Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

[goodpsusearch]

The label says it is 430W but the pcb says “HEC-250TN”. At least it has a full input filter, only the thermistor is missing. The primary caps real capacitance is 470uF. That makes me wonder if someone did fake Teapos.

-Passive pfc

-GBL406 4A bridge rectifier

-UC3843B PWM controller

-S2N60 N-Channel MOSFET for 5vsb

-TPS3510P IC for psu protection

-heatsinks only good for 350W max provided there is good airflow and high efficiency

-main transformer is ERL 35 size. Not good for 430W in my opinion.

-I was unable to read the main mosfet's marking. There is something like green glue attached to it. It is in small package and there is only one there. I bet it can't do more than 350W without blowing up.

-psu fan is thermally controlled and "Hec" brand.

-All caps are Teapo, DON and Su'scon

-1x MBR3060 @ 12V

-1x STPS3045C @ 5V

-3.3V: I think current passes through an STPS3045C and then a 09N03LA Power-Transistor, I am not sure, I am not familiar with this design... Could someone explain this to me please?

-Soldering is ok

-Of course it is not dual rail as the label states. It is single rail psu.

What do you think its actually wattage is?

[TELVM]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

Excuse the potentially stupid question: Is that what's called 'single transistor forward' topology?

[Pentium4]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

-3.3V: I think current passes through an STPS3045C and then a 09N03LA Power-Transistor, I am not sure, I am not familiar with this design... Could someone explain this to me please?

It looks like the STPS3045C technically outputs the power for the 3.3V, and the 09N03LA acts as a voltage regulator to step the power down from the 5V rail

What do you think its actually wattage is?

350W seems about right.

Excuse the potentially stupid question: Is that what's called 'single transistor forward' topology?

Yes

EDIT: I've seen those fake Teapo's too! https://www.badcaps.net/forum/showpo...6&postcount=51 Are you gonna recap that? It doesn't look bad.

[Th3_uN1Qu3]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

I've seen that platform before. It's actually quite decent. Certainly worth a recap. I had the 350W version (and of course, labeled as 500W), brand was Rasurbo IIRC. Primary switching FET, UC3842 and gate resistor were blown on mine. The PSU will do well if that FET is replaced with something more beefy.

[goodpsusearch]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

It looks like the STPS3045C technically outputs the power for the 3.3V, and the 09N03LA acts as a voltage regulator to step the power down from the 5V rail


350W seems about right.


Yes

EDIT: I've seen those fake Teapo's too! https://www.badcaps.net/forum/showpo...6&postcount=51 Are you gonna recap that? It doesn't look bad.

The owner of the computer just wants it to work again. Unfortunately, it seems he destroyed something trying to clean PC from dog hair..

The psu works without a problem.

[momaka]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

-3.3V: I think current passes through an STPS3045C and then a 09N03LA Power-Transistor, I am not sure, I am not familiar with this design... Could someone explain this to me please?

This is an independent linear 3.3V rail design. With this design, the transformer's 5V tap is taken and rectified (with the STPS3045C, in your case) and filtered. Then this voltage rail (which is separate from the main 5V rail) is stepped down in a linear fashion with a MOSFET (very inefficient at moderate-to-high loads, but very clean and stable voltage output). There's usually another cap after the MOSFET to further smooth power. You can recognize this design by the lack of torroid coils for the 3.3V rail.

There's also the dependent linear 3.3V rail design. This design is usually found in very cheap PSUs because it uses very few parts (i.e. cheap). It's the same as the design above, except you don't have a rectifier. Rather, the MOSFET is few directly from the main 5V rail of the PSU, and the MOSFET simply steps it down to produce 3.3V. You still get the same clean output voltage, but the downside is that the 5V rectifier needs to be bigger since it is used for providing power to both the 3.3V and 5V rails (so the current rating of the 5V rectifier defines the combined power on 3.3 and 5V rails).

Both of these designs are referred to as linear 3.3V designs. The other design is the one you're probably more familiar with - a small torroid coil for the magamp regulator and a bigger torroid coil next to the 5V/12V torroid coil.

What do you think its actually wattage is?

Depends on what the main MOSFET on the primary is rated for. The Thermaltake TR2-430W (XP550-NP) uses the same platform, and although rated at 430W by Thermaltake, it is only capable of 350W max (PSU has OPP if I remember correctly, so it shuts down if you try to pull more than that out of it).

The psu works without a problem.

That may be. But it has a 2-transistor 5VSB design that DOES have a critical cap and is capable of being a motherboard killer (I think even someone on Badcaps posted about his/hers having a voltage on the 5VSB above 6V). So make sure to change that critical cap with something good.

EDIT: I've seen those fake Teapo's too! https://www.badcaps.net/forum/showpo...6&postcount=51 Are you gonna recap that? It doesn't look bad.

Confirmed
My TR2-430W has two 680 uF Teapo caps as well, and both measured around 470 uF. See the chart at the end of my post here:
https://www.badcaps.net/forum/showpo...5&postcount=57

[Wester547]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

This is an independent linear 3.3V rail design. With this design, the transformer's 5V tap is taken and rectified (with the STPS3045C, in your case) and filtered. Then this voltage rail (which is separate from the main 5V rail) is stepped down in a linear fashion with a MOSFET (very inefficient at moderate-to-high loads, but very clean and stable voltage output). There's usually another cap after the MOSFET to further smooth power. You can recognize this design by the lack of torroid coils for the 3.3V rail.

Do you mean that there are two 5V outputs, in a way, in such a linear regulated design? I ask because I wonder if the Hipro units that use linear regulation to generate +3.3V actually do have a separate transformer tap for that rail or if it's just taken from +5V in the manner you mention (that would dissipate quite a bit more power... if it's taken from +5V, even 10A @ +3.3V = 17W of power dissipated [through the power transistor]! But if it's taken from a transformer tap that ultimately measures closer to +4V before it's stepped down, only 7W would be dissipated at the same load)... I guess the only way to find out for sure would be to unwind the transformer (without checking the rectified output through other means), but I don't know of anyone who's done that. There does 'appear' to be a separate transformer winding in the transformers, in those Hipro PSUs, but there appears to be one in these Thermaltakes as well, which is why I ask. I don't mean to derail the thread by asking.

[momaka]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

If there is a separate tap, then I would guess it's lower than 5V in order to increase efficiency.
You don't really have to unwind the transformer to check the voltage - just solder a piece of wire to the rectified rail and measure voltage with respect to ground when PSU is on.

I'm lazy, so I never gotten around to doing that.

[Wester547]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

Thanks for the advice. But I believe you once said in the another thread (I believe another one regarding a Thermaltake power supply, since that's the original topic at hand) that with the independent method, the DC output is usually higher than +3.3V and maybe even +4V. Coupled with the observation that there seems to be another tap in the transformer, I thought it was more or less stepped down from +4V. I guess that's possible, I know the magamp is used as a buck regulator of sorts to step down the higher voltages in the transformer taps after they've been rectified, for the other rails, besides cleaning up the ripple current... so I wonder if the magamp is used to any extent in the independent design, in terms of regulation, before the MOSFET finally regulates it down to +3.3V.

Another Thermaltake, a rather different unit, was reviewed several years back at Hardware Secrets. It also uses linear regulation to generate +3.3V and its efficiency never went below 70% (the +3.3V rail was loaded to 16A in the last test... it managed 350W with good voltage regulation but shut down at 430W), so maybe there is another tap.... but that could just be a byproduct of a higher duty cycle at lower loads (higher duty cycle = less conduction losses and higher efficiency). Or it could be the fact that it uses a FET with a very good ON resistance rating to regulate the voltage down to +3.3V, or maybe just that it has both an exhaust and intake 80mm fan for augmented airflow. That power supply performed very poorly otherwise, though, especially for a unit that appears to be single switch forward... the ripple voltage on all rails, especially as the load increased, quickly started touching the ATX spec limits and beyond. Something is definitely amuck with it.

I believe Pentium4 has a very similar PSU and stated that it only has a single 2200uF 16V Teapo SC capacitor and no ferrite coil for the +12V rail, so it doesn't really have a proper or complete output filter.... that might be why... yet in the review, the +12V ripple voltage jumped to 680mV when they try to overload it even a bit. Then again, the +3.3V rail is shown to be outputting an unusual amount of ripple voltage in the review, so it might not even be linear regulation in these PSUs. Goodpsusearch's PSU looks to be similar in terms of light output filtering.

[senz_90]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

geez. i think this teapo 680uF attach to my psu too!!

unfortunately, i don't have caps meter to measure it.

[goodpsusearch]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

This power supply belongs to me now and I am planning to do a recap.

I was able to read the marking on the main mosfet. The Arctic psu uses an STP10NK60Z (TO-220 package) in a single transistor forward configuration

Drain Current (continuous) at TC=25°C ID=10 A
Drain Current (continuous) at TC=100°C ID=5.7 A
Drain Current (pulsed) ID=36 A
RDS(on)=0.65Ω
Total Dissipation at TC = 25°C Ptot=115W

I have some STW9NB90 (TO-247 package, RDS(on)=0.85Ω) available from dead Antec SmartPower psus.

Is it worth to replace STP10NK60Z with STW9NB90 or should I order something better from http://gr.rsdelivers.com/?

[kc8adu]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

lower rdson is better

[c_hegge]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

I thought you said earlier that the PSU worked. If the switcher isn't blown, then I wouldn't worry about replacing it. I would just recap the PSU and keep it as a 350 watter. It would do fine in a basic home or office PC.

[goodpsusearch]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

The psu works but I thought I could get some extra juice from it if I replace the TO-220 STP10NK60Z mosfet with something better as Th3_uN1Qu3 suggested.

For example this psu you have reviewed was able to output more than 400 watt using 1x FQA10N80 MOSFET (RDS(on)=1.1Ω) in a single transistor forward topology.

I am also going to replace the fake Teapos with real 680uF primary capacitors.

[Heihachi_73]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

The Teapos I have of that shape are blue 560uF SK series dated 07/03, and are still in spec at 499uF and 505uF respectively.

Get rid of anything which has "VENT" or "TEAPO" on it and your trouble should go away.

[goodpsusearch]

Re: Fake Teapo 680uF primary caps-470uF real capacity (Arctic A-430BX)

I partially recapped that HEC OEM power supply. I used Panasonic, Sanyo, Nichicon, Rubycon and Teapo capacitors.

I also tweaked the fan controller. The psu still remains one of the most silent I own and it doesn't overheat.

I replaced the fake 680uF primary caps with real 680uf caps (measured capacity 725uF & 690uF).

I replaced STP10NK60Z with STW12NK80Z. More info about that operation here:
https://www.badcaps.net/forum/showthread.php?t=45884

Do you think it can do 350W without blowing?