Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 33

Thread: EVGA 450 BT PSU Review: Amazing Value At $25 @ Tom's Hardware

  1. #21
    Join Date
    Nov 2015
    Location
    U.S.A.
    Posts
    1,264
    Thanks Thanks Given 
    327
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by Samueru Sama View Post
    Are you implying that the 450 bt could kill a pc?
    Every PSU can, it's only a matter of probability and what probabilities you find acceptable. I'd think even a protection IC itself can fail. And there isn't a backup.

    Stefan, if I recall the fuse didn't even blow on the B3's, so it's likely in my mind that perhaps the B3 failure could also cause damage to the PC.

    For the EVGA 450BT, yes I would be fine with risking it killing my whole PC with low voltage. I'd rather have that actually than something exploding; not that I find it likely but I'd never want a spark to go flying and causing a house fire

    I suppose if the whole PC gets killed it's just a situation of bad luck. It happens.

  2. #22
    Join Date
    Mar 2018
    Location
    Maracaibo, Venezuela
    Posts
    59
    Thanks Thanks Given 
    25
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by Tator Tot View Post
    Don't dismiss the cost of those components. Likewise, you need to keep in mind that SuperFlower will need to pay the contractor who built the unit, and also make a profit themselves.
    Well since there are PSUs on the market for 20$ that have a fan, wiring, transformer and coils (obviously not the best parts) I see plenty of margin there.

    And this unit isn't specially cutting corners, -12V has more filtering than other psus do on the 12 rail, 5vsb has a rectifier more powerful than other psus do on their 12V rail. Heck some components on the primary side are even better than the ones used on 750 B3.

    It is completely possible to make profit out of a good unit for 25$.

    Quote Originally Posted by turkey3_scratch View Post
    Every PSU can, it's only a matter of probability and what probabilities you find acceptable. I'd think even a protection IC itself can fail. And there isn't a backup.

    Stefan, if I recall the fuse didn't even blow on the B3's, so it's likely in my mind that perhaps the B3 failure could also cause damage to the PC.

    For the EVGA 450BT, yes I would be fine with risking it killing my whole PC with low voltage. I'd rather have that actually than something exploding; not that I find it likely but I'd never want a spark to go flying and causing a house fire

    I suppose if the whole PC gets killed it's just a situation of bad luck. It happens.
    I work repairing electronic stuff, and I've fixed several ATXs and laptop PSUs, to this date I haven't gotten a single unit that has killed a PC, or any part of it.

    I would say that it is as unlikely as one being hit by lighting.

    Seriously think of it, since switching is done at the primary side any failure there could at worst just destroy the primary side. If the secondary fails the capacitors will short the output of the unit.

    The worst could be that the protection IC is damaged by default in a way that it still doesn't shut the unit and the 3.3V mag amp transistor fails open (Putting 5V on 3.3V).

    Other would be that the feedback opens (next to impossible), disable the protection IC and then power unit to get +30V on the output. Which btw, the 450 BT has some sort of open feedback protection at its primary because I accidentally discovered it while trying to fix the 450 BT that I had that died (I fixed again btw, I'm waiting for some halogen lamps to test it properly and post the results).

    Other could be overshoot issues in the feedback, I've seen several units reviewed here having that issue, however for that to be considered an issue the overshoot would need to be test with at least 100 uF of extra capacitance, motherboards have a lot of it. Actually I wonder if PSU reviews sites add the 10 uF + 0.1 uF ceramic cap to their noise measurements as ATX12V specifies.

    Obviously I'm talking the average group regulated PSU, not the DC-DC + Active rectification ones, on those the protection IC has to really work, specially given how sensitive mosfets are to esd.

    Actually how do people even determine that a PSU killed something and not the other way around? For example If you motherboard and CPU die it is very likely that the problem was a shorted buck converter on the motherboard.

  3. #23
    Join Date
    Jul 2009
    Location
    Germany
    Posts
    3,498
    Thanks Thanks Given 
    6
    Thanks Thanks Received 
    31
    Thanked in
    28 Posts

    Default

    Quote Originally Posted by Samueru Sama View Post
    Are you implying that the 450 bt could kill a pc?
    I'm implying that Protections are important and can prevent unnecessary damage to the components due to a fault situation...

    If the PSU doesn't switch off, its very very bad.

    Just look at this Review:
    https://www.hardwareluxx.de/index.ph...t.html?start=3

    You might want to use a Translator because its in German!
    But its pretty damning because they were able to pull 100A at 6V off the PSU.
    And Ripple/Noise goes through the roof at around 60A.

    In short:
    Due to missing protection it doesn't switch off or die under overload conditions wich will cause damage to the components.

    With the EVGA B3 it just dies. Voltages don't go out of spec before that, it just dies. You can argue that that is better than not dying because of the voltages staying in spec...

  4. #24
    Join Date
    Mar 2018
    Location
    Maracaibo, Venezuela
    Posts
    59
    Thanks Thanks Given 
    25
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by Stefan Payne View Post
    I'm implying that Protections are important and can prevent unnecessary damage to the components due to a fault situation...

    If the PSU doesn't switch off, its very very bad.

    Just look at this Review:
    https://www.hardwareluxx.de/index.ph...t.html?start=3

    You might want to use a Translator because its in German!
    But its pretty damning because they were able to pull 100A at 6V off the PSU.
    And Ripple/Noise goes through the roof at around 60A.

    In short:
    Due to missing protection it doesn't switch off or die under overload conditions wich will cause damage to the components.

    With the EVGA B3 it just dies. Voltages don't go out of spec before that, it just dies. You can argue that that is better than not dying because of the voltages staying in spec...
    I would say the pretty damaging part would be ripple and noise going to the roof, if the pc somehow doesn't crash by that low input voltage (the hdd motor) the board regulators will just increase the duty cycle to compensate for the low voltage, this will in turn pull more current of the PSU, until the primary or secondary dies.

  5. #25
    Join Date
    Jul 2009
    Location
    Germany
    Posts
    3,498
    Thanks Thanks Given 
    6
    Thanks Thanks Received 
    31
    Thanked in
    28 Posts

    Default

    Quote Originally Posted by Samueru Sama View Post
    I work repairing electronic stuff, and I've fixed several ATXs and laptop PSUs, to this date I haven't gotten a single unit that has killed a PC, or any part of it.
    You don't know that, because if something dies, you usually don't know what caused it.
    But there are some instances where a PSU could be the cause of components dying prematurely because of bad voltages.

    It just takes time to see that. One possibility is that Hard Drives die rather quickly. But also other things are possible.

    Its not that easy sadly...

    Quote Originally Posted by Samueru Sama View Post
    I would say that it is as unlikely as one being hit by lighting.
    I'd say that it depends on the PSU, protections and Situation.
    If you don't have not working protection and a Problem, it certainly can kill components.

    And there are instances where a PSU with not working protection can make a fault worse than it already is.

    There is a Thread at OC.com "why Single Rail is NOT Better than Multi Rail", but the Pictures are gone thants Photobucket...

    In this case a high Side MOSFET shorted and that caused a pretty nice fire...

    Quote Originally Posted by Samueru Sama View Post
    Obviously I'm talking the average group regulated PSU, not the DC-DC + Active rectification ones, on those the protection IC has to really work, specially given how sensitive mosfets are to esd.
    You are missing high ripple (in the Volt area like 1-2V Ripple) either due to failed capacitors or due to overload and not working Protection.

    That is what was also mentioned in the Review I've linked, though he did not say how high it was....

    Quote Originally Posted by Samueru Sama View Post
    Actually how do people even determine that a PSU killed something and not the other way around? For example If you motherboard and CPU die it is very likely that the problem was a shorted buck converter on the motherboard.
    In some cases you can give an educated guess, in others the PSU dies and other components are dead too.

    But sometimes its just fantasy...

  6. #26
    Join Date
    Mar 2018
    Location
    Maracaibo, Venezuela
    Posts
    59
    Thanks Thanks Given 
    25
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    I'm not missing ripple, I mentioned the overshoot part and I said right there that I wonder if PSU reviews sites add the 10 uF + 0.1 uF ceramic cap to their noise measurements as ATX12V specifies. And also told you about it when talking about the german review.

    Ripple could be an issue, but to determine if it is an issue for the PC it needs to be measured with the capacitance the board adds. Which many times does help a lot, IIRC here I read of a crappy PSU that literary had no capacitors on the 12V, it relied on board capacitance. xd

    Also I meant dead PSU with dead CPU and board at the end, if the board fets fail short the CPU will have full 12V (and full short circuit in other words) which could kill the PSU, specially given how very cheap PSUs are prompt to primary side failure when overloaded.
    Last edited by Samueru Sama; 06-05-2018 at 12:50 AM.

  7. #27
    Join Date
    May 2010
    Location
    Missouri; USA
    Posts
    5,054
    Thanks Thanks Given 
    410
    Thanks Thanks Received 
    14
    Thanked in
    9 Posts

    Default

    Quote Originally Posted by Samueru Sama View Post
    Well since there are PSUs on the market for 20$ that have a fan, wiring, transformer and coils (obviously not the best parts) I see plenty of margin there.
    You don't see any margin though, because you don't know the difference in cost between the components. You're just assuming that the absolute garbage has the same BoM as this thing.

    Quote Originally Posted by Samueru Sama;149689[B
    And this unit isn't specially cutting corners[/B], -12V has more filtering than other psus do on the 12 rail, 5vsb has a rectifier more powerful than other psus do on their 12V rail. Heck some components on the primary side are even better than the ones used on 750 B3.

    Which is what makes me skeptical of EVGA's long term plans. It's a half decent design that's heavily antiquated by today's standards and is being sold through retail partners at such a low price that it seems questionable if the company will actually be able to continue long term production and support of the product.

    We have already witnessed EVGA selling their other more premium products at a loss to gain marketshare. So it is not as if this is a new move by them.



    Quote Originally Posted by Samueru Sama;149689It is completely possible to make profit out of a [U
    good[/U] unit for 25$.
    Possible, but not likely in this case. The MSRP is $45; and this is a $20 discount ontop of that price. With a retail middle man in there as well.

  8. #28
    Join Date
    Mar 2018
    Location
    Maracaibo, Venezuela
    Posts
    59
    Thanks Thanks Given 
    25
    Thanks Thanks Received 
    0
    Thanked in
    0 Posts

    Default

    Quote Originally Posted by Tator Tot View Post
    You don't see any margin though, because you don't know the difference in cost between the components. You're just assuming that the absolute garbage has the same BoM as this thing.

    I said right there: (obviously not the best parts)

    Anyway, lets keep doing it, remember we were at 11$:

    Transformer: https://katalog.we-online.de/pcd/datasheet/070-5783.pdf

    That one is a 500W transformer (on average) (depends on frequency and topology).

    https://www.alibaba.com/product-deta...160037e9GeYeBb

    0.59$ Lets assume it is a piece a crap, so lets double the price for a proper one, 1.2$.

    Common mode coils, you get a dozen for 1.5$, lets add 0.4$.

    The F3AA012E costs 0.85$ on mouser.

    Wiring, here the meter of 18 gauge wire costs 0.1$, the 24 pin atx connector measures 55 cm, so 55 cm * 23 = 12.65 meters (23 because there's no -5V).

    PCIE Power cable measures, 65 cm, 65 cm * 6 = 3.9 meters.

    CPU Connector, 62 cm * 4 = 2.48 meters.

    Two sata cables 60 cm, that's (60 cm * 5) * 2 = 6 meters.

    One molex of 80 cm: 80 cm * 4 = 3.2 meter.

    That gives 28.23 meters of wire -> 2.8$.

    (And this unit doesn't use 18 gauge for all its wires btw).

    Fan, can't find any globe one at whole sale, the Chinese ball bearing ones cost 2$: So 2$.

    That's 7.25$, add the previous 11$ and that gives 18.25$.




    Quote Originally Posted by Tator Tot View Post
    Which is what makes me skeptical of EVGA's long term plans. It's a half decent design that's heavily antiquated by today's standards and is being sold through retail partners at such a low price that it seems questionable if the company will actually be able to continue long term production and support of the product.


    We have already witnessed EVGA selling their other more premium products at a loss to gain marketshare. So it is not as if this is a new move by them.
    Yeah, that was when it was for 10$.


    Quote Originally Posted by Tator Tot View Post
    Possible, but not likely in this case. The MSRP is $45; and this is a $20 discount ontop of that price. With a retail middle man in there as well.
    Well at least you've dropped insane from it.

    Remember that those prices I gave you are mostly from an US supplier, they take their cut as well. And some parts like the caps are more expensive since I can't find the teapos on mouser, also didn't bother to determine which wires are 20 gauge which are 18 gauge.

  9. #29
    Join Date
    May 2010
    Location
    Missouri; USA
    Posts
    5,054
    Thanks Thanks Given 
    410
    Thanks Thanks Received 
    14
    Thanked in
    9 Posts

    Default

    Quote Originally Posted by Samueru Sama View Post
    That's 7.25$, add the previous 11$ and that gives 18.25$.

    Remember that those prices I gave you are mostly from an US supplier, they take their cut as well. And some parts like the caps are more expensive since I can't find the teapos on mouser, also didn't bother to determine which wires are 20 gauge which are 18 gauge.

    You just further illustrate the point. Even assuming you take a 30% markup off the top of that parts cost, you're left with a basic BoM of 12.775; but for the ease of math lets just say $12.75.

    That's just a basic materials cost and doesn't cover the assembly, QC, testing, etc that has to go on at a factory. We also know that SuperFlower is outsourcing a lot of their work, so you have to keep in mind that you've actually added in two middle men for profit sharing to this equation.



    So again, with a $12.75 BoM, if we assume 30% margin on each middle man:
    SuperFlower 12.75 x 30% margin = 18.21
    Contract Manufacturer 18.21 x 30% margin = 26.01
    Retailer 26.01 x 30% margin = 37.16


    Given that the MSRP is $44.99, that ~$37-38 USD cost seems relatively accurate as if you add on 30% to that then you're just around $50 USD so the actual margins before this step may be a bit lower but we're definitely in the right ballpark.



    If we wanted $25 to be profitable then the margins would need to be as low as 10% at each step which would give a cost to retailer of 21.59 and with another 10% margin from them a total cost to end user of $23.99.

  10. #30
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    1,832
    Thanks Thanks Given 
    96
    Thanks Thanks Received 
    31
    Thanked in
    29 Posts

    Default

    Quote Originally Posted by Tator Tot View Post
    At $25 they're probably selling the unit way below cost which makes me question EVGA's long term plans in the market. A warranty means nothing if the company goes belly up and a very low end build like this does not inspire confidence.

    Sometimes companies flood the market at a loss to damage a competitor.

Similar Threads

  1. EVGA 850 B3 review @ Tom's Hardware
    By Orion in forum PC Power Supply Discussion
    Replies: 28
    Last Post: 09-29-2017, 07:04 PM
  2. EVGA SuperNOVA 850 G3 - Review @Tom's Hardware
    By Internet in forum PC Power Supply Discussion
    Replies: 7
    Last Post: 03-17-2017, 01:47 PM
  3. Amazing features of Google Earth !
    By chikiwighi in forum Off Topic
    Replies: 1
    Last Post: 01-18-2010, 06:47 AM
  4. Wow - Amazing UPS Tracking
    By SuperSix in forum Off Topic
    Replies: 4
    Last Post: 07-02-2007, 11:14 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •