Power supply HELP PLEASE!

[slotbot]

Hi Everyone!

I am trying to make myself a nice little powerbrick since almost all of my Boss Adaptors have called it quits.

unfortuantely i am having MASSIVE problems. And im not quite sure why. So i thought maybe someone could help me out.

First of all i think my transformer may not be OK. Its supposed to have 117 V in and 15 V out. (im assuming they mean RMS)

I get117 V in and like 8 V outs (On my dmm) if i jsut let one lead float. but if i hook up both leads of the DMM to both outs of the transformer i get zero volts output.  

Maybe i screwed up my tranny by measuring it this way as well?

The other thing i want to mention is that im only using 2 wires from the socket. not all 3 (no ground) becasue my transformer is not center tapped on either side (as far as i can tell).

Most schematics ive seen in a textbook i have show the third prong going to ground. Is this required?

Basically here is the power supply i came up with...

the above mentioned transformer => a bridge rectifier made with 1n4003's => 3300uF 25 V cap => LM317 variable voltage regulator set up to output 9.6 V.

I have tested the 317 setup on a breadboard and it works ok if you feed it a DC voltage out of my bench supply.

So i believe the problem is in getting the 117 line voltage to a DC voltage.

also i have a side question about bridge rectifiers.

im trying to make one on my breadboard and then use my signal generator to test it. the output always only looks liek a halfwave rectifier. the second half of the wave is not getting rectified. The way i have it set up is 4 1n4003's  and then a resistor inbetween where the ground connection would be and the output.

then i put 1 side of the function generator output where each side of the incoming AC signal would go.

And then my scome on the output and ground.

Should this work? im not sure if its suppsed to so  im kinda stuck there aswell

IF ANYONE can give me any help here it would be MUCH apprciated.

THANKS!

[spongebob]

I get117 V in and like 8 V outs (On my dmm) if i jsut let one lead float. but if i hook up both leads of the DMM to both outs of the transformer i get zero volts output.

What lead do you leave floating? Your transformer has a single secondary pair I assume? Then you should have to connect the DMM to both secondary wires to get a reading, where would you connect the DMM ground otherwise?

The other thing i want to mention is that im only using 2 wires from the socket. not all 3 (no ground) becasue my transformer is not center tapped on either side (as far as i can tell).

The 3rd wire should be called Earth, not Ground, though it's often conneced to power ground. But the power supply give the proper readings even without a Earth connection, assuming you wired everything else correctly.

Nice page on PSUs: http://sound.westhost.com/psu-wiring.htm

[niftydog]

I get117 V in and like 8 V outs (On my dmm) if i jsut let one lead float. but if i hook up both leads of the DMM to both outs of the transformer i get zero volts output.

To get proper readings on a DMM, both leads must be connected to something. You must make a circuit!

Is your DMM set to read AC voltages?

Maybe i screwed up my tranny by measuring it this way as well?

highly unlikely.

im only using 2 wires from the socket. not all 3 (no ground) becasue my transformer is not center tapped on either side

Center tapped transformers have NOTHING to do with the earth wire from your mains power outlet. You are begining to concern me with your lack of knowledge on the subject, are you certain that you are competent to be working with mains power?!?! This is not the project to be experimenting on, mains power can be lethal!!!

The earth lead is included for safety and is often not connected to the circuitry at all. It is most often connected to the metal casing so that any dangerous short circuits to the case are shunted to ground, thus tripping the circuit breaker.

the output always only looks liek a halfwave rectifier. the second half of the wave is not getting rectified.

far more foolproof to get yourself a purpose built bridge rectifier package. Sounds to me like you have the diodes wired incorrectly.

then i put 1 side of the function generator output where each side of the incoming AC signal would go.

And then my scome on the output and ground.

Should this work? im not sure if its suppsed to so im kinda stuck there aswell

the phrase "scrabbling for purchase" springs to mind. I strongly encourage you to revisit the theory of power supplies rather than stabbing blindly in the dark with potentially lethal voltages! You are doing some things right, like testing the bridge rectifier with a relatively safe voltage level, but if you can't get that right, you certainly should not be going anywhere near mains voltages.

by the way, if you have a scope, that will give you far more accurate readings than your DMM... and you won't be fooled by trying to read an AC voltage on a DC voltage range and getting zero.

[slotbot]

Ok i solved most of the problems.

The transformer was no good. i got a replacement today and that took care of most of the problem.

As for the bridge rectifier not working on the bread board i MAY have found an explantion. And that is that the output of the function generator is earth grounded. its not really an ac supply. So it was liek having a ground where thereshould have been an ac signal?

im not sure if thats what could have been doing it. But in any case the bridge rectifier works fine on my perf board supply with the transformer.

the only thing left is that i dont know what the VA rating of my transformer is. The 317 can handle 1.5A but i doubt the transformer can. its about 2 inches cubed. im guessing its around 10 VA so that gives me around 666 mA to go with? (cause its a 15 v output. sooo 10VA = 15V x .666A)

and if i were to fuse that on the side plugging into the wall of the transformer i would want a fuse around .1A (.085 to be more exact). (cause 10VA = 117V x .085 A)

Is this the correct way to come up with what fuse to use?

Also if anyone elsse want to take an educated guess on the VA rating of my transformer that woudl be appreciated. the coil windings are abotthe size of a golf ball. and then withthe metal brace around it its about 2 by 2 and abit (inches) with a height of just under 2 inches.

[slotbot]

also niftydog just FYI i was aware of all these things. the reason i posted it so bluntly is to get more liek a verification that YES it SHOULD be working. But thanks for the concern. it is nice to see that people are looking out for each others safety!

The most confusing thing was the bridge rectifier not even working on the breadboard. i was dumbfounded. but further analizing today i realized it was probably becasue teh function generator i have has a ground on one side of the lead. so its not really like an ac source in that aspect. and it was putting a ground where teh - ac line would be. thus effectively making a half wave rectivier. (which is the waveform i was getting)

Thanks

scott.

[niftydog]

the output of the function generator is earth grounded. its not really an ac supply.

not sure what you mean by "threshold" but an AC signal passes "through ground". That is to say that it's positive and negative. So, if your generator is fluctuating positive only, it is said to have a DC shift. This can be eliminated using a series capacitor.

it is nice to see that people are looking out for each others safety!

Cool, it's just my obligatory warning/concern comment I put in most threads about mains voltage.

teh function generator i have has a ground on one side of the lead. so its not really like an ac source in that aspect.

Well... the AC signal could still go below ground if it wanted to! Suss it out on the scope. Find the ground point then plug in the signal.

And, as for the fuse, I'd suggest that it wouldn't need to be too close to the "maximum" current calculated. The LM317 has good protection built in and that will save the day most of the time. I would suggest that the fuse is there only in the case of a short circuit occuring before the LM317. In this case, the current coming in from the AC mains will be enormous and the breaker should trip. Still, it's good practice to have a fuse there, maybe make it 1.5 times the maximum current.

Oops, almost forgot, here's a great VA rating resource.

[slotbot]

Hi,

I thought i had this supply working but now i have a new problem (YAY!)

The problem is that i can only draw about 100 mA before the voltage totaly sags to about 3.5 volts. THe drop isnt instant. itll be like 9.6 for a few seconds then slowly start dropping to 9.5 then start dropping quicker until it hits about  3.5.

The resitors im using in the 317 set up are 220 ohms and about 1.5K (trim pot adjsuted to give 9.6v out.)

When i remove the load the voltage goes back up to around 9.4V.

Later i monitored it for along period with a low current draw and noticed that it was VERY slowly dropping in output voltage aswell. I left it overnight and it went from 9.6 to about 9.4 with a low current draw (bazz fuzz with LED on).

Any ideas?

Scott.

[Eric H]

The problem is that i can only draw about 100 mA before the voltage totaly sags to about 3.5 volts. THe drop isnt instant. itll be like 9.6 for a few seconds then slowly start dropping to 9.5 then start dropping quicker until it hits about  3.5.



Any ideas?

Scott.

What is the AC voltage to the regulator? Sounds like it's too low.

-Eric

[beans_amps]

Is the IC regulator a LM317 or LM317T?  The Plain version is rated for 300mA while the T version is rated for 1.5 Amps.

I your incoming DC voltage high enough?  It needs to be about 2.5 volts above the output.  Check both sides of the regulator under load.  If the incoming side is sagging, either a sereis resistor has too large a value or the PT is too small.  IF it only sags on the output side; change the LM317 with a another and see if that helps.  

Also measure your 220 Ohm sense resistor and ensure it is the right value.  A 2200 ohm resitor will limit your output current by a factor of ten.

Sean Weatherford
Bean's Amp Repair
Central, SC

[Eric H]

I your incoming DC voltage high enough?  It needs to be about 2.5 volts above the output.  Check both sides of the regulator under load.  If the incoming side is sagging, either a sereis resistor has too large a value or the PT is too small.  

Exactly.

2.5 volts is the minimum, though, and the incoming voltage will drop significantly with load.

I meant DC in my post ;-)

-Eric

[niftydog]

you need to work out whether it's the input or the output that's sagging. Measure both sides of the regulator while under load.

Someone mentioned a series resistor, do you have a series resistor in circuit?

[slotbot]

Ok. I did some scope measurements.

for the test i used a 2 watt 47 ohm resistor as the load.

So that should draw 204 ma (assuming 9.6 is my set Vout).

When the load is applied the voltage at the input sags for a second or 2 but then goes back up to around 18V with a nominal amount of ripple (1 volt or so)

My cap is a 3300uF rated for 25 V. Also the 317 is heating up like KERAZY even with the heatsink i have on it it is way to hot to touch (while the voltage is sagging). i dont know if thats normal or not. i know voltage regulateors get hot.

int terms of a series resistor hmm. heres my design.

transformer => bridge rectifier made of 4003's => 3300uF cap (polarity is correct i have checked) => lm 317 Input => a 220 on Vout to ADJ and  a 100k trim in llel with a 2.2k resistor with the trim adjusted to give a total resistance of about 1.5k from ADJ to ground => the vout goes straight to my dc out jacks.

i was playing around withit today and i think the regulator blew. i am going to get more parts in the next couple of days and try again.

I was thinking of jsut getting a 10V regulator though instead. even though they handle less current, i find 1A to be more then acceptable.

But of course id REALLY lke to get this going with the "correct" voltage. becase it *should* work right?

Also here are soem otehr measurments i took today before teh regulator went down.

There is almost no current coming out of the regulator when the voltage is "sagging" (i think imeasured around 6mA.)

Vin unloaded is 22 V
Vin loaded is around 18 V with a small ripple.

I DO NOT have any of the input output caps on the LM317 because from what i have read on the spec sheet it doesnt sound like i need them.

Thank you for all your help so far. hopefully i can get this thing going.

scott.

[spongebob]

Does the regulator get hot without a load? I had this problem once, adding input and output capacitors to the LM317 pins fixed it, if you look in the datasheet it recommends these capacitors to avoid stability problems.

Here's the configuration I normally use with no problems (straight from the datasheet):


But the regulator will get warm during normal operation too, total power consumption will still be 18V * 0.2A = 3.6W, but if the transformer's output voltage stays up under load I would search for problems at the regulator.

[Torchy]

crp

[niftydog]

the input sags for a second or 2 but then goes back up to around 18V with a nominal amount of ripple (1 volt or so)

Could be too much input voltage. At 200mA draw, the regulator is dissipating 1.7W of energy just to regulate the 18VDC down to 9.6VDC.

Let alone what it dissipates to supply the current at 9.6VDC.

How do the waveforms look? Is the output stable and free from high frequency guff?

Try adding the caps like spongebob said.

Also, physically measure the resistance from the output to ground. It may be that there's something shorting. This might account for why you only measured 6mA... the REAL current is taking a different path. ie; not through your multimeter.

[slotbot]

OK. Thank you all for your help!

After quite a bit of tinkering i *think* i have it.

i added 1 100 uF accross the output and the input and output ones suggested. I also pt in a brand new 317 with a nice little heatsink. AND i changed the trim pot setup to a 1k trim in series with a 1k resistor.

i misplaced my 2w resistor so i cant do that test on it right now but before i relized the craping out when i tried to power 2 digitech x series and a danelectro vibrato and a small line mixer i made. (caused INSTANT CRAP OUT) now however im sitting here with 4 digitech x series pedals (2 delays bass synth & metal master) AND another large pedal i have made (pulls about 150 mA on its own) and the voltage is sitting steady at 9.65. jiggles up and down a bit every so often but there is definately no tendancy either way.

really i dont thinkill ever exceed 1A i dont EVER use that many pedals at once. So im goingto grab a high power 10 ohm resistor next time m in the local shop and see if this thing can handle it.

so ill keep my fingers crossed for now and post if anythign  goes down.

Thanks!

Scott.

[niftydog]

sweet! So, can't say for sure but it could have been oscillating, the caps would have stopped that.

the 10 ohm resistor... sure the regulator will handle it with a decent heatsink, but the resistor had better be something special! 1A into 10ohms is 10Watts!