jonnyGURU Forums

jonnyGURU Forums (http://www.jonnyguru.com/forums/index.php)
-   PC Power Supply Discussion (http://www.jonnyguru.com/forums/forumdisplay.php?f=3)
-   -   Seasonic Prime 850W Review (http://www.jonnyguru.com/forums/showthread.php?t=13643)

Orion 09-22-2016 09:01 AM

Quote:

Originally Posted by Journeyman (Post 130292)
Hee hee, already removed (cowards). :D

Huh. Why'd they do that?

ITelektro 09-23-2016 03:11 AM

Quote:

Originally Posted by Oklahoma Wolf (Post 130229)
No, I think those spikes are not important.

Oh, yes, those
Do not understand me wrong, I completely agree. Probably this is a random spikes of high frequency which does not generated in the power supply. The higher the frequency, the less important. But, each spike within 20Mhz must be include, no matter what we think.

Quote:

As for the fan and soldering, I can't really score units I don't have.
I think all Primes 850w have the same fans.

And finally, the idea of scoring a turn-on analysis. Not quite true that this result meets Intel's requirements or that is within specification.

There must be a smooth and continuous ramp of each DC output voltage from 10% to 90% of its final set point within the regulation band, while loaded as specified.

IMO, also think that it should be just pass / fail evaluation. But if you want to introduce a rating you should analyze in sections and give an average rating for the turn-on time, rise time, overshoot, wave form.
A lot of work, it may be better to spend time at hold-up time and / or inrush current.

Quote:

Originally Posted by Journeyman (Post 130292)
Hee hee, already removed (cowards). :D

:D
Quote:

I can understand your irony, but hell yeah... Every watt counts :p
In this case, literally:D

Oklahoma Wolf 09-23-2016 11:44 AM

Quote:

Originally Posted by ITelektro (Post 130329)
There must be a smooth and continuous ramp of each DC output voltage from 10% to 90% of its final set point within the regulation band, while loaded as specified.

The wording is vague on that point, so a lot of OEMs likely get to a certain result, say it's good enough that it could be considered smooth and continuous, and let it go at that.

If I start scoring on those tests, this is exactly where I would want to do it. If, that is, I can think of a good way to get it into the point deductions.

ITelektro 09-23-2016 02:57 PM

Quote:

Originally Posted by Oklahoma Wolf (Post 130351)
The wording is vague on that point, so a lot of OEMs likely get to a certain result, say it's good enough that it could be considered smooth and continuous, and let it go at that.

I did not quoted the whole text, below is exactly define vague part.
There must be a smooth and continuous ramp of each DC output voltage from 10% to 90% of its final set point within the regulation band, while loaded as specified. The smooth turn-on requires that, during the 10% to 90% portion of the rise time, the slope of the turn-on waveform must be positive and have a value of between 0 V/ms and [Vout, nominal / 0.1] V/ms. Also, for any 5 ms segment of the 10% to 90% rise time waveform, a straight line drawn between the end points of the waveform segment must have a slope ≥[Vout, nominal / 20] V/ms.

Although on the modern oscilloscopes this is relatively easy to do that, but it is lot of work.

According to this, your test sample does not meet Intel specification!?
On my sample results were considerably better, not perfect, but better.

Seasonic Rep 09-24-2016 06:03 AM

Quote:

Originally Posted by ehume (Post 130191)

@Seasonic Rep: the one thing I wish you included was a PSU turn-on socket. Your support page suggests we test the PSU first by grounding the Power-Good line. But female 24-hole ATX sockets for motherboards are available with the Power-Good and Ground positions pre-connected. I have this adapter -- I bought one from a shop. But a PSU maker includes them routinely. I'll bet one intercepted return would pay for including them routinely.

Hi Ehume,

Thank you for the input, yes, it does seem to be a good idea. I will pass this to our PM at HQ and check on the feasibility.

Seasonic Rep 09-24-2016 06:36 AM

Quote:

Originally Posted by ITelektro (Post 130220)
I tested this power supply for the one web portal in Croatia. Plenty complaints went to the quality of the fan. Double lower life expectancy than the rest of the device. What is your opinion?
Thank you for your answers.

On my copy I found a couple of really badly soldered points. According to the experience would say that someone manually soldered. I'm glad to see that it was only in my copy. In the copy which is jonny test, soldering is flawless.

Hi ITelektro,

Can you please provide the following information so we can follow up and improve the quality of our products.
> PRIME Model Number
> Serial Number
> For which Croatian portal did you make the testing
> Where are your test results posted
> Link to the feedback / complaints about the fan
> Photo of the "really badly soldered joints"

This is very important to us because we want to provide the information to our QA so we can implement change and improvement.

SOLDERING: From time to time, hand soldering is required for either touch up or component change but we try our best to maintain the highest level of workmanship. Please provide a photo so we can document it and make improvements.

FDB FAN: As for the FDB fan, this is one of the better fans we have found so I am a bit surprised when you write... "Plenty complaints went to the quality of the fan."

Also I am not sure if I understand what you mean by Double lower life expectancy.... Considering the fan is the only moving mechanical part in the whole PSU, it will have the lowest component MTBF. What would you recommend as an alternative?

Please kindly reply to the requested information so we can make improvements where applicable.

Thank you.

quest for silence 09-24-2016 07:20 AM

Quote:

Originally Posted by Seasonic Rep (Post 130409)
FDB FAN: As for the FDB fan, this is one of the better fans we have found

Just as a side note: that doesn't mean it's enough for such a flagship unit.

It might worth to note that the Seasonic brand kindles the highest expectations (it's one of your marketing goal, isn't it?), so maybe that's a fan which we might expect by someone like Thermaltake, but not exactly by you.

In this respect I agree with those (the russian review now disappeared? I don't recall, atm) who wrote that the previous Sanyo Denki SanAces (or even the Protechnic MGA13512XF-025 you mounted on the XP-based CM Vanguard series) were probably more in line with the Seasonic status and track, than the current Hongh Hua one (too often seen, even on cheaper PSUs).

ITelektro 09-24-2016 08:41 AM

Quote:

Originally Posted by Seasonic Rep (Post 130409)
Hi ITelektro,
Can you please provide the following information so we can follow up and improve the quality of our products.

I appreciate your approach.
Quote:

PRIME Model Number
; SSR-850TD (Prime 850 Titanium)
Quote:

S/N
:R1606TA106460003
Quote:

For which Croatian portal did you make the testing
www.pcekspert.com
Quote:

Where are your test results posted
http://www.pcekspert.com/clanak/seas...titanium-test/
Photos are available on my website. These are out our selection images, there are more, I will gladly share them if these are not enough.
Quote:

Link to the feedback / complaints about the fan
Complaints about the fan are not by me, I'm not sure how much are justified. Therefore, I ask for an opinion.
These are reactions to the test, used google translate;
http://forum.pcekspert.com/showthread.php?t=280892
Recommendation was SanAce.
Photo of the "really badly soldered joints"
http://www.itelektronika.com/blog-se...owall=&start=8
As I wrote, I think it will not be a problem. In all the available tests soldering is excellent. My copy is the exception. I understand that sometimes it is necessary to hand soldering but clean the flux residues and optical check the connection.
My opinion;
This is the best power supply according performance that has been with me on my workbench. Price of the Prime is quite high in Cro (about 350USD) so I do not hide that I'm looking for the tiniest faults and asking for the opinion of others in order to be as objective as possible.
I like criticism, and if I made a mistake somewhere, I expect criticism with gratitude.
Sorry for my bad English

Seasonic Rep 09-24-2016 08:45 AM

quest for silence,

The RU review is... under review because we are checking test parameters.

SanyoDenki does not make a 13 or 13.5 or 14cm fans with 25mm high frame, not that we know of anyway... If we continue with the 12cm, we might not hear the end of it... As well, the SD fans are optimized for reliability and life time and not necessarily noise. This is similar for Papst or Nidec or Panaflo/NMB... all great fans but not necessarily optimized for low noise. From the production side, we also have to consider lead time, delivery, etc. but noise and reliability are our major concerns.

I will ask our PM about the Protechnic, but he did tell me that the Hong Hua is an excellent FDB fan and seeing it in lessor PSUs does not make it a lessor fan.

quest for silence 09-24-2016 10:56 AM

Quote:

Originally Posted by Seasonic Rep (Post 130419)
the SD fans are optimized for reliability and life time and not necessarily noise. This is similar for Papst or Nidec or Panaflo/NMB... all great fans but not necessarily optimized for low noise. From the production side, we also have to consider lead time, delivery, etc. but noise and reliability are our major concerns.

I'm pretty sure there are supply chain related reasons behind the Hong Hua choice, but from my point of view there are too many well appreciated fans in the market to plainly accept a plain vanillla HH (and not an individually balanced SanAce) on your Prime Titanium.

And, about noise, IIRC you implemented the same HH "real" FDB onto the Antec Edge units, and the general consensus is that it was a badly sounding unit.

As far as I understand, optimized for low noise mostly means "not a ball bearing, please": as a matter of fact, even the expensive Be Quiet Silent Wings 3 (like the BQ SIW3-13525-HF) or Noctua iPPC (like the A14) can be noisy (they are rated up to 2.1/3k rpm, Aris recorded the first up to 47dB IIRC), but I won't call them as noisy when properly operated, neither unreliable. But I guess you already know that far better than me.


Quote:

Originally Posted by Seasonic Rep (Post 130419)
I will ask our PM about the Protechnic, but he did tell me that the Hong Hua is an excellent FDB fan and seeing it in lessor PSUs does not make it a lessor fan.

If I recall correctly what Phaedrus wrote when he worked for CM, that Protechnic should sport the latest Matsushita FDB gen. 2 bearing: in case, it could be a viable candidate for any high end PSU.

But above all, with all the due respect, sir, your one looks like sort of a moot point.

The more (money) you ask, the more demanding the people are: said differently, what if there were non japanese electrolytic caps inside a Prime Titanium? Deep down lots of experts here wrote how good some Samxon, Teapo, Capxon (!), Taicon (!!), Ltec (!!!), Elite (!!!!), JunFu (!!!!!) series may be, and lots of good engineers (think to Delta's ones) somehow agree with them using those caps.

What's the point of a flagship Prime Titanium over a high end units like Corsair RMx or EVGA P2? Up to twice the price for what? It should give us some tidbits, not just some tiny bits of more performance, I guess.

When more than fifteen years ago the Jaguar S-type sedan come out lots of people complained about the cheap Ford small parts in it... what's with Jaguar now? Look at how much the Prime Titanium costs (outside USA, of course)...


All times are GMT -4. The time now is 02:50 PM.

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Copyright © JonnyGURU.com 2000 -