from the secondary winding of a transformer, and the voltage out of that transformer depends on how much power you and your neighbors are drawing from it.
Depending on the circumstances, you might find your mains voltage as high as 125 V or as low as 115 V.
-
So long as you have enough headroom above your regulated voltage, these regulators will not be affected by small changes in input voltage. The output voltage is determined entirely via the feedback from the voltage divider between Vout and Vadj, completely independent of Vin.
An LM317, for example, will constantly adjust its output voltage to maintain 1.25V on Vin, regardless of any change in input voltage. The output voltage is determined by the ratio between R1 and R2, chosen as required to produce 1.25V in the voltage divider at the desired output voltage.
(http://www.circuit-innovations.co.uk/LM317T.jpg)
-
As for the line voltage becoming higher when switching on your pump, if you have a large amount of switch-mode devices without PFC then an inductive load could be improving your power factor and thus lowering the voltage drop to your house,
Also, if there is a heavy load on one half of a split phase supply it could pull the neutral to one side and increase the voltage on the other side?
-
Mains voltage will vary. The biggest thing that will make it change is load. More amps means larger voltage drop.
The POCO usually tries to keep it within a certain range, but their official ranges tend to be fairly broad. Here on a 120V circuit, the POCO is okay with anywhere from 110V to 125V
-
I spent a while searching around for answers which I already pretty much knew but just wanted to make sure but asking people on this forum got me better answers so I'm glad I asked.
I finally found a transformer from a stereo with a power switch and fuse already wired into the cord that gives me around +/- 23V DC after setting up a Bridge rectifier and some big caps but I was still curious about voltage fluctuations and adjustable regulators.
Thanks for the well worded info Eevblog regulars :)
Now I just need to find a good circuit I like. They're all extremely similar because it's not like it's terribly complicated but I just want to find one I'm happy with before moving off the breadboard and moving to something permanently soldered. I've found some schematics that omit protection diodes (I'm not gonna do that) and some that have them in slightly different places and others have resistors at the outputs to ground like http://www.musicfromouterspace.com/analogsynth_new/POWERSUPPLY2009/POWERSUPPLY2009.php (http://www.musicfromouterspace.com/analogsynth_new/POWERSUPPLY2009/POWERSUPPLY2009.php)
-
Now I just need to find a good circuit I like. They're all extremely similar because it's not like it's terribly complicated but I just want to find one I'm happy with before moving off the breadboard and moving to something permanently soldered.
Start with the examples in the datasheet (e.g. TI (http://www.ti.com/lit/ds/symlink/lm317.pdf)). The big problem with many circuits you'll find around the place is that they're based on circuits which are based on circuits which are based on the circuits in the datasheet, and stupidity creeps in with each iteration.
You've already shown your smarter than that by deciding to keep the diodes :)
-
In the circuit in the last link.
R1 and R5 should actually be about 100-120 ohm, not 240 ohm. Some versions of LM317 or LM337 (depends on manufacturer) need a bit more current going into the adjust pin, so by lowering the resistance you do that.
R3 and R7 (1kOhm each) are the resistors that make the circuit output a reasonable output range, by restricting the minimum voltage.
minimum ... Vout = 1.25 ( 1+ (R3+R4)/R1) = 1.25(1 + (1000+0 ) /240) = 6.45v
maximum ... Vout = 1.25 ( 1+ (R3+R4)/R1) = 1.25(1 + (1000+2000 ) /240) = 16.875v
If you change R1 and R5 to 100 ohm to make your circuit more compatible with various LM317/337 chips and drop R3 and R7 to smaller values (100ohm for example)
minimum ... Vout = 1.25 ( 1+ (R3+R4)/R1) = 1.25(1 + (100+0 ) /100) = 2.5v
maximum ... Vout = 1.25 ( 1+ (R3+R4)/R1) = 1.25(1 + (100+2000 ) /100) = 27.5v
Even if you'd decrease the potentiometer to 1k, the maximum configurable voltage would be 15v.
I suspect the additional 4.7k resistors on the output R2 and R6 are there just as minimum loads, to consume some current. Like I said above, some LM317/LM337 will only output a stable voltage with a minimum current used.
V = I x R ... at 15v out, you'd have 15 = I x 4700 -> I = 0.003A = 3 mA.
3mA plus the current in the adjust pin is enough to make the regulators stable.
If you change those 240 ohm resistors to 100-120 ohm, there will be more current going into the adjust pin and therefore you could reduce the current consumed in those output resistors if you really want to. For example, going with 10k resistors instead of 4.7k would probably work just fine.
And C1 and C9 don't have to be 1uF .. you can go higher, to 10-47uF, 35-63v rated. It won't give you any benefit performance wise, but such electrolytic capacitors are much easier to find than 1uF ones.
-
this document may be relevant to help you understand if your mains supply is within tolerance.
http://www.saskpower.com/wp-content/uploads/power_quality_efficiency_guide.pdf (http://www.saskpower.com/wp-content/uploads/power_quality_efficiency_guide.pdf)
-
Great info mariush
If I have time tomorrow I'm gonna see what I can come up with now that I have all this info.
I just have to find the appropriate pots in my bins of stuff. I'll mess with different values in simulation first then build it for real.
http://sound.westhost.com/project44.htm (http://sound.westhost.com/project44.htm) Fig.2 shows a way of using larger pots to adjust a 317/337 with transistors so I won't have to use a strange sized pot I don't already own.