Fortron Zen 300 W incredible ripple

[Behemot]

I got this passive Zen to have a look at it. It had problem to turn on with no load. This seems to be a feature more than a bug, or to be precise, it has to have some load on either +12 V or +3,3 V. If there is load only at -12 V or +5 V (or higher than on other rails), protection kicks in.

Other than that, there was bad cap at +12 V. I have filled it with Samxon RS 3300 uF/16 V, it was tough due to D8 positions but I made it. I've accidentally changed polarity of one at +3,3 V, this is good now, but there is still incredibly huge ripple (more of it on other rails actually) and I don't really know what to do with that.

Total capacitance is aprox. 14,5 mF/+12 V, 6,5 mF/+3,3 V, 7,5 mf/+5 V, 700 uF/-12 V and 4,5 mF/+5 V SB. But other rails than +12 V are very noisy, check screens. It uses one output rail and magamps for both +3,3 and +5 V. Do you think it's bug or future as well? It get's better without load at -12 V but it's still out of spec by almost 100 %…

[Behemot]

Re: Fortron Zen 300 W incredible ripple

I've been playing around with it and it seems like the problem is in the switching part. The ripple seems to be at 108 kHz (double the swtching frequency), measuring it directly in molex revealed it is there and the decoupling caps are just able to filter part of it out, making such blocks you can see on the images.

Adding 3300uF did not help at all.

[Behemot]

Re: Fortron Zen 300 W incredible ripple

Been thinking, as those are tall but short high-frequency spikes, could adding small ceramics in parallel with the big electrolytes help?

[tom66]

Re: Fortron Zen 300 W incredible ripple

Sounds more like a probing issue. Try limiting bandwidth to under 1MHz.

[Behemot]

Re: Fortron Zen 300 W incredible ripple

No it's not probing issue ofc. It occured to me as a first thing so I've tried two different PSUs just to check if I am not taking bad data. The other two were both just fine around 20-30 mV. As for limiting under 1 MHz, I believe the scope has only 20 MHz limiter which is also spec'd in ATX guidelines.

[PeteS in CA]

Re: Fortron Zen 300 W incredible ripple

Does it have ferrite rod inductors for the O/Ps (in addition to the big toroids)? If not, I think that is the purpose of ferrite rod inductors. If it does, try ceramic caps on the O/P side of the ferrite rods. Also, check to be sure all the connections to chassis ground are tight. A poor ground connection can show up as noise. Another possible cause might be a snubber that isn't doing its job - e.g. a bad solder joint or toasted resistor - or was left uninstalled to cut costs.

[Behemot]

Re: Fortron Zen 300 W incredible ripple

Well now that you're asking, there are ferrite rod inductors and two "air core" inductors (just more wire without any core) what FSP uses. If the two are for +3,3 and +5 V, do you think that's it?

ADD// seems like the inductors without core are on the two +12 V rails so +3,3 and +5 V have ferite rod ones. Heres an image.

[tom66]

Re: Fortron Zen 300 W incredible ripple

...There is a bad cap on that PSU.

[Behemot]

Re: Fortron Zen 300 W incredible ripple

...There is a bad cap on that PSU.

Other than that, there was bad cap at +12 V. I have filled it with Samxon RS 3300 uF/16 V, it was tough due to D8 positions but I made it.

orly

[Behemot]

Re: Fortron Zen 300 W incredible ripple

Oh my lord! Something is very very wrong with this PSU! I first though I have accidentally soldered the cap on +3,3 V with reversed polarity but now it seems I haven't. The other Samxon RS 3300 uF/16 V i have soldered in on +3,3 V output has bulged again!

What does make a brand new quality cap bulge in matter of hours?!

ADD// I've been thinking that maybe it is little bit oscillating so I've downgraded the output capacitance a litle bit on +3,3 and +5 V (KY 1200 uF/6,3 V instead of RS 3300 uF/16 V) but that increased the ripple even more…

ADD2// I have found one bad SMD zener diode on what seems to be a +12 V feedback. It has three colro bands: black, yellow, yellow.