Cheap PSU: Piece of junk or semi-decent?

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Err ... again excuse the noob question ... how can I find which resistor works for which rectifier? Maybe beeping my way somehow ?

Exacty. One end of resistor should be on ground, the other on the particular rail.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

Exacty. One end of resistor should be on ground, the other on the particular rail.

Coul you be a bit more specific, as in 'one leg of resistor should beep with the black cable in molex, and the other leg of resistor should beep with the center leg of rectifier', or so ?

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Yeah, that too

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

Ok this is very confusing, I did my best.

7.5Ω:
· Both resistor legs beep between them.
· Both resistor legs beep to black cable in molex.
· Leg 'A' of resistor beeps to center leg of +3.3V rectifier.
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier AND to center leg of +5V rectifier.

1.1Ω:
· Both resistor legs beep between them.
· Both resistor legs beep to black cable in molex.
· Leg 'A' of resistor beeps to center leg of +3.3V rectifier.
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier.

93Ω:
· Resistor legs do NOT beep between them.
· Leg 'A' of resistor does NOT beep to black cable in molex.
· Leg 'A' of resistor beeps to center leg of +12V rectifier.
· Leg 'B' of resistor beeps to black cable in molex.
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier AND to center leg of +5V rectifier.

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Well under 10 ohms it usually beeps. In these low values, go not only for sound, but also for value on DMMs display. There should be 0.0 for it to be direct joint;-) Mine DMM also shows 0 if it is under 1,5 ohm - than check that with resistance meter.

Judging from their position, 93 is on +12 V, the other two on +3,3 V.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

Thanks Behemot, that cleared my rookie brainstorm .

7.5Ω (+3.3V rail)
· Leg 'A' of resistor beeps to center leg of +3.3V rectifier (.000)
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier AND to center leg of +5V rectifier (.015)

1.1Ω (+3.3V rail)
· Leg 'A' of resistor beeps to center leg of +3.3V rectifier (.001)
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier (.000)

93Ω (+12V rail)
· Leg 'A' of resistor beeps to center leg of +12V rectifier (.000)
· Leg 'B' of resistor beeps to center leg of +3.3V rectifier (.007) AND to center leg of +5V rectifier (.015).


Are the multimeter resistance readings valid while the resistors are soldered in the PCB?



BTW always in the pursuit of improving my execrable english, I've discovered that I'm incorrectly titling myself a 'noob'. What I am (or would like to be ) is a 'newbie':

'Contrary to the belief of many, a noob/n00b and a newbie/newb are not the same thing. Newbs are those who are new to some task and are very beginner at it, possibly a little overconfident about it, but they are willing to learn and fix their errors to move out of that stage.

Noobs, on the other hand, know little and have no will to learn any more. They expect people to do the work for them and then expect to get praised about it, and make up a unique species of their own.'

Urban Dictionary - Noob

[momaka]

Re: Cheap PSU: Piece of junk or semi-decent?

May wanna do flying resistor mod too

That will work as well if you don't have any suitable replacement resistors. Personally, I don't like "flying" components, though, because if they are not secured well, they can short out on something. And in my experience, the short circuit protection in these Deer/Allied/L&C isn't very sensitive (actually, not at all).

Argh, there goes my 'productive +12V dummy load'

It is a nice 3.3V/5V load though. I have several Radeon 9700 video cards and they pull similar power from the 3.3V/5V rails. They make the PSU rails very balanced if used in a 12V-based PC.

Are the multimeter resistance readings valid while the resistors are soldered in the PCB?

For the most part, YES. They may be off by a few Ohms, but that's okay. I usually go by the printed color code on them. And to see which rail they are on, I just remove the PSU board from its case and trace where they go to.

For these older Deer/Allied/L&C PSUs, I can tell you for sure that the "93" Ohm resistor is actually a 100 Ohm resistor (color code Brown, Black, Brown, Gold - just Google resistor color codes) and it's for the 12V rail. The "7.5" Ohm is likely for the 3.3V rail. The first band on its color code is blue, though, so it's actually 6.x Ohm resistor. Perhaps 6.8 Ohms? (Blue, Gray, Gold, Gold).
As for the "1.1" Ohm resistor, I'm not sure what that is on. It has too low of a resistance to be on any of the rails. If it was to be on the 3.3V rail, then it would dissipate (3.3 V ^2) / 1.1 Ohms = 9.9 Watts - and that can't be the case because judging by the resistor size, that resistor is probably only a 2 or 3 Watt rated resistor. At 9.9 Watts, it wouldn't last more than a few moments without burning. And if it was on the 5V or 12V rail, forget it - the thing would burn as soon as the power is turned on - so it can't be on those rails either.
So, for the "1.1" Ohm resistor, you would have to remove the PSU PCB from the case and follow the traces. It may be that the "7.5" Ohm resistor is connected in series with the "1.1" Ohm resistor.

If you don't want to do any PCB tracing, another way to measure the resistance of the resistors on those Deer/Allied/L&C PSUs is to connect one of your multimeter probes to a ground on the PSU (a black wire on a drive connector is perfect), and then with the other probe measure the resistance for the 12V rail (yellow wire), 5V rail (red), and 3.3V rail (orange).
Report what resistance measurements you get this way.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

I'm not yet familiar with resistor color codes, but I'll try:

· 93Ω: Dark brown / very dark brown or black / dark brown or dark red / gold

· 7.5Ω: Very dark blue or faded black / light brown or yellowish green / dark brown or dark red / gold

· 1.1Ω: Light red or orange / white / silver / gold

If you don't want to do any PCB tracing ...

Errr not sure now what you mean by tracing. I looked for full continuity (beep + .000 in display) between the three resistors and the three rectifiers, and posted the results above.

Perhaps you mean visual tracing?:

... another way to measure the resistance of the resistors on those Deer/Allied/L&C PSUs is to connect one of your multimeter probes to a ground on the PSU (a black wire on a drive connector is perfect), and then with the other probe measure the resistance for the 12V rail (yellow wire), 5V rail (red), and 3.3V rail (orange).
Report what resistance measurements you get this way.

Aye aye, here we go:

· Black molex ground to yellow molex +12V: 93Ω

· Black molex ground to red molex +5V: 15.6Ω

· Black molex ground to orange mobo connector +3.3V: 7.5Ω

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

That will work as well if you don't have any suitable replacement resistors. Personally, I don't like "flying" components, though, because if they are not secured well, they can short out on something. And in my experience, the short circuit protection in these Deer/Allied/L&C isn't very sensitive (actually, not at all).

I always use heatshrink or hot glue on the conductive parts. Not a suicider ya know On the other hand keeping even the original resistor in the airflow guarantees it will not cook itself anymore.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

Ha, spotted a hidden resistor lurking between the output wires. And guess what, resistance between its legs reads 15.6Ω. And guess what, it zero-beeps to the center leg of the +5V rectifier:

I've perpetrated a couple guetto-mods. A header for the PSU cooling fan:

And a header for the main wires:

Now removing the PCB for fiddling is plug&play, much easier.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

... The one you have is a Deer/Allied/L&C unit ...

You got a hawk's eye Momaka. Found this picture of an Allied PSU from forumer Pentium4:

And mine is almost identical:

Also here is a video of a guy recapping his Allied 350W, again pretty similar to mine.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

... You will need to do a few modifications on it, though. For example, the likely reason why one of its original caps cooked is because of the minimum load resistors. Using this photo as a reference:
http://www2.picturepush.com/photo/a/...g/11286455.jpg
note the 3 large resistors...
- 1 just below the bad Koshin cap you replaced
- 2 to the right of the orange wires
These are the minimum load resistors. More often than not, Deer/Allied likes to use low values for these so they end up generating a lot of heat and cook the caps. If you are going to use this PSU with a computer, then you can change those resistors.
Assuming you'd want to use this with your Pentium 3 PC, you can make the following changes to the load resistors:
3.3V rail: swap for 33 to 100 Ohms resistor (preferably 33 to 47 Ohms), 1 Watt or higher rated resistor
5V rail: swap for 56 to 270 Ohms resistor (100 Ohms works well, but you can go even higher than 270 Ohms if you're going to use this PSU only on 5V-based computers). Again, for power rating, you'll want 1 W or higher.
12V rail: 270 Ohms resistor or higher, 1 W rated ...

So if I get it correctly the problem with these resistors is that they dissipate much heat, and this heat can cook close-by components like capacitors. That's why you recommend replacing them with higher Ω, and hence lower W & lower heat resistors.

But these resistors must be there for some good reason (otherwise they'd discard them to save costs). Maybe they are for better crossloading management? If so perhaps it isn't such a good idea to lower its Ω.

I can think of several ways to tame the heat problem.

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Minimum load- doesn't it say pretty much everything? On these older topologies you had to provide minimum load on every rail for the PSU to run the rails properly. Otherwise strange things could happen (voltage fluctuations, too low/too high voltage etc.). I usually do flying mod - just add some wire and pick them up into the airflow, away from caps where they can be cooled better as well.

[TELVM]

Re: Cheap PSU: Piece of junk or semi-decent?

... strange things could happen (voltage fluctuations, too low/too high voltage etc.) ...

I usually do flying mod - just add some wire and pick them up into the airflow, away from caps where they can be cooled better as well.

I'm thinking on something similar but a bit more radical . The cilindrical body of a resistor isn't conductive, is it?

[momaka]

Re: Cheap PSU: Piece of junk or semi-decent?

I'm not yet familiar with resistor color codes, but I'll try:

Like I mentioned, Google resistor color codes and you will know how to read them. It's not hard and comes in very handy when you start messing with circuits more often.

Errr not sure now what you mean by tracing... Perhaps you mean visual tracing?

The multimeter can give confusing results sometimes, so I rather go by what I see.

Aye aye, here we go:

· Black molex ground to yellow molex +12V: 93Ω

· Black molex ground to red molex +5V: 15.6Ω

· Black molex ground to orange mobo connector +3.3V: 7.5Ω

Boom! perfect. You got a 100 Ohm resistor on the 12V rail, 15 Ohm on the 5V rail, and 6.8 Ohm on the 3.3V rail. Judging by your underside pictures, the "1.1" Ohm resistor along with the 2 diodes next to it appears to be in some kind of load-balancing circuit between the 3.3V rail and the 5V rail. I *think* you can leave that one alone as it probably won't cause harm since it's far away from nearby capacitors.

As for the other resistors, you can do a little swapping around...
- Remove the 100 Ohm resistor from the 12V rail and put it in place of the 15 Ohm resistor. The 15 Ohm resistor you can put in place of the 6.8 Ohm resistor or put the two in series for a total of 21.8 Ohms. The latter will dissipate less heat (good for the caps) but it may also degrade the 3.3V rail regulation slightly.

On the other hand keeping even the original resistor in the airflow guarantees it will not cook itself anymore.

Well, in this case here, you have 3 resistors dissipating about 1.5W each. It may not seem like a lot, but for the size of those resistors it is. Moreover, that's 4.5W of power your wasting for nothing. Even worse, it's 4.5W that heats the other components inside the PSU.

You got a hawk's eye Momaka. Found this picture of an Allied PSU from forumer Pentium4:

I have the same PSU as well (the LC-B300ATX 300W one below).
https://www.badcaps.net/forum/showpo...&postcount=491
You may not recognize it because it has less and crappier components, but the PCB design is exactly the same.
The good thing is, this PSU has extra holes on the PCB for another 5V rectifier (in case 30A is not enough ).

On these older topologies you had to provide minimum load on every rail for the PSU to run the rails properly. Otherwise strange things could happen (voltage fluctuations, too low/too high voltage etc.).

If the power supply is in a computer and you are loading the "right" rail (the one that the PSU was designed to be loaded on the most - for this PSU it's the 5V rail), you can go without minimum load resistors. I have a Task PSU for example, where I changed many of the load resistors. With a heavy 12V load, the PSU whines and oscillates like crazy. With a heavy 5V load, it's quiet and stable.

Also, on my LC-B300ATX, with the exception of the 3.3V rail load resistor, the other load resistors had much higher resistance. IIRC 270 Ohms for the 5V rail and 470 Ohms for the 12V rail.

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

But that's it, once you will use it on 5V based system, than twice on 12V system. I better play it safe, couple watts is not such a big deal

If you manage proper airflow for the resistors, you can disipate much more heat than what they are rated for without them burning down.

TELVM: there is usually non-conductive stuff on them (either some color or ceramic-liek thing). But if you happen to damage it, you may get conductive places on the package and that is not good. Especially the ceramic-like stuff breaks easily.

[momaka]

Re: Cheap PSU: Piece of junk or semi-decent?

But that's it, once you will sue it on 5V based system, than twice on 12V system. I better play it safe, couple watts is not such a big deal

If it's cooking my caps, it is a big deal! The load resistors on these Deer/Allied/L&C PSUs often get hot enough to discolor the PCB over time. I rather play it "unsafe" and remove/replace the resistors than be greeted with freshly baked caps the next time I open the PSU . Most of the power supplies I did this to never complained. Only my TASK PSU oscillated when there wasn't enough of a load on the 5V rail.

If you manage proper airflow for the resistors, you can disipate much more heat than what they are rated for without them burning down.

True. In fact, most resistors can usually withstand 2 to 4x of what they are rated for. That said, I still don't recommend to get anywhere close to what they are rated for in order to keep their temperature down. Usually 1/2 of their rating or less is what I aim for.

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Man, on my PSU loader I am building ATM, I am having some 10W rated resistors with 16W planned load. They have been one third the price of 20W rated

[tom66]

Re: Cheap PSU: Piece of junk or semi-decent?

It is bad engineering practice to use resistors above their rated load.

It is bad engineering practice to use resistors above their rated load.

It is bad engineering practice to use resistors above their rated load.

...and don't think that stuck between those caps and inductors they are going to get any significant cooling. You are kidding yourself.

[Behemot]

Re: Cheap PSU: Piece of junk or semi-decent?

Write that to LC/Deer, I know that?!