Need help with Antec TruePower2.0 380W

[etnietering]

Re: Need help with Antec TruePower2.0 380W

The changes in the readings are due to ripple voltage. Your voltmeter samples the voltage every so often (dt), which might be like 500ms or 250ms or something. At these points in which it samples, the voltage is on different places in the output waveform.
For example, say the output is a pure cosine wave at 1Hz, and you sample every 250ms. You'll get values 1, 0, -1, 0, 1, 0, etc... The output of the PSU is a lot more complex than this, but it's the same idea.

Judging by those numbers, it looks like the ripple out of the PSU is:
+5vsb : 30mv
3.3v : 20mv
5v : 30mv
12v : 30mv
All of which are in spec, if I remember the spec correctly. So, it's nothing to worry about. As the new caps in the PSU 'burn-in', the ripple current might go down. Speaking of which, I love the sight of a whole bunch of Panny T vents in a PSU! Good job!

[i4004]

Re: Need help with Antec TruePower2.0 380W

yes, my thoughts too.
and i thought how could that be quite a decent way to actually measure ripple, although more tests with few multimeters would be required too...
and few crappy psus...heh...

when tested prior to recap he got

5V----------------------5.05-5.09--------5.06-5.10

and as you say new caps will probably "settle" somewhat to bring that even a bit lower.

that's mighty interesting as it suggests that close inspection of voltage can give a ballpark reading about the ripple...
offcourse, it would be good to devise some voltage divider to get better resolution, ie to measure in the 2000mV range etc.

requirements

from

page22

[i4004]

Re: Need help with Antec TruePower2.0 380W

or..wait, goodpsu already did that right here
https://www.badcaps.net/forum/showpo...6&postcount=52
and got results consistant with overall quality of those psus, ie codegen worst, macron best.

very nice, goodpsusearch, VERY nice!

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

Speaking of which, I love the sight of a whole bunch of Panny T vents in a PSU! Good job!

...And this would never be done without the help I got from that forum

I would like you to know how grateful I feel.. Thank you!

[PCBONEZ]

Re: Need help with Antec TruePower2.0 380W

We had no luck modding the fan controller. We added resistors in parallel with the thermal sensors but no change was observed. When added a very small one, the psu did not start at all (overheat protection was enabled).

If adding a resistor in parallel kicks in overheat protection then it is telling the fan controller the unit is hotter that it is and that would speed up the fan, not slow it down.
Less resistance = higher temp -> speed up fan.

Instead measure DC resistance of fan [it will be tiny] and then use the parallel resistance and Ohms Law equations to determine what resistor in parallel with fan will reduce current through the fan by X-many %.
.

[i4004]

Re: Need help with Antec TruePower2.0 380W

what a crappy idea: putting resistor parallel to fan will just bring down the rpms of the fan and that's quite the opposite of what he wants.
(bonez still seems to think you want silent psu...uh...just can't derail him from that...)

overall bonez suggestions here about the 7v mod(being mighty dangerous) and fan speed modding proved to be useless. electronics is really not his field, and he didin't really understood what they were doing.

modding should be done in general direction they attempted (by varying the voltage of the sensor, ie to "trick" circuit into thinking it's hotter than it really is, to bring RPMs up) but by first monitoring voltage on it(while also monitoring the heat) and then see if it's ntc or ptc, and then perhaps serial resistor would be needed, etc.
best would be to put trimmer so it can be varied...

but anyway if 7v is ok why bother?

[momaka]

Re: Need help with Antec TruePower2.0 380W

We had no luck modding the fan controller. We added resistors in parallel with the thermal sensors but no change was observed. When added a very small one, the psu did not start at all (overheat protection was enabled).

You need to look at wheather the thermistor is PTC or NTC (most likely it will say so on the PCB).
If it's PTC, add a resistor in series with the termistor.
If it's NTC, add a resistor in parallel with the termistor.
The hard part is picking the value of the resistor. This you will have to do by measuring the resistance of the PTC/NTC at different temperatures (for example, when the PSU is cold and when it is warm).

[PCBONEZ]

Re: Need help with Antec TruePower2.0 380W

momaka is correcting at the sensor.
I was telling how to correct at the fan motor.
Either will work but to me when the issue is the fan is too fast it's easier to fix at the motor.
[Note: Correcting at the motor only works with two wire fans.]
.

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

I am satisfied with the 7v solution at the moment. It gets warm but not too hot as earlier, and it is not hearable at all.

Thank you all

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

btw. if you're gonan do this often, esr meter would be handy.
this way you're just guessing what's ok, what not, what needs to be changed etc.
let me guess, no money for that?
<wink>

Well, I got ESR micro v4.0!

And I have a number of questions about antec old caps readings...

First of all, I measured 4x4700uF 10V 10mm Fuhjyyu:
capacitance varied from 4400 to 4800 and esr was 0.01

So, those caps were actually good! Low esr and the capacitance claimed was true.

It was said in that forum that their actual capacitance was not more than 3100uF, but that did not happen here..

Given that they were replaced with 4x3300uF 6.3V Panasonic FK (esr 0.03, capacitance 3020uF) I am wondering if that results in higher ripple in the 5V and 3.3V output.

What do you think?

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

Some more testings of the original psu caps:

4700uF 10V Fuhjyyu: 4400 to 4800, esr 0.01
3300uF 16V Fuhjyyu: 3100, esr 0.01
470uF 25V Fuhjyyu: 468, esr 0.05
220uF 16V Koshin: 224, esr 0.20
1000uF 10V bulging Fuhjyyu: 7.3, esr 15.8
1000uF 10V bulging Fuhjyyu: 0.6, esr 51
1uF 50V Koshin: 1, esr 25
10uF 50V Fuhjyyu: 10.9, esr 3.0

In the last 2 pictures of the first post, notice the PCB discoloration near the 5vsb circuit (particularly the resistor and diode next to 5vsb transformer). This mean you definitely need to change the critical 5vsb cap (which is usually either 10, 22, or 47uF). Seems to be the one right next to U6 (5vsb transistor).

The following three caps were placed near the 5vsb circuit. I think momaka was right when he said they definitely needed recap:

100uF 25V Fuhjyyu: 0.3, esr - (see attached photo)
22uF 50V Koshin: 19.1, esr 27
22uF 16V Fuhjyyu: 12.9, esr 40

I need your opinion about the esr numbers... For example is 1uF 50V Koshin esr 25 normal?
What numbers are expected from low esr caps and normal esr?

Also, attached photo of 3300uF 6.3V Panasonic FK, the type I replaced the Fuhjyyu 4700uF 10V caps with:

https://www.badcaps.net/forum/attach...id=14179&stc=1

Can anyone tell if replacing 4700uF 0.01esr with 3300uF 0.03 esr caps could result in noticeable higher 5V and 3.3V output ripple?

[PCBONEZ]

Re: Need help with Antec TruePower2.0 380W

The problem with Fuhjyyu [for that matter all 'bad brands'] isn't that they are guaranteed to be bad but that they are inconsistent.
-
I did a number of Antec TP2-550w a while back and in some every 10mm Fuhjyyu was bloated and leaking with all the 6mm & 8mm still okay, in others the 10mm were all okay and the small ones all popped, some had all bad, none had all good.

You are much better off with the FK's in there.
.

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

Thanks for the reply PCBONEZ.

One last thing I'd like to know is the range of esr in low esr caps.

What's the biggest number one can find in a low esr capacitor and still think of it as good one?

Also, if you like, check my questions about another psu jou jye electronic 460watt

[PCBONEZ]

Re: Need help with Antec TruePower2.0 380W

What you are asking is what I call 'grade'.
'Grade' in my context compares ESR and Ripple ratings from one brand or series to another.
.
The way to tell grade is to look at the can size, the ESR, and the ripple.
In good brands nearly all of the same can size [regardless of uF & volts] will have the same ESR and Ripple. [Not 100% in series that cover a huge range of values but nearly so.]
- You then look at the same can sizes in the series you are comparing to and determine which series is the better grade.
-
The problem with that is that many crap brands 'lie' in their data sheets. I don't think it's always deliberate, I think usually their QC and testing methods just suck so they end up with bogus data in their tables.

To see what I mean look at the data sheets for 8x20mm in Rubycon MBZ, MCZ, ZL, and Chemicon KZE.

8x20mm [uF and volts doesn't matter for this.]
MCZ - 2350 / .012
MBZ - 1870 / .019
ZL - 1250 / .041
KZE - 1250 / .041

MCZ is higher 'grade' than MBZ which is higher 'grade' than ZL & KZE.
But ZL & KZE are the same 'grade'.

But knowing the grade only tells you which data sheets [series] are most likely to have what you need. There will be cases where a lesser 'grade series' has a better rated cap because the can size is bigger in the lesser grade.
[This is why you see so many 10v caps where a 6.3v would do. The 10v caps tend to have *larger cans* which gives them better ESR and ripple withing a given grade [and cost]. It's a cheap upgrade since [say] an 8x20mm 10v cap doesn't really cost anymore than [say] a 8x16mm 6.3v cap.]

For original caps in PSU OP filters what I usually see are about the grade of Rubycon ZL -or better- up to about Panasonic FM.
- There are A LOT of exceptions to that because there are so many PSU designs out there but if there is a 'normal' for that application then the range from ZL to FM would be it.
[As I recall Panny FK are slightly better grade than KZE & ZL.]

- Important -
It is not good to drastically change the ESR of filter caps that have a coil [inductor] in the circuit. The ESR+uF and uH of the coil are chosen to maximize the filtering at the expected ripple frequency. [People get taught that the freq response this is based only on uF [Xc] & uH [Xl] but that isn't true. Electronics training for filters is usually based on Radio Frequencies which are much higher than the Ripple-Noise frequency in a SMPS. At Radio Frequencies ESR is nill so they don't talk about it. Down at SMPS Ripple frequencies ESR becomes as large [and even larger] than Xc and needs to be included in the calculations.]
- This is because Xc [the reactance, or ~the AC resistance~] gets smaller as frequency gets smaller and ESR gets larger.
Down at low SMPS Ripple Frequencies (ESR+Xc) & uH are what controls filter effectiveness/efficiency.
---- And so, if you lower uF you will shift the effectiveness of the filter to a higher frequency leaving more SMPS Noise in the circuit. To compensate you need to lower the ESR too.
~~
To do the math you have to think of the Ohms from Xc and the Ohms from ESR as parallel resistors. As with actual resistors it's the SMALLEST Ohms value that has the most 'control' over the current. [In this case Ripple Current.] - And because both ESR & Xc change with frequency you have to re-do the math for each frequency you care about.
~~

It's all quite messy.
Easier to just stick to original values when you can.
.

The 6 & 8 mm [non-OP filter] caps found originally in PSUs are -usually- around or just below the grade of Panny FC or Nichicon PW. There is occasionally a wringer that's up around FM.
[These usually don't have any coils in circuit.]

The little 4 & 5 mm guys are usually just barely low ESR or even basic GP caps.
I usually use FC or PW because they are better and cheap. [Sometimes I get FM instead because although FM are much better that FC or PW they are often less expensive.]
[These usually don't have any coils in circuit.]
.

[goodpsusearch]

Re: Need help with Antec TruePower2.0 380W

Thank you a lot PCBONEZ. It's all clear to me now.

In fact, I read your post many times until I understood the best of it.

I saved it in bookmarks for future reference!

[PCBONEZ]

Re: Need help with Antec TruePower2.0 380W

Congrats!
Once you 'get it', it's pretty easy. Glad you got it now.

I make a wording mistake over and over which probably confuses things a little.
[Sorry]
Often when I say something like "ESR gets big" or "Xc gets big" what I actually mean is "the effect of _____ gets big".
For both Xc and ESR when the effect gets big it is because the value got small.
[The smallest value resistor [in parallel] has the most effect on current.]

When ESR is tiny it has the most effect on current.

.