Part Analysis
| General Data | |
| Manufacturer (OEM) | Seasonic |
| PCB Type | Double-Sided |
| Primary Side | |
| Transient Filter | 6x Y caps, 3x X caps, 2x CM chokes, 1x MOV |
| Inrush Protection | 2x NTC Thermistor MF72-20D20M (20 Ohm) & Relay |
| Rectifier FETs |
4x IPB60R040C7 600V ,1x LL25XB60 25A 600V
|
| APFC MOSFETs |
4x Infineon IPA60R099P6 (600V, 24A @ 100°C, Rds(on): 0.099Ohm)
|
| APFC Boost Diode |
2x CREE C6D10065A (650V, 10A @ 155°C)
|
| Bulk Cap(s) |
3x Nippon Chemi-Con (420V, 820uF each or 2460uF combined, 2,000h @ 105°C, KHE)
|
| Main Switchers |
4x Infineon IPA60R080P7 (600V, 23A @ 100°C, Rds(on): 0.08Ohm)
|
| Drivers IC | 2x Silicon Labs Si8233BD |
| APFC Controller | Texas Instruments UCD28070 |
| Resonant Controller | Champion CM6901T2X |
| Topology |
Primary side: Bridgeless, Interleaved PFC, Full-Bridge & LLC converter
Secondary side: Synchronous Rectification & DC-DC converters |
| Secondary Side | |
| +12V MOSFETs | 16x Nexperia PSMN1R0-40YLD (40V, 198A @ 100°C, Rds(on): 1.93mOhm) |
| 5V & 3.3V | DC-DC Converters |
| Filtering Capacitors | Electrolytic: 5x Nippon Chemi-Con (105°C, W), 1x Nippon Chemi-Con (5-6,000h @ 105°C, KZH), 1x Nippon Chemi-Con (2-5,000h @ 105°C, KZE), 3x Rubycon (6-10,000h @ 105°C, ZLH), 1x Rubycon (3-6,000h @ 105°C, YXG) Polymer: 28x Nippon Chemi-Con, 2x FPCAP, 8x Evercon |
| Supervisor IC | Weltrend WT7527RA (OCP, OVP, UVP, SCP, PG) |
| Fan Controller | Nuvoton M031 |
| Fan Model | Noctua NF-A12x25 (120mm, 12V, 0.14A, Hydraulic Bearing Fan) |
| 5VSB Circuit | |
| Rectifier |
1x Infineon BSC100N06LS3 FET (60V, 36A @ 100°C, Rds(on): 10mOhm)
|
| Standby PWM Controller | Power Integrations INN3164C |
No changes were made to the main PCB compared to the “plain” TX-1600. Only the fan and the top cover are different. Despite the huge PCB, the platform is densely populated. Moreover, I usually don’t find large heatsinks in modern PSUs. Although Titanium efficiency allows for low thermal loads, Seasonic used large heatsinks to make the fan’s noise as low as possible. A bridgeless, interleaved PFC is used to achieve high efficiency, along with a full-bridge topology and an LLC resonant converter.
The 12V rail is generated through 16x powerful Nexperia FETs. The same rail also feeds a pair of DC-DC converters, which generate the minor rails. Fully desoldering this platform makes it easier to identify all its parts, but I didn’t want to do that since I need it to remain fully operational for future testing.
The build quality is top, and the same goes for the soldering quality. Seasonic used Japanese caps everywhere, both electrolytic and polymer, and the bulk caps have a vast combined capacity, close to 2500uF! Because of the large bulk caps and their combined capacity, not one but two NTC thermistors had to be used, with 20 Ohm resistance each! Despite the large PCB, space is still an issue, so both are installed onto a vertical board.
The cooling fan is by Noctua, a high-speed version of the NF-A12x25, which is not commercially available. You can find the full review of this fan here:
I just bought this for an upcoming build
I wanted a Titanium PSU.
I wanted the high quality Prime goodness.
I wanted a Seasonic specifically to match all my other PSUs (all Prime Titaniums) that share Seasonic’s cable compatibility.
I will be powering an RTX 5090 when it’s released. I was going to go with the TX-1300, but I wanted ATX 3.1 (specifically the new connector).
My UPS can only do 900w, which should be enough for the system (hopefully?). I really wish they made a TX-1000 ATX 3.1 Prime. I bought more PSU than I need, or will conceivably ever need. Certainly more than my UPS can support.
NZXT C1500 or Seasonic Noctua TX-1600?
depends on the price
Which would you choose for your rig?
Hello!
quick question, as I could not find this info anywhere on the internet:
the 12v-2×6 cable connectors- are they H+ or H++ labeled?
Asking because my PX-2200 12v-2×6 are labeled H+(the 2 with 2×6 at both ends) for 2 of them, and H++ for the other 2(the ones with 2x8pin at one end, and 2×6 at the other).
H++ normally
Thanks a lot for your reviews! I am searching for a new PSU after my 8 years old Seasonic XM2 1250w fan started making weird noises and I thought it’s time for a new PSU since it’s more than 8 years old now anyway.
I found the PX-1600 ATX 3.0 PSU locally in my country at a reasonable price ($480, and that’s a good price considering the high taxes here), do you think I should wait for this PSU to come to my region? it will probably be ~120$ more than the PX but if it’s much better, I can afford to pay $120 more since I replace a PSU almost every decade.
PX-1600 is fine! Go for it, no reason to wait or pay more for the TX-1600
That’s really good to know! I will buy it tomorrow.
ATX 3.0 vs ATX 3.1 shouldn’t really matter because I just read some of your previous reviews and you said the changes are mainly only in the plug and not on the cable, and since new GPUs are coming with the new plug it’s not a problem. The cable won’t be moving on the PSU side and once it’s plugged firmly it will be fine.
Got the PX-1600 ATX 3.0 today and installed it and wow it really has huge dimensions, and even though I was prepared I still was shocked when I saw it in person haha, it barely fit inside my Lian Li O11 Dynamic XL.
It really feels high quality! it’s dead silent and the fan didn’t even spin at all at ~600w load, it only started spinning on a low speed when my watt meter (connected to the wall socket) hit 640-650w
I’m torn between this and the AX1600i. Is the ATX 3.1 spec and the new cabling worth the lower efficiency and ripple suppression vs Corsair?
it is a tough choice, indeed.
I have the exact same problem here 😔
If you would have to decide, which one would you choose and why? I’m leaning towards AX1600i because it is fully digital and allows for software monitoring. But could the absence of the ATX 3.x be a problem in the future? Thanks
I’ve just ordered the Seasonic Prime Noctua TX-1600, despite the excellent performance AX is not ATX 3.x compliant.
The EU price is actually pretty great relatively. The regular Seasonic TX-1600 retails for 500€ too and it has older 12VHPWR cables, ATX 3.0, PCIe 5.0, and a louder fan.
How can it be quieter at 1440 rpm at 100% load versus the standard version and 980 rpm? Even it 120mm fan.
Different fan, bearing, and top cover can affect noise output notably!
https://ibb.co/8dW4pNy
https://ibb.co/37ysWM0
Similarities with the PC desktop graphics or processor market, which has massive stagnation due to illegal collusion, price market fixing etc. since this 2006 disclosure.
https://www.tomshardware.com/news/nvidia-amd-ati-graphics,6311.html
Especially since 2011 (CPU market intel amd) and 20180 gpu market (nvidia amd). All these corpos doing a farce of marketing pretending to be competitors, but in reality friends who scam customers to maximise profits for many years.
It’s a duopoloy (intel amd) and Stackelberg duopoly (nvidia amd): If there is no serious threat from the other duopolist (e.g. intel 2011 – 2017), or now nvidia dominitating graphics space since 2018, the execs of these corpos just collude to make their products bad on purpose or rise prices by + 100 – 200 %.
Want acutal meaningful improvements with a new graphics card in 2018 – 2024? Buy stupidly overpriced 2080 Ti, 3090, 4090, 5090. Everything below is heavly castrated.
Even AMD in PC diy is selling frigging 8-core cpus for 500 bucks in 2024. Same as intel execs did 2011 – 2017 with 4-cores for 350 bucks.
Similar with psus now: If there isn’t a new technolgy standard coming out (ATX 3.0 now), it’s stagnation for many years.
Have to buy psus north of 1000 watts for > 400 bucks (which hardly any desktop consumer needs) to get some advancements.
Rest in the 100 – 200 price range is stuck in cybenetics silver – platinum ratings, noise fans etc. for years.
Remember seasonic booth in 2018 at that computex, showing an engineering prototype of psu achieving 96 % efficiency-load during the whole 1 – 100 % utilization rate? “Ready in 2 -3 years” they said.
Never heard a single word of again.
I wonder why they decided to go with 120mm fan. The SeaSonic PRIME PX-1600 can easily be modified to use Noctua NF-A14x25r G2, which would perform better. The mod is cheaper than buying the Noctua PSU as well.
Thank you for the through review. Did you test for EMF? I don’t recall seeing it in the review.
EMI? Nope.
I’ve seen you mention EMI in previous reviews. Is it no longer part of your testing procedures?
it’s up there with the 1kW (or more) passive PSU on the: impressive feat, but what for? list
with such a high load components you use absolutely will make tons of noise, even the beast won’t be able to handle them, so… it’s beyond flagship, it’s just a showoff
unless some users use custom water cooling with silent fans and huge radiators.
PC cooling pumps aren’t in the same ballpark as A12x25 at 600RPM when it comes to silence and I doubt there are other industries that would make them quieter than that
I would love them to exist, that would allow for some extra headroom, but for now we’re stuck with Asus x Noctua cards with reduced power and D15G2 (or some 120mm dual tower fanswapped to A12x25 or T30)
this would probably get around 500-600W in total, which is a lot of power, but you could power two machines like that with one PSU and still stay below these 600RPM
that’s impressive engineering feat for sure, but impractical
well, maybe fan reliability justifies it