#lm338t

It might be the long week, and it may also be that I was mentally prepared to shelve the power supply project, since you know, I thought it was done!  But...

http://www.tymkrs.com/forum/viewtopic.php?id=27&p=2

So a comment from Youtube came:

"Your fuses are on the wrong side of the transformer. You'd want to protect the voltage regulator. What you usually would do is put a 1A fuse (if you want to protect it if it goes higher than 1A) after the capacitor. The current drawn from the 120 AC is much much lower than the current coming out of the voltage regulator."

Roy said: I think you want fuses where they are, but you might want an additional fuse where he says.  Every power supply I have been inside of has had a fuse between the wall and the transformer.  He's right that the current is higher on the output of the regulator than on the input to the transformer... but that's just because you have high voltage lower current on input and lower voltage and more current  on the output side

Brainwagon: Uh, the reason you have a fuse is so that if the transformer shorts, you don't start a fire.

Me: The transformer would limit it to 1.5A right?

Brainwagon: It doesn't limit anything.  That's what it's rated for.

And this throws my head into a tizzy.  I thought that what was coming out of the transformer was 18v 1.5A.  But apparently it's not?  It can draw more?

So I should have worked from this calculation:

18v * 1.5a = 115 @ x     and solving for x = 0.235A 

And gotten a fuse of 0.250A.

Essentially, the fuse that I have now would allow the transformer to draw up to 6.389A which is above its specified rating:

115V @ 1A = 18V @ 6.389A <--above rating of 1.5

Now to address the difference between the voltage regulator and I guess the transformer.

18V @ 1.5A = 28V @ 0.99A       (MAX Voltage Output) 18V @ 1.5A = 1.2V @ 22.5A      (MIN Voltage Output)

But if I have a circuit and it wants 1.2V, it won't necessarily pull 22.5A from this.  So as long as my circuit doesn't require more than 5-8A (max output of the LM338T chip), it should be okay?

But how do I account for that? The fuse protects the transformer from the wall, but doesn't protect the voltage regulator chip.  So what's the calculation to protect the voltage regulator chip?

Questionable equations:

1.2V @ 5A = 18V @ 0.333A

28V @ 5A = 18V @ 7.778A  (which is higher than the rating)

I'm genuinely confused.

So to a certain degree, when the transformer is sitting there without a load, it kind of just sits.  The current that the circuit after the transformer pulls will affect how much current goes from the wall into the transformer and therefore into the circuit.

So the LM338T is rated for...*figures out how to find that value in a datasheet* anddd thanks to @mgburr:

So this is the datasheet for the LM338 and you see a parameter called "Current Limit" where it looks like the typical current draw is 8amps.  But if output is 30v then 1Amp.  12v can be handled at a surge for 500 milliseconds.

Now, the chip will require a minimum current to be able to function as a regulator. Here we see that it's 3.5mA.  The circuit has to draw at least the minimum otherwise the chip won't turn on.

So this is nice and all but now I have some uber awesome newbie questions.

So in "current limit" you have a max of 8-12A which is fine.  But then you also have a limit minimum of 5-7A.  Okay but now there's also a minimum load current parameter.

In "minimum load current" you have the minimum of 3.5mA but a maximum-minimum of 5mA.

What's the point of saying "MIN" 5-7?  We're talking limits here, you have the minimum load current which should be the MIN and the current limit should be the MAX right?  So why isn't the minimum current 3.5mA and the maximum 12A?

@mgburr smartly said: "Input current draw is not the same as output current load."

------Minimum Load Current----[LM338T]----Current Limit------

When you use a regulator, it switches the input at a frequency to make an output based on pulse width modulation.  If you don't have a load on the output, then the PWM will not work correctly and cause it to work improperly and possibly fail.

The minimum load enables the regulator to continue proper operation without upsetting the PWM balance in that it generates normal pulses which prevent the circuitry from going into runaway or current overload.  In our case it'd be the 3.5-5mA or "minimum load current".  This could be considered the input current draw as well.

Now the output current is the maximum current that can be passed.  In our case static 5-8A with pulses of 7-12A.  The pulses have a rise and fall that the circuit has problems keeping up with, which is why the 7 is lower than the 8.  And this could be considered the output current load.

And when you do step up or down, the currents are usually different between input and output.  For example.  With a 9V 3A source voltage, you can provide 27w of power.  If you step the voltage down to 5V, your current will change.

Kinda confusing until you get a Mike to explain it to you!

So @johns_az looked at the schematic I was going to use for the LM338T power regulator and immediately said, "There's no fuse."  And being the awesome guy and amazing schematic drawer he is, he drew this:

Bigger version here: http://www.tymkrs.com/forum/viewtopic.php?id=27

Anyway. So I'm to the point where I'm looking at fuses.  And John suggested I get this one from Radio Shack: http://www.radioshack.com/product/index.jsp?productId=2103770

It's rated for 250VAC 1.0A.  That's cool, but I found that we had some fuses that we got from some Electronic Goldmine random box and they're rated at 250VAC 2.0A.  So I wanted to see if these would work just as well.

Well the main thing then is to figure out whether the transformer can handle 2 amps of current (which to me sounds like a tiny amount).  So I asked around and tried looking for a datasheet which gave me this:

Now this was where I got /really/ confused.  Because you almost never not find a datasheet for some electrical part.  So okay, to the twitter folks!

@atdiy: I'm looking for the max current this transformer can handle.  http://parts.digikey.com/1/parts/999535-transformer-115vac-18vct-1-5a-166k18.html

@eevblog: Trick Q? It's rated at 27VA.

@IdeaPDish: Click on the link in your tweet. Read the box next to max. Lookup 'VA' in Wikipedia.

So I looked up volt-amps and it says: "A volt-ampere (VA) is the unit used for the apparent power in an electrical circuit, equal to the product of root-mean-square (RMS) voltage and RMS current."

Fine no problem, I have (secondary) RMS voltage and RMS current.  Multiply those to get VA:

VA = RMS(volt) * RMS(amp) = 18*1.5 = 27VA

So then. The root mean square amperage IS the max amperage?  And this would mean that the MAXIMUM input current is the output current of the transformer? 

Anyway, this would seem to say that the fuses we have around the house (250vac 2amp) would allow too much current to pass through for our 1.5amp transformer.

So this is one of the other big projects that @johns_az sent me in his box of goodies: A power regulator with the LM338T chip. But I don't normally think much about power supplies and regulators, which is probably dangerous since I'll be working with household AC current (120VDC I think).

So what I think I know:

Power (watts) = Current (amps) x Voltage (volts)

So according to the below picture, output is 24V DC at 100mA.  This would mean that the power output is 2.4watts.

And this Macbook Air power supply is rated for 14.5V at 3.1A --> 45W of power!

Now the interesting thing I've found out is that a lot of power supplies are a conglomeration of different components. 

There is usually a transformer which converts household AC (120 or 240V AC) into something a bit more manageable.  Essentially whatever the rectifier will accept. 

The rectifier takes this AC, and like a diode that only allows current to go in one direction, it changes it to DC current. 

Then a power regulator circuit takes that pulsing current (known as rippling) and essentially smooths it out so that the resulting thing you're supplying power to gets a smooth voltage source.

So you can see from this schematic - we've got the 230VAC from the household current (where you plug your plugs in).  It goes to the coily transformer, then to the boxy rectifier and from there it goes to the regulating circuit which has a lot of capacitors.  I imagine they help to smooth out the voltage as capacitors are known for doing.

Interestingly, some power supplies don't even have regulators in them?!!  According to @whixr, old school tube amplifiers work in this way.  They didn't have silicon based regulator chips, but they did have a number of rectifying tubes which through magic lead to smooth current.  A VERY simplified example:

Here - instead of D1 - you have rectifying tubes.  A whole bunch of them.  So that it's a whole bunch of peaks at different times making a smoother curve like in the three phase six pulse portion below.

Rock on!