Weird PSU problem

[phn]

I have built this PSU: http://dogbreath.de/PS1/LinearPSU_V2/LinearPSU_V2.html

It's the first one, "T1."

I have a 120VAC trafo (way oversized for the purpose) with four secondaries. Two are connected in series, two paralleled. Regardless of what the trafo says, I get about 8 and 16V without load.

One PSU section has 7.6V out (as "T1") and the other 3.6V.

Now the weird stuff begins. When I connect it to the PlayStation the 3.6V drops to 1.5V and the 7.6V becomes anything from 5.5 to 12V! That is, I turn on the switch and the 7.6V becomes, say, 5.5V. I turn it off and then on again and it's 12V!?

I have spent hours trying to figure what's going on, checking the PSU, resoldered and rerouted leads. I have tried the most farfetched (i.e. stupid) things.

The guy designing the PSU uses it with his PlayStation without problem. I have tried mine on two PlayStations with the same result.

I have pretty much given up on the PSU and have decided to use another design. My sentiment is, "screw this." But it's not that easy. Since there's no reasonable explanation for this that I can see, I might end up with the same problem again. So before ordering parts for a new PSU, I thought I at least should ask you guys.

[DragonMaster]

Tried with an other transformer or voltage source?

You said in the PSOne topic that removing the resistors help, so it really looks like the diodes and/or transformers are miswired.
(The - bar is where + is on diodes, I keep forgetting this)

[Mick_F]

Did you EXACTLY follow the schematic shown, or did you deviate in some points? If you deviated, please specify the differences.

Mick

[jamesrnz]

is it a 220 transformer? how does it load with a resistor on the secondaries without the regulator?

with the load what is the input to the regulators, for the lm series you really need 1.5 to 2 volts higher than the expected regulated output.

if the voltage before the regulators drop with the correct load the pri/sec turns ratio is wrong.

you need to verify the transformer voltage and current delivery (AC) first, then if it correct then it is wired wrong the lm's are straight forward.


make a schematic from your mock up to see if you miswired it from mark's

[phn]

If you call raising the voltage from 1.4 to 1.8 by going from a 220Ohm resistor to no resistor improvement then there was.

I even tried using only a voltage regulator, a resistor and a trimpot, same thing. I did use the same trafo and rectifiers. So those could be broken. But if they are they are faking it well.

The design is the same. I use a 10uF poly cap instead of 22uF and have added a 10uF inductor.

I could get a new trafo and replace the rectifiers. But without any indication that they are broken I don't feel like spending €50 only to find out that it makes no difference.

[jamesrnz]

the voltage on the input of the lm's needs to be 1.5 to 2 volts higher than the lm's output under load. in order for them to regulate properly.

the cap change is unimportant.

where did you add the inductor?like i said before..

if loaded and voltage is correct at lm's input you have wiring or component problem after/or including lm.

if the voltage to the lm is less than 1.5 to 2 volts than the loaded output is supposed to be. then you have a problem before the/or including the lm.

you will need to disconnect the lm's and test load the rectifier transformer section.

to get 9 volts rms from the rectifiers at say 750ma you will need a 12 ohm resistor at 7 watts.

to get 5 volts rms at 500ma you will need a 10 ohm resistor also at 5 watts.

-----------in summary, if you cannot supply the regulators with this minimum amount of power the regulators will not power the load.--------- don't even try to get the regulators going until you get the transformer rectifier part working........

you can test the transformer without the rectifiers just to eliminate this..


by the way if you want take a 12 volt battery or 12 v dc power supply and feed the rectifiers across the ac input of the bridge. the rest of should work ok, then yor trans is wrong or wired wrong.

if the rectifiers and regulators do not work you have a component/wiring error.


----there are two ways to check yourself.

---additionally do not assume you wired it correctly, draw yourself a schematic from your work to see how it comes out..........

[DragonMaster]

Are you sure that the transformer rails are wired correctly?

[phn]

Minutes after my post here it hit me. "Idiot!" I didn't mean idiot in a necessarily negative sense.

I thought, "Don't you realize what you just wrote? Are you sure you wired the voltage regulators right?"

I have now tried it. It still doesn't work properly. The 3.6 drops to 2.6 or so. But we can now conclude that I was at error here. So it looks like I will get this to work.

Thanks guys. And sorry for having wasted your time. All just because I took things for granted and didn't double-check. Idiot!

Edit: The inductors are at the outputs.

[DragonMaster]

You don't like the fact that LM317 isn't in the same pin order on a scheme than the TO-220 package?

Unplug the 3.3v rail in case it's messing up the 8.3v one.