antec sp 450

[retiredcaps]

Re: antec sp 450

Both of those caps are for the 5vsb. Get them replaced and move forward.

Okay, I used 2 of my 6 remaining 1000uF 16V Rubycon MCZ 8mm to replace a Asia'X EC bulged 2200uF 10V (EC9) and Teapo SC 1000uF 10V (EC10). The Asia'X measured 10.3uF and the Teapo measured 1036uF on my multimeter. I don't have an ESR meter so I don't know ESR readings.

I didn't have a 2200uF 10V in 8mm and the other thread suggested that a 1000uF was fine for EC9.

The standby (purple) reads a solid 5.00V DC with the Rubycon MCZ (it was 4.98V DC with the previous Asia'X and Teapo combo). I tried shorting green and black and I still get no fan startup and all the other DC voltages are 0.00.

I know in the LCD SMPS world, there is a startup cap usually around 22uF 50V that plays an important part in the SMPS startup process and I see one at EC5. I don't have a 22uF 50V. I have a 47uF 50V and I might have a 22uF 25V. Would either suffice for a test? Or am I looking in the wrong place?

[Toasty]

Re: antec sp 450

47@50 should be fine for that 22@50. That's the one out near the mains plug and fuse for the board, yes?

I have no Cap ID's on my sheets. Just values and placement.

Difference between 4.98v and 5v is not a difference. The leads had better contact this time than before, or they weren't laying so near the switcher before, or.....

If it was 4.8 and 5v, then I'd say yes, perhaps.

If you replaced the 2 caps in between the heatsinks at the edge, then the 5vsb is fine. I did replace the other 5 caps in between those heatsinks.

The next batch would be the output caps on the other side of the fan control board. All the little 1uF and such, can be left alone.

>>I didn't have a 2200uF 10V in 8mm and the other thread suggested that a 1000uF was fine for EC9.<<
If that means you replaced the output 2200/10 with a 1000, you need to rethink that. The filter is set up to use a 1000 on input and 2200 on output. Radical changes can throw the voltages off and introduce a lot of ripple.

If you can do a cap map (drawn) and take a pic, then I can follow your EC#'s better.

Toast

[retiredcaps]

Re: antec sp 450

47@50 should be fine for that 22@50. That's the one out near the mains plug and fuse for the board, yes?

Yes, it is next to the mains plug and fuse.

If you replaced the 2 caps in between the heatsinks at the edge, then the 5vsb is fine. I did replace the other 5 caps in between those heatsinks.

Yes, I replaced the 2 caps in between the heatsink near the edge with the Rubycon MCZs.

The next batch would be the output caps on the other side of the fan control board. All the little 1uF and such, can be left alone.

I don't have any of those values 3300uF, 2200uF, etc at home so I will have to wait to aggregate a big enough order to justify the $8 USD shipping costs to Canada. I do have ten 1uF 50V Rubycon caps though.

At some point in this exercise, I may have to add up all the caps (+ shipping) required for a total recap vs the cost of a new quality name 400W PS (80 bronze). A good name brand on sale will be around $35-40 CDN.

Right now, I'm willing to spend some money to learn something new to me.

>>I didn't have a 2200uF 10V in 8mm and the other thread suggested that a 1000uF was fine for EC9.<<
If that means you replaced the output 2200/10 with a 1000, you need to rethink that. The filter is set up to use a 1000 on input and 2200 on output. Radical changes can throw the voltages off and introduce a lot of ripple.

I thought the drop from 2200uF to 1000uF was odd too, but the other thread says that two 1000uF in the 5V SB voltage rail was done and tested many times to be working. I don't have a scope and don't plan to acquire anytime soon. Once I place an order, I will get the 2200uF 8mm or I can try and shoehorn in the 2200uF 10mm cap I have. I'll have to check the spacing later.

If you can do a cap map (drawn) and take a pic, then I can follow your EC#'s better.

Toast

I will try to get it done in the next day or two.

[c_hegge]

Re: antec sp 450

When your done add a 2-2.4K resistor in parallel with the thermistor on the secondary, that way you know it won't cook your new caps.

[retiredcaps]

Re: antec sp 450

Just replaced the 22uF 50V cap by the mains plug/fuse with a Rubycon 47uF 50V and still no power. 5V SB is still 5.00V DC, but other voltages are 0.00V DC when shorting green to black.

Also, there is no way I can shoehorn in a 2200uF 10mm in the 5V SB rail. It is already a very tight fit with 8mm caps. I'll have to place it on the future order list.

Hmm, I'm going to have to put this one on the "todo pile". I don't have any of the larger uF caps at home and I have company coming over the next week.

Even though these are crappy caps, I'm starting to suspect some transistor or IC is shorted. I will try to check for shorted components , but my spatial mapping skills aren't the greatest.

I'm not giving up, but some time away might help.

edit: The original 22uF cap measures 20.1uF on my multimeter. Here is a pic of the two 8mm Rubycon's on the 5V SB rail. As you can see, there is no room for a 10mm cap.

[Toasty]

Re: antec sp 450

>>Also, there is no way I can shoehorn in a 2200uF 10mm in the 5V SB rail. <<



I think this got derailed somehow, somewhere. I don't know how to say it any clearer.

The only 2200/10 I have on my map is by the OUTPUTS between the RED and YELLOW wires.

If yours had a 2200/10 on the 5vsb, that is different than mine had.
Mine had (2) 1000/10's

Yes, we will have to track the fault down. As I no longer have this unit in front of me, I can't offer "look for this voltage there" exacting kind of help. Any testing will be interesting and will rely on photos and your skills with a meter.

Enjoy your company! Look forward to picking this up later.

Toast

[retiredcaps]

Re: antec sp 450

I think this got derailed somehow, somewhere. I don't know how to say it any clearer.

The only 2200/10 I have on my map is by the OUTPUTS between the RED and YELLOW wires.

If yours had a 2200/10 on the 5vsb, that is different than mine had.
Mine had (2) 1000/10's

When I got the unit, the Antec warranty sticker was still intact so I'm pretty sure I'm the first to open it. The bloated cap was a 2200uF 10V cap. I'm holding it in front of me now. Clearly marked 2200uF. It was where the "inner" Rubycon sits now. The Teapo 1000uF 10V was sitting where the "outer" Rubycon sits now.

Maybe I have a different revision or something?

[retiredcaps]

Re: antec sp 450

Any testing will be interesting and will rely on photos and your skills with a meter.

Tracking down voltages using a multimeter is not a problem. I do this in the LCD SMPS world and know generally what to look for, but in the ATX SMPS world, I find the confined space and the soldered mains wiring hard to deal with SAFELY. The typical LCD SMPS are laid out in a more flat space and I have access to the components safely with my multimeter probes.

I can't figure out a way to measure the back of the ATX PS PCB without possibly shorting out something. If you or others have suggestions on how to do this safely, let me know. Pictures of how you test or "bread board" would help a lot.

[Toasty]

Re: antec sp 450

I have some very fine, sharp probes that I use when the traces get narrow and close. I also have a hand-held, all-in-one digital meter with a pigtail lead for work around IC's. Again, very fine sharp point.

I use clip on meter leads to latch on to component legs, especially when electrically HOT.

I just picked up a one piece, tweezer-style, meter lead set off Amazon that is nice for measuring SMD caps and resistors. Squeeze, read meter, next!

If push comes to shove, then I tack-solder some wire and use some alligator clips out to the meter(s). Nothing wilder that 4 or 5 meters and a scope all tied to one piece of equipment, watching it run its paces. Post-its anyone?

Toast

[Consider investing in a multi-voltage isolation transformer.]

[retiredcaps]

Re: antec sp 450

Well, I started with power off and unplugged probing around for shorts. I couldn't find any so I balanced the power supply with plastic shielding and cardboard so I could probe the backside. I started with the secondary side and couldn't find any voltages. So I moved to the hotside and found my 162V DC (steady) across the main filter cap.

Then I moved back over secondary side to the ICs and found 4.74 and 4.54V DC. I moved onto the next IC and found 5.00V and 11.99V DC. At this point, it did not dawn upon me that the PS might be working. I was confused at seeing 11.99V and did a bit more probing for voltages. When I looked up I saw the PS internal fan spin. The fan is super quiet and I didn't hear it. I then probed the secondary voltages and to my utter surprise, I found 3.34, 5.02, and 12.22V DC.

I didn't change a single thing so I suspect I might have found a cold joint by probing around. I hooked up the PS to an old 2001 motherboard and it works fine. The secondary voltages are all stable.

I'll test it for a few more hours and then I'll order proper replacement caps for all the Fuhjyyu still inside.

The only wierd thing is that with my Amprobe AM-60 multimeter, when I measure the 12V rail, I get 0L followed by a beep for HV alert, a reading of 451V DC, and then it settles on 12V. I turned AM-60 on/off and got the same readings. I then pulled out the Fluke 75 and Fluke 12 and got 0L followed by 12V DC.

Curious, I pulled out another PS and measured the 12V rail. For all 3 multimeters I got 0L and then 12V DC. I'm wondering if I found a glitch in the AM-60? I did purchase it brand new in the original packaging (from an USA ebay seller) and am confident that is not damaged in anyway. Other than this 451V voltage alert, it is works great so far (it has the sharpest probes out of all my meters).

[b700029]

Re: antec sp 450

Are you in a dry area? Static electricity accumulating on the probes?

451V spikes on the 12V rail do not sound normal at all, it might be a good idea to check with an oscilloscope (if you have one).

[retiredcaps]

Re: antec sp 450

Are you in a dry area?

Yes. It is quite dry where I live.

Static electricity accumulating on the probes?

Not sure. See below.

451V spikes on the 12V rail do not sound normal at all, it might be a good idea to check with an oscilloscope (if you have one).

I should point out that the 451V spike reading is recorded this way.

1) Antec power switch is turned off on the back.
2) Plug dmm black probe into black molex lead.
3) Plug dmm red probe into yellow molex lead.
4) Turn on power switch on back of Antec.
5) Press computer power button at front.
6) See and hear Amprobe AM-60 go OL, then 451, then 12.2V DC.
7) Fluke 12/75 goes OL, then 12.2V DC.

If the power supply is already on and I plug dmm black probe into black molex and dmm red probe into yellow molex, I get 12.2V DC right away on the AM-60 (no OL or overload indication).

Now I do still have those crappy Fuhjyy caps on the secondary side. Once I order caps and replace them, I will retest.

And I still could have some dodgy cold solder joints so once I replace the caps, I will touch up all the joints.

[retiredcaps]

Re: antec sp 450

it might be a good idea to check with an oscilloscope (if you have one).

I was "this close" to getting a MSRP $1,900 used handheld dmm/scope for 94% off, but someone beat me to it. So no, I don't have a scope.

I should also note that this behaviour does not happen on the 5V rail.

[Toasty]

Re: antec sp 450

I don't want to say it's not -possible-, but it's not possible to get 451v on the secondary side. It has to be something with the meter or the leads. The number is -TOO- consistent.

You may be seeing some induced interference or high-frequency spikes in the test leads.

[Krankshaft]

Re: antec sp 450

When the two mains filter caps charge up after being dead on my Antec Smartpower (after flipping the rear power switch) the incandescent lights on the same circuit dim slightly for a split second as the caps charge.

So I'm sure this supply has a fair bit of inrush current. Could you somehow be measuring an inrush spike?

That would line up with why it only happens when the main power switch is turned off since it would allow the main filters to discharge.

[retiredcaps]

Re: antec sp 450

It has to be something with the meter or the leads.

Okay, mystery solved.

In auto ranging mode, both the Amprobe AM-60 and Fluke 12 show overload before displaying the proper DC voltage. If I switch to manual range mode, both the Amprobe AM-60 and Fluke 12 show no overload (0L) upon Antec PS startup.

I was also curious about the AC side, so I put the Amprobe AM-60 back to auto ranging, plugged the leads into a power bar strip and turned on the power bar. The Amprobe read 860, 77, 140, before settling down to 121V AC. The Fluke 12 read 0L and then 121V AC for the same test. Again, if I switch both to manual range, they both read 121V AC fairly quickly.

In re-watching some of Dave Jones' eevblog multimeter videos (both $50 an d$99 shootout), the Fluke 87-5 does the same behaviour with displaying 0L and then the correct voltage. The Amprobe AM-60 has some overshoot before settling down.

[Toasty]

Re: antec sp 450

Another reason to dislike digital meters....

[Lapres_3]

Re: antec sp 450

I wanted to add to this post, because it seems like a great starting point if one is troubleshooting an Antec SP-450. I apologise if this seems outside the scope of the thread.

I have two Antec SP-450s. One slightly older than the other. The older one has two 1000uf Caps, the newer one had a 1000uf and a 2200uf (EC9 and EC10). One supply slowly developed a whine when off, and it remained constant when on. Also became difficult to turn on the computer. 2200uF cap was replaced and fixed all problems. *

Second supply made noise rather suddenly, loud wine too. Both 1000uf caps were bulged and replaced, but the problem was not solved. On measurement, +5SB was at +13V. Luckily, installed 16V caps instead of 10V. Found that cap near the AC input filter (22uf 50V) read about 4uF. Replaced that with a 33uF 50V (lazy) and all is well.

Debating replacing all caps. Not sure how much life 4 year old supplies have. Yes, may be on borrowed time with those crapacitors.

*After 2nd supply was repaired, 1st supply- 22uF 50V- cap was replaced

Hope this helps someone

[PCBONEZ]

Re: antec sp 450

The Smart Powers had issues with +5vsb and that was probably the 22uF cap.
The noise thing was often associated with it.
.
A full recap would probably get you 5+ years more use out of them.
OTOH they aren't high efficiency supplies - if that matters to you.
.
[I'm assuming EC10 is the 2200uF here.]
If the EC10 position in both PSUs has the same sized coil associated with it then I'd upgrade the 1000uF EC10 to 2200uF.
.

[c_hegge]

Re: antec sp 450

If I find one with a 2200uF and a 1000uF on the 5sb and can't squeeze 10mm caps in, then I usually use two 1500uF 6.3v 8mm panny FKs to achieve about the same total capacitance.