Question about 12V ripple

[Pentium4]

Worried ripple might be too high on this system I'm using daily. It's my Xigmatek PSU https://www.badcaps.net/forum/showpo...postcount=1156

I think it's good for 350W at least but I'm worried about the 12V ripple. All it has on the 12V is 1 pi coil and one 4700uF Panny FR cap. It's powering a Core 2 Duo 2.4GHz proc, 4GB RAM with a BFG GTX 260 and a single SSD drive. Sometimes when doing medium gaming, the PSU gets around 200W. Efficiency is around 77% I don't have an oscilloscope but as a guess would you say the ripple isn't that bad?

[Wester547]

Re: Question about 12V ripple

Though I would recommend at least two filtering capacitors for each output, 4700uF x1 should be theoretically enough for the +12V rail... so long as the ripple is well under 120mV (that's quite a bit of electrical noise and heat), you should be okay. Any PSU that gives off 100mV of ripple on the +12V rail isn't one that I'd feel safe with, though. I would guess that the heat output from the PSU itself isn't bad, overall? If you don't feel safe with it right now, I don't think it would hurt to add an extra 2200uF capacitor on the +12V output (if there's room). Note that 120mV is the limit for the +12V rail and 50mV the specified limit for the +3.3V and +5V rails.

[Pentium4]

Re: Question about 12V ripple

Yeah I wish there was another slot for a cap but there wasn't, that's why I put a huge 16V cap in the only spot I figured the coil would help some too though, even at 200W it doesn't kick out a ton of heat because my room is around around 15C

[ben7]

Re: Question about 12V ripple

Just a note - adding/decreasing capacitance and/or esr can cause MORE ripple in some psus!

[Wester547]

Re: Question about 12V ripple

^ I would guess that augmenting capacitance can increase ripple because as capacitance goes up in capacitors, ESR goes lower (larger can sizes, usually, so higher ripple current and lower ESR rating...).

[c_hegge]

Re: Question about 12V ripple

I have load tested a similar PSU (http://www.hardwareinsights.com/wp/a...520c-review/2/). The ripple was a little high, but still in spec.

[ben7]

Re: Question about 12V ripple

^ I would guess that augmenting capacitance can increase ripple because as capacitance goes up in capacitors ESR goes lower (larger can sizes, usually, so higher ripple current and lower ESR rating...).

It affects the way the control loop works, and the frequencies that get filtered out.

[Pentium4]

Re: Question about 12V ripple

Shoot....I hope the ripple isn't higher! I used my little ESR meter and the Panny cap was actually 0.01 higher than the Teapo if I remember correctly

[c_hegge]

Re: Question about 12V ripple

^
a 0.01 ohm difference isn't going to make much difference.

[momaka]

Re: Question about 12V ripple

Just a note - adding/decreasing capacitance and/or esr can cause MORE ripple in some psus!

Only somewhat true.

Increasing the capacitance alone without lowering the ESR or increasing the ripple current will actually decrease the output noise and ripple. However, Wester547 stated a valid point - the can size of the capacitor often becomes larger with increasing capacitance, so that means the ESR is lowered and the ripple current increased - and this is exactly what can cause the PSU to actually output more noise.

The 4700uF Panasonic FR should be fine, though - of course, all depending on the TDP rating of the Core 2 Duo and the 260 GTX. I've successfully ran a 107 W TDP CPU with a Macron MPT-301 PSU that has only a single 3300uF 16V Chemicon KZE cap for the 12V rail.

[Wester547]

Re: Question about 12V ripple

EDIT: Nevermind, his PSU is a half bridge too...

[Pentium4]

Re: Question about 12V ripple

Only somewhat true.

Increasing the capacitance alone without lowering the ESR or increasing the ripple current will actually decrease the output noise and ripple. However, Wester547 stated a valid point - the can size of the capacitor often becomes larger with increasing capacitance, so that means the ESR is lowered and the ripple current increased - and this is exactly what can cause the PSU to actually output more noise.

The 4700uF Panasonic FR should be fine, though - of course, all depending on the TDP rating of the Core 2 Duo and the 260 GTX. I've successfully ran a 107 W TDP CPU with a Macron MPT-301 PSU that has only a single 3300uF 16V Chemicon KZE cap for the 12V rail.

Cool They're both running at stock speeds and the case has good airflow. And at least it has an SSD instead of a spinning disk, so slightly less 12V needed

[Wester547]

Re: Question about 12V ripple

the can size of the capacitor often becomes larger with increasing capacitance, so that means the ESR is lowered and the ripple current increased - and this is exactly what can cause the PSU to actually output more noise.

I was of the impression that there wasn't such a thing as "too high a ripple current" in capacitors... if only because it increases the maximum amount of ripple they can handle (the maximum permissible ripple current for the capacitors, though I know that lower ESR ratings and higher ripple current ratings in capacitors are almost unavoidably synonymous with larger can sizes), but then a good design will be such that there won't be much ripple to begin with.

However, in saying that, if capacitors allow more ripple into the PSU (and thereby out) since they can handle it with a higher ripple current, it does make sense... unless you mean that the higher the ripple current, the more ripple has to be "released" to ground.

[ben7]

Re: Question about 12V ripple

I think it is more the different esr and/or capacitance messes with the pole in the feedback loop. Unique sure could tell us.