Same model, different PPFC chokes. Are they real? (MPT-350P)

[goodpsusearch]

The power supply is Macron Power MPT-350P. I already had a 2004 model and now I got another one made in 2005.

There are some differences between them and one of them is the different pfc chokes. Are they real?

If they are not fake, which one is the best?

[c_hegge]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

They both look real - I can see some coil wire in them. I'm guessing the one with black and white wires is probably better, but the only way to know is to try a PSU with both of them and see which one gives you a better power factor.

[goodpsusearch]

More differences

They both look real - I can see some coil wire in them. I'm guessing the one with black and white wires is probably better, but the only way to know is to try a PSU with both of them and see which one gives you a better power factor.

Thank you! Another difference between the 2 models is that the 2004 one has 18AWG wire except the 12V connector for the CPU being 20AWG.

The 2004 model also has the ancient AUX connector. The 2005 model includes a 20+4 ATX connector and every wire is 18AWG.

One last difference is that the primary caps of the 2004 model are Fuhjyyu 1000uF while the 2005 model has G-Luxon 1000uF primary caps that are slightly smaller that the Fuhjyyu caps. And that of course has an impact at capacity:
Fuhjyyu capacitors were found to be 1000uF caps while the G-Luxon caps' real capacitance was only 909uF

I attached some photos. I am tempted by the idea to replace the 12V rectifier (F16C20C) with something bigger. Will it make any change? What do you think?

[Khron]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

For what it's worth, i've seen Fuhjyyu primaries bulge, so... I wouldn't trust'em either way, even if they measured ok. Like C(r)apXon, they're BOUND to fail (sooner rather than later).

And about the G-Luxon measuring a bit low, electrolytics have standard +/-20% tolerance, so it would still be in spec anywhere between 800uF and 1200uF

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

For what it's worth, i've seen Fuhjyyu primaries bulge, so... I wouldn't trust'em either way, even if they measured ok. Like C(r)apXon, they're BOUND to fail (sooner rather than later).

Those were in Antec ovens with APFC boosting voltage to 400V DC.

[Khron]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Those were in Antec ovens with APFC boosting voltage to 400V DC.

Point taken But if i can rid / save myself from a headache with a wee bit of preventive maintenance, i would / will

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

There is an empty place for a second 12V rectifier on PCB.

I can use the following parts from dead power supplies:

MBR20100CT
F12C20C
F16C20C

I can't decide which of the following I should do:
-replace existing F16C20C with an MBR20100CT (20A)
-replace F16C20C with 2xF12C20C (24A total)
-add a F16C20C in parallel with the existing F16C20C (32A total)

How many Amp do you think this psu could provide on the 12V?

[c_hegge]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Go with option 3 (adding the 2nd F16C20C). Even if you aren't loading it to 32A, they will be more efficient when not heavily loaded. The PSU would probably handle 20-25A OK

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Thanks! Does it hurt that the 2 Mospec F16C20Cs are not from the same batch?

[c_hegge]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

I doubt it. I've used similar rectifiers from different batches before and it works fine.

[momaka]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Go with option 3 (adding the 2nd F16C20C). Even if you aren't loading it to 32A, they will be more efficient when not heavily loaded. The PSU would probably handle 20-25A OK

Actually, the MBR20100CT by itself may be just a tad more efficient because it's a Schottky rectifier and has a lower forward voltage drop. (the forward voltage drop times the current passing through the rectifier gives you the power that is wasted as heat).

If you look at the typical forward voltage graphs for these rectifiers, you can actually tell how much power your be wasting as heat per given current draw.

Let's say we draw 20A from the 12V rail on the PSU.
- With the MBR20100CT, the forward voltage drop will be just a bit less than 0.85V - and that's worst case scenario. So for 20A, the rectifier will be wasting 0.85x20 = 17W as heat.
- With 2x F16C20C rectifiers in parallel, the forward voltage drop will be very close to 1V (each rectifier will be handling only 10A, so you're looking for the forward voltage drop at 10A). Therefore, each rectifier will be wasting 1x10=10W as heat, or 20W total for the two in parallel.

So, as you can see, the MBR20100CT is a bit more efficient. Now try repeating the same excercise for 10A and 5A of load. In each case, the MBR20100CT should be be more efficient.

There is, however, something else that's important too.
First, the 2x F16C20C in parallel will be contacting the heat sink over a larger area, so they may actually run cooler.
Second, the single MBR20100CT is limited to 20A of current. Beyond that, it will likely begin to produce a lot more heat and may eventually become less efficient than the 2x F16C20C rectifiers. Also, the 2x F16C20C rectifiers will be able to handle up to 32A of current. The rest of the PSU may not be able to supply that kind of current, but you're not limited to 20A anymore so that's also something to consider.

In the end, though, I think I too would recommend you go with the 2x F16C20C rectifiers just because the PSU looks like it may handle 20 to 25A fine.

[Behemot]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Yep, exactly what I was to post before I've read momaka's post

Best solution would be to use two lower-current schottkies, or one higher-current one (preferably in bigger physical package as well). I tend to collect more and more super fasts, because, what to do with them when there are shottkies?

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

I did the replacement momaka and c_hegge suggested and everything works fine

I let the psu run fanless for some minutes and 10W load at 12V and the secondary heatsink remained cool.

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Actually, the MBR20100CT by itself may be just a tad more efficient because it's a Schottky rectifier and has a lower forward voltage drop. (the forward voltage drop times the current passing through the rectifier gives you the power that is wasted as heat).

If you look at the typical forward voltage graphs for these rectifiers, you can actually tell how much power your be wasting as heat per given current draw.

Let's say we draw 20A from the 12V rail on the PSU.
- With the MBR20100CT, the forward voltage drop will be just a bit less than 0.85V - and that's worst case scenario. So for 20A, the rectifier will be wasting 0.85x20 = 17W as heat.
- With 2x F16C20C rectifiers in parallel, the forward voltage drop will be very close to 1V (each rectifier will be handling only 10A, so you're looking for the forward voltage drop at 10A). Therefore, each rectifier will be wasting 1x10=10W as heat, or 20W total for the two in parallel.

So, as you can see, the MBR20100CT is a bit more efficient. Now try repeating the same excercise for 10A and 5A of load. In each case, the MBR20100CT should be be more efficient.

There is, however, something else that's important too.
First, the 2x F16C20C in parallel will be contacting the heat sink over a larger area, so they may actually run cooler.
Second, the single MBR20100CT is limited to 20A of current. Beyond that, it will likely begin to produce a lot more heat and may eventually become less efficient than the 2x F16C20C rectifiers. Also, the 2x F16C20C rectifiers will be able to handle up to 32A of current. The rest of the PSU may not be able to supply that kind of current, but you're not limited to 20A anymore so that's also something to consider.

In the end, though, I think I too would recommend you go with the 2x F16C20C rectifiers just because the PSU looks like it may handle 20 to 25A fine.

That's a great source of info!


I did a full recap on that thing and now it powers an Athlon 64 3000+ without even getting hot and its fan remains fairly silent.

I did not change the orange 1uF caps and the primary 1000uF G-Luxon caps (too expensive).

2x1000uF 200V G-Luxon -> left alone

1uF 50V G-Luxon -> 1uF Nichicon PW 50V

3x1uF 50V Jpce-tur -> left alone

4x10uF 50V G-Luxon SM -> 4x10uF 50V UCC KY

1x470uF 25V -> 1x 470uF 35V Nichicon HE

+12V
1x3300uF 16V G-Luxon LZ -> 1x3300uF 16V Panasonic FR

+5Vsb
2x2200uF 10V GL -> 2x2200uF 16V Panasonic HFQ (from an old Intel PII motherboard)

+5V
2x2200uF 10V GL -> 2x3300uF 6.3V Panasonic FK

+3.3V
2x2200uF 10V GL -> 1x2200uF 16V Panasonic HFQ and 1x2200uF 16V Nichicon PW

-12V and -5V:
2x220uF 16V GL RK -> 2x220uF 50V Nichicon HE

It is very hard to fit 12.5mm diameter caps on those 10mm pcb spots.

I would like to thank everyone for your help

[c_hegge]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

It is very hard to fit 12.5mm diameter caps on those 10mm pcb spots.

I know what you mean. I think I'll do this if I have to to use 12.5mm caps in the future (http://*************/photo/my-images/32/img0020rp.jpg/). You might also be able to squeeze a few coils on there too if you're modding cheapies

[momaka]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

I did a full recap on that thing and now it powers an Athlon 64 3000+ without even getting hot and its fan remains fairly silent.

Cool

As I might have mentioned (probably many times before), I have the MPT-301 (300W). It's currently sitting in a PC with an AMD Athlon 64 3200+. I even tried an Athlon 64 FX-53 CPU (105W TDP) in that PC, and the 12V rail never dipped below 12V. 5V rail didn't go high either. I expected it to cross-load badly, but it didn't. Not only that, but it also remained fairly cool (even though it actually has much smaller heat sinks than yours). I also tried that same power supply in a PC with an Intel i7 2600k processor. Got very similar results as I did with the FX-53. Aside from the 2-transistor 5VSB, I think these are fairly solid PSUs.

BTW, I also left the original primary caps in mine (200V 470uF Fuhjyyu), along with the small orange PCE-TUR caps.

*EDIT*
Hmm, apparently I never posted pictures of my MPT-301 on the PSU pictorial thread. I swear I remember doing so, though.
I need some sleep....

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Aside from the 2-transistor 5VSB, I think these are fairly solid PSUs.

And the good thing about its 5vsb circuit is that somehow they managed to do it without a critical cap.


BTW I tested the different passive PFCs today with 65-92 watt load (the Athlon 64 3000+). There wasn't any significant difference between them at that load. The 2005 pfc gave 0.7-0.71 power factor and the 2004 pfc gave 0.68-0.7.

Equipment used for testing: https://www.badcaps.net/forum/showthread.php?t=13768

[momaka]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

And the good thing about its 5vsb circuit is that somehow they managed to do it without a critical cap.

Indeed.
Although, the 5VSB on mine is a bit high - 5.15V or 5.20 IIRC. It also squeals a little when PC is OFF or in stand-by so I have been turning OFF the PSU from the back when I'm not using the PC. I did leave one of the original 10V 2200uF Fuhjyyu filter caps in the 5VSB, though I don't think that's the problem because it's the 2nd cap in the filter and the first cap is a 6.3V 2200uF Panasonic FM, which should be taking most of the ripple. But who knows. Maybe I'll replace it someday just to see if that brings the 5VSB a little closer to 5V. Again, it likely isn't a problem, but most 2-transistor 5VSB circuits I recapped always came back to 5V exactly, give or take 50mV.

[goodpsusearch]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

Panasonic FM is not suitable for 5vsb of that power supply. Those Macron power supplies use general purpose caps in every output! I had a similar problem after recapping MPT-301 (this was my first recap), I used Panasonic FK 2700uF but the psu still felt they were too low esr for 5vsb.

[momaka]

Re: Same model, different PPFC chokes. Are they real? (MPT-350P)

That's weird, because self-oscillating circuits like that tend to benefit from caps with low ESR. I recapped the 2-transistor 5VSB circuit of a Morex MXA-235PTF1 with 2x custom 6.3V 820uF Rubycon MFZ caps from an Xbox 360 (these are equivalent in ESR to 6.3V 1500uF Nichicon HN). PSU was dead quiet for all load ranges on the 5VSB (I tried 0.01A to 1.1A, which is 100mA over what the label says the PSU can do). The 5VSB also ran much much cooler.

Perhaps it could also be the high capacitance? I think I might do some tests on mine.

I really need to get me some more not-so-low ESR caps too . I have lots of ultra-low ESR caps from motherboards (2x 1KG jars). Enough to probably do 20 computers.