Hardware Insights Forum

LongRunner wrote:I think these PSUs do require low ESR caps.

It depends on what the feedback loop is tuned to and it depends on what other methods the PSU uses to suppress ripple. You could probably get away with anything as high as 75-85 ohms rated capacitors ESR wise. True, having lower ESR capacitors does mean an ultimately lower ripple voltage.

PCBONEZ's explanation (half ESR = twice the ripple current) assumed that the ripple voltage remains the same

Well, it can't, because lower ESR capacitors keep the ripple voltage lower anyway. But in order to do that, they have to work even harder.

If you replace (say) [b]one cap out of eight in parallel in a VRM with one of lower ESR, you will indeed get more ripple current through that individual cap, because they are at the same voltage and lower resistance across the same voltage does indeed mean more current, but if you replace all of the caps with lower ESR units the ripple voltage will drop, and the current through the group will (hypothetically) remain the same. (In practice it's not that simple, but I'm sure it's closer to the truth than what PCBONEZ said.)

PCBONEZ did say that one 3300uF capacitor would overheat where two 1500uF capacitors in parallel were designed to work. I could understand why.

But in order to do that, they have to work even harder.

PCBONEZ did say that one 3300uF capacitor would overheat where two 1500uF capacitors in parallel were designed to work. I could understand why.

I'm not sure specifically why he said that but it might be because ESR and ripple aren't linearly related to case size even within a series. In Panasonic FR for example 10x20mm is 20mΩ but 12.5x30mm (just over twice the volume) is 13mΩ, not 10mΩ. The lower you go in ESR, the more significant this effect becomes.

(Edited as having read the engineering documents, there isn't much more to impedance than ESR at the switching frequency. It was a "higher-ranking" Badcaps.net member that told me there was.)

LongRunner wrote:

But in order to do that, they have to work even harder.

The lower the ESR rating, the more ripple is passed (the more ripple is "dumped" to the ground through the capacitor, so it has to dispense of more ripple, thereby more internal heating).

Continuing with the MPT-4012...

The +3.3V and +5V have the same rectifiers as in MPT-xxx. The +12V now has 2 x MBR20100CT in parallel. The +5VSB has an SBL1040CT.

This time there are pi coils on all the rails (except -12V which uses a 7912 as previously mentioned), and the one on the +12V is massive. All of the main rails have OCP. +12V is split into two - one for the CPU (15A, black stripes) and one for everything else (14A; the combined limit is unspecified).

The main switchers are 2 x 2SK2765 in parallel and the 5VSB switcher is a KHB4D5N60F.

Unlike the older MPT-xxx units with their distinctive colour scheme (purple for +3.3V, orange for Power Good, brown for +5VSB and grey for on/off), MPT-xxx2 uses the standard colours.

Having seen an MPT-301 with G-Luxon LZ on +3.3V, +5V and +5VSB, they are 25mm tall, not 20mm as I guessed at in the first post. I have fixed it.

Any luck obtaining a review sample???

Pentium wrote:Power Logic fans seem to be reliable too which I've seen in every Macron

At least those I've checked under the label on have actual oil in the bearing, not that grease that sometimes goes sticky.

The second set of attachments is here (due to the 3-per-post limit). As for the circuit used by CWT and a few others (including Jou-Jye IIRC) that blows 2.2µF 400V electrolytics by putting them next to hot resistors, read this review. I severely doubt a circuit with two hot resistors is going to improve efficiency, though and it won't boost the voltage noticeably with the main supply running anyway.

The MOVs are between the primary capacitors and the heatsink. The two unlabelled through-holes are seemingly for the PFC coil terminations had it been installed, which would require cutting a PCB trace between them (odd that they didn't just use a jumper for the no-PFC version).

And of course, there's this one...

I'll probably just get one of these - http://www.ebay.com.au/itm/Macron-400W- ... 2a32b3ce51

Go ahead. I was thinking along the lines of the MPT-xxx2 being worthy of a full review, though. You could get both it and the old one and include the latter in the cheapo round-up.

I might do that, actually, although I do have on in this year's roundup which is actually a kind of OK 450-500W 2TF design. Too bad it had an 850W Label.