Rosewill RP550-2-S

[370forlife]

Re: Rosewill RP550-2-S

Can you post a quick description or picture of the pinout on the transformer? I got that AP-300 here doing nothing and I have no problem getting rid of it.

[everell]

Re: Rosewill RP550-2-S

Here is the pinout pattern of the transformer.

Left side, top to bottom: Ground, 5vsb

Right side, top to bottom: Common, feedback to main pwm chip, feedback power to pwm oscillator or pwm chip, drain of 5vsb transistor or pwm chip, and +320 volts from the mains capacitor.

[everell]

Re: Rosewill RP550-2-S

Trying to fix that 5vsb circuit has gone nowhere. So I finally resorted to temporarily wiring an outboard 5vsb circuit. Powered it up, and all voltages present and measure good......but the psu is on even when it should be off. Started measuring voltages on the turn-on transistor switch and BANG. One of the switching transistors shorted and arced to one of the mains capacitors. Looks almost like arc welding. Damage to the capacitor case makes me suspicious that it is a goner. Also blew the fuse.

The outboarded 5vsb still works OK. But now will have to replace a switching transistor, a burned out resistor in its circuit, and maybe the pwm chip.

Now for the problem......the numbers on the transistors and chips have been removed (ground off). Does anyone have one of these with readable numbers? I will need to know switching transistors, pwm chip, and possibly the supervisory chip.

[SIDMX]

Re: Rosewill RP550-2-S

I think i have one of this PSUs in my wife's computer, im like 99% sure about the model but it's a 550W from Rosewill for sure because i picked all the hardware and that model number in the stick sounded too familiar to me.

Right now i'm not in home, but in a few minutes i will and then i could tell if it's or not.

[SIDMX]

Re: Rosewill RP550-2-S

Very close but it's not the same model this one is labeled RP550V2-S-SL the guts looks very close too, main transformer is labeled diferent but the 5vsb transformer match perfectly so it may use the same 5vsb design ...what do you think?

[everell]

Re: Rosewill RP550-2-S

They look very much the same. Parts very crowded. My psu has a blown (and gutted) 5vsb, a blown main switcher, and a blown main capacitor. I think I will put it aside til much later. For now, check the spacing between the main switching transistor and the mains capacitors. Might be a good idea to insert a piece of plastic for insulation if they are very close or touching. Here are the pictures from damage to my psu. The transistor is black and damaged....nothing unusual. The other picture shows damage to the main capacitor under the microscope. The arc damage blew a hole completely thru the metal can and got inside the capacitor. This leads me to think that there may have been more than one arc, possibly why I got the bad psu in the first place.

[SIDMX]

Re: Rosewill RP550-2-S

Might be a good idea to insert a piece of plastic for insulation if they are very close or touching.

Thanks for the tip i will do it this weekend.

[everell]

Re: Rosewill RP550-2-S

Success at last! The DM311 mod did NOT work on this power supply. I thought at first that the 5vsb transformer was bad. Finally I set up a breadboard circuit with only the key 5vsb components. What I found was that the feedback winding providing the Vdd voltage to the DM311 was going above 20 volts, which was shutting down the DM311 operation. So the overvoltage protection was working just fine! So next I tried wiring a TNY277 chip into the circuit. Still had problems until I determined that the optocoupler was leaky. Replaced the optocoupler, and now the 5vsb circuit works. Now the Vdd feedback voltage to the TNY277 provided by the transformer feedback winding is measuring 25 to 35 volts, depending on amount of loading.

Now for a new problem. The 5vsb transformer feedback winding which provides power to the pwm chip is putting out about 26 volts when the 5vsb has a 2.5 ohm load (for a 2 amp output). To close for comfort on the UC3842/45 series of pwm chips. So I will need to insert a voltage regulator between the DC voltage source and the pwm chip. I found that the Bestec ATX-250 12E and 12Z both have a great regulator circuit for this purpose. Bestec scores again!

[370forlife]

Re: Rosewill RP550-2-S

I like how bestec has such a horrible reputation yet are actually quite well built.

[c_hegge]

Re: Rosewill RP550-2-S

^
Hipro are better again. I compared a bestec ATX-300-12Z wih a Hipro HP-D3057F3H, since HP seem to use them interchangably. The bestec had higher ripple, and exploded at 400W. the Hipro did 450W and shut down when I asked for more.

The irony is though, that the Bestec uses capxon caps, runs somewhat hotter, has higher ripple, but the caps last much longer than the Teapos and LTECs in the Hipro, which stresses them less (and uses bigger caps). It leads me to believe one thing - teapo

[everell]

Re: Rosewill RP550-2-S

How many Bestec ATX-300 12Z did you test, and how many HiPro did you test? 300 watt Bestec blowing up at 400 watts sounds much better than the Deer/LC/Solytec lines, such as the famous PowMax.

Some day I would like to hear about someone testing the Bestec ATX-250 12E which has the DM311 mod. I think that would be much more of a "fair" fight since we already know that most of its problems were caused by bad 5vsb design. Fix the 5vsb and then see what it can do!

[momaka]

Re: Rosewill RP550-2-S

^
Hipro are better again.

+1
Most HiPro PSUs I've seen have space for 12.5 mm caps on most of the rails, which is very nice IMO. The heatsinks and also always much thicker than other PSUs of the same wattage rating. And I like how there's no SMD components on the solder side of the HiPro PSU either (at least not in the older Dell 200/250W PSUs) - makes the solder side look cleaner.

[370forlife]

Re: Rosewill RP550-2-S

How many Bestec ATX-300 12Z did you test, and how many HiPro did you test? 300 watt Bestec blowing up at 400 watts sounds much better than the Deer/LC/Solytec lines, such as the famous PowMax.

Some day I would like to hear about someone testing the Bestec ATX-250 12E which has the DM311 mod. I think that would be much more of a "fair" fight since we already know that most of its problems were caused by bad 5vsb design. Fix the 5vsb and then see what it can do!

Oklahoma Wolf tested a 250-12E on jonnyguru.com, and he couldn't kill it, even with bad capacitors and a failed 5vsb.

http://www.jonnyguru.com/modules.php...Story&reid=154

[c_hegge]

Re: Rosewill RP550-2-S

I won't deny that the 300-12Z is a well designed power supply, and it will do the promised 300W in spec which is great, but when a comparable Hipro does 450W and shuts down when I ask for more, as well as having better ripple suppression, while the bestec blows up at 400W, what am I supposed to believe?

[2lazy]

Re: Rosewill RP550-2-S

Does this psu have a resistor on the bottom of the pcb, near mid center? If so any way I can get the value of it off you? Plz Plz Plz...

[everell]

Re: Rosewill RP550-2-S

Does this psu have a resistor on the bottom of the pcb, near mid center? If so any way I can get the value of it off you? Plz Plz Plz...

If you are talking about the Rosewill RP550-2-S

It does not have any resistor on the bottom of the pc board......there are NO SMT components on the bottom of the pc board. All components are on the top side.

Tell us the model number of your power supply, and post some pictures.

[Shocker]

Re: Rosewill RP550-2-S

The irony is though, that the Bestec uses capxon caps, runs somewhat hotter, has higher ripple, but the caps last much longer than the Teapos and LTECs in the Hipro, which stresses them less (and uses bigger caps). It leads me to believe one thing - teapo

Not on the +5VSB IIRC. The Bestec has two 2200uF 10mm caps there, while the Hipro only has a 1000uF and a 470uF (both 8mm). Perhaps the 1000uF in the Hipro can't take the ripple???

FYI, two of the 16V caps are for the +12V and the other two for the -12V.

But, yes, it is ironic that the Bestec's CapXons last longer considering the working conditions.

After the Bestec blew out, what did you do with it???

BTW, I don't burn whole PSUs no matter how bad they are. I try to salvage working components such as:

X capacitors
Y capacitors
EMI filter coils
Rectifier bridges
Primary switchers
Secondary rectifiers

Funny thing is that I've never seen a bloated Fuhjyyu LP in person. But I do have a Bliss "250W" (Macron OEM) with bulging "JP" (not JPCON nor CEC, even though both those brands exist in the same PSU - who knows who they are) 330uF 200V 18mm primaries , and I've also seen a King World PSU with a leaking from the bottom SWCON primary cap (also 330uF 200V 18mm), which I promptly binned... (all the other caps were SWCON too)

SWCON
SWCON
SWCON
SWCON
SWCON

As a matter of fact, I got two 0.1uF X2 caps, an input coil, an F12C20C, an S30D40C, and a 40N03P (IIRC) out of there, and I've still left the switchers in place. (The 40N03P is a MOSFET used to regulate the +3.3V in some old and/or cheap designs.)

You're missing out on a lot of stuff if you burn up PSUs without salvaging anything. On my desk I have 11 X2 caps, 20 Y2 caps, 6 EE-25 coupled coils, 4 other coupled coils, 4 bridges, 8 fast rectifiers, and I can't even count the number of Schottky rectifiers. (Better clear it off soon...)

Are you going to follow me and put a on my face??? Or are you going to stay with your old ways and leave a on my face???

Also, my system is currently running on an HP-D3057F3H with original caps .

Have I missed anything???

[momaka]

Re: Rosewill RP550-2-S

You're missing out on a lot of stuff if you burn up PSUs without salvaging anything.

+1
I even keep my crappy PSU's intact. Sometimes I need to set up a quick experiment for a circuit, and these PSUs do okay for that.
Old broken CRT TVs are also a good source for parts. Got 2x 200v, 330uF Nichicon caps from one, as well as a coil, 3 X caps, a 4W stereo amp IC, and many other goodies. I'm thinking of building a small amp for my headphones with that stereo amp IC .

[c_hegge]

Re: Rosewill RP550-2-S

Not on the +5VSB IIRC. The Bestec has two 2200uF 10mm caps there, while the Hipro only has a 1000uF and a 470uF (both 8mm). Perhaps the 1000uF in the Hipro can't take the ripple???

There's not a lot of ripple. With a good cap, it's under 10mV with 1A load. good caps like Chemi-con KY and KZE seem to hold up much better, but it's not just the 5vsb caps which fail. I've seen the 2200uFs die as well.

After the Bestec blew out, what did you do with it???

IIRC, i just scavenged the good parts, since I have lots of 300 watters as it is.

BTW, I don't burn whole PSUs no matter how bad they are. I try to salvage working components such as:

I don't often burn the whole PSU if there are parts on it that are worth scavenging. You do find the odd one with no EMI filtering, 2A four diode treatment and only 10A rectifiers on all of the rails. There's nothing woirth scavenging on those, and those do just get burned or thrown in the garbage as they are.

[Shocker]

Re: Rosewill RP550-2-S

I don't often burn the whole PSU if there are parts on it that are worth scavenging. You do find the odd one with no EMI filtering, 2A four diode treatment and only 10A rectifiers on all of the rails. There's nothing woirth scavenging on those, and those do just get burned or thrown in the garbage as they are.

I scavenge the rectifiers regardless of rating. I often find F06C20C fast rectifiers on the +5VSB in Casing Macron PSUs (mostly MPT-301s). I replace them with 10A Schottky rectifiers from the +3.3V and +5V of L&Cs. (In case you're wondering, I do keep the F06C20Cs.) It's not just about efficiency. Lower forward voltage = less heat = more reliable, and that matters with the fan off.

I did include 4 F06C20Cs when I counted 8 fast rectifiers, so it might not be a useful count, but I also included 3 F12C20Cs (better than nothing for old +5V heavy PSUs) and a "C25M" that I can't find any info on. There were also an FR1004G, an SF1002G, and two SF1004Gs that I didn't count at the time.

In any event, all the smoke from the burning can't be too good for you.