Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 35

Thread: Smaller PSU in the future?

  1. #21
    Join Date
    Jun 2010
    Location
    The Netherlands
    Posts
    5,228
    Thanks Thanks Given 
    374
    Thanks Thanks Received 
    71
    Thanked in
    61 Posts

    Default

    Quote Originally Posted by ashiekh View Post
    The output voltage is low (5 super-capacitors in series covers 12V)
    The problem was the low holdup time, because there isn't enough room for a bigger primary cap. How does the use of many more capacitors fixes that? That takes even more room, not less.

  2. The Following User Says Thank You to -The_Mask- For This Useful Post:

    Jon Gerow (03-05-2019)

  3. #22
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    15,861
    Thanks Thanks Given 
    525
    Thanks Thanks Received 
    178
    Thanked in
    128 Posts

    Default

    Quote Originally Posted by ashiekh View Post
    What if one put super-capacitors on the output for holdup time?
    Quote Originally Posted by -The_Mask- View Post
    The problem was the low holdup time, because there isn't enough room for a bigger primary cap. How does the use of many more capacitors fixes that? That takes even more room, not less.
    Yeah. Let's not bring this up again.

  4. #23
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    1,860
    Thanks Thanks Given 
    104
    Thanks Thanks Received 
    32
    Thanked in
    30 Posts

    Default

    Quote Originally Posted by -The_Mask- View Post
    The problem was the low holdup time, because there isn't enough room for a bigger primary cap. How does the use of many more capacitors fixes that? That takes even more room, not less.
    Super-capacitors are so called because they are MUCH more compact (energy wise), by an enormous factor; so they would fit in the space, even if they were in a different place (on the secondary rather than the primary).

    The capacitors needed would be tiny (physically), and would go on the 12V line (for a dc-dc unit).


    Forgive me for insisting, but I think there is an important point here, in the sense that I feel this is a valid solution.
    Last edited by ashiekh; 03-05-2019 at 04:00 PM.

  5. #24
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    15,861
    Thanks Thanks Given 
    525
    Thanks Thanks Received 
    178
    Thanked in
    128 Posts

    Default

    Quote Originally Posted by ashiekh View Post
    What if one put super-capacitors on the output for holdup time?
    Quote Originally Posted by ashiekh View Post
    Super-capacitors are so called because they are MUCH more compact (energy wise), by an enormous factor; so they would fit in the space, even if they were in a different place (on the secondary rather than the primary).

    The capacitors needed would be tiny (physically), and would go on the 12V line (for a dc-dc unit).


    Forgive me for insisting, but I think there is an important point here, in the sense that I feel this is a valid solution.
    And if you don't regulate the output from the capacitors, how do manage the tremendous drop in voltage when the cap is 50% discharged?

    So you're already proposing an expensive solution, now add to that an output regulation circuit and you're not saving any space at all.

    You're better off having the larger primary cap on a daughter board, perpendicular to the main PCB, so it clears the fan's "keep out zone". But that's assuming you have space on the PCB to do that and that the cap doesn't block the airflow of any components on the main PCB.

    Like so:


  6. #25
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    1,860
    Thanks Thanks Given 
    104
    Thanks Thanks Received 
    32
    Thanked in
    30 Posts

    Default

    Quote Originally Posted by jonnyGURU View Post
    And if you don't regulate the output from the capacitors, how do manage the tremendous drop in voltage when the cap is 50% discharged?
    With a slight design change

    At the moment the 3.3 and 5V are bucked from the 12V

    If instead one generates the 3.3, 5 and 12V from a basic 12V the problem is solved.


    I though the idea of these forums was to be able to kick around possible solutions.

  7. #26
    Join Date
    Jun 2010
    Location
    The Netherlands
    Posts
    5,228
    Thanks Thanks Given 
    374
    Thanks Thanks Received 
    71
    Thanked in
    61 Posts

    Default

    Well first I actually don't see a problem with a hold-up time that's a bit shorter. 16ms at 80% load isn't that bad. I don't see anyone getting issues because of that.

    Second it's not a slight design change. Probably a big design change. And you also don't know how it affects te performance. And last, it would be really expensive, as does caps aren't cheap.

    It would be a lot easier to "just" make room for a bigger or second hold-up cap.

  8. #27
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    1,860
    Thanks Thanks Given 
    104
    Thanks Thanks Received 
    32
    Thanked in
    30 Posts

    Default

    The ATX specs perhaps

    Small super-caps are not so expensive, we only want to compete with the original primaries for energy.

    Quote Originally Posted by jonnyGURU View Post
    ...and due to the limited size for a proper bulk cap, could not have 16ms hold up time @ 80% load.
    Last edited by ashiekh; 03-05-2019 at 06:20 PM.

  9. #28
    Join Date
    Oct 2006
    Location
    jonnyGURU forums, of course!
    Posts
    15,861
    Thanks Thanks Given 
    525
    Thanks Thanks Received 
    178
    Thanked in
    128 Posts

    Default

    Quote Originally Posted by ashiekh View Post
    With a slight design change

    At the moment the 3.3 and 5V are bucked from the 12V

    If instead one generates the 3.3, 5 and 12V from a basic 12V the problem is solved.
    Even with a single output voltage, I would want the output of the capacitors to be regulated and that's going to take up a bit of space. You're essentially going to have to introduce a whole other DC to DC circuit to the PSU.

    Quote Originally Posted by ashiekh View Post
    I though the idea of these forums was to be able to kick around possible solutions.
    It is. But this idea of yours keeps coming back to haunt me.

    Quote Originally Posted by ashiekh View Post
    Small super-caps are not so expensive, we only want to compete with the original primaries for energy.


    AFAIK, super-caps are twice as expensive as similar capacity Li-Ion batteries. I wouldn't necessarily use a Li-Ion inside a PSU to increase hold up time... just saying.

    Quote Originally Posted by -The_Mask- View Post
    Well first I actually don't see a problem with a hold-up time that's a bit shorter. 16ms at 80% load isn't that bad. I don't see anyone getting issues because of that.

    No. Of course not. MOST people either have reliable mains and/or a UPS.

    It's the marketing/reviews that kill it. What sells higher end PSUs (which a power dense PSU would be considered "enthusiast grade" since there's a cost adder for the power density)? Reviews and marketing bullet points. Scenario: Hold up time isn't to Intel spec. Minus one point for performance. Everything else perfect. Final score: 9/10." Typical Internet review reader: "9/10? I want 10/10. Hard pass!"



  10. #29
    Join Date
    Oct 2017
    Location
    Grand Junction, CO
    Posts
    1,860
    Thanks Thanks Given 
    104
    Thanks Thanks Received 
    32
    Thanked in
    30 Posts

    Default

    Let's do a calculation

    Say the primary we would want is 450V 600uF, that's 60J; now lets have 5 super-caps at 2.7V, each would need to be 4F to hold 70J, which could supply 1kW for 16 ms, about what is needed.

    On eBay I see 5 pieces of 10F for $9 with shipping (25 x 10mm) so I imagine in bulk the price is comparable with the high voltage capacitor one is replacing; this is over 2 times what is needed and getting closer to a 50 ms hold up time (Such a small Li-Ion battery could not be pushed to 1KW).


    Anyhow, this is all for fun; it goes without saying that all my back-of-the-envelop calculations need checking.
    Last edited by ashiekh; 03-05-2019 at 08:53 PM.

  11. #30
    Join Date
    Nov 2010
    Posts
    379
    Thanks Thanks Given 
    0
    Thanks Thanks Received 
    10
    Thanked in
    8 Posts

    Default

    and how efficient would the 90v AC .. 250v AC to 12v be?
    how efficient would conversion from 12v to 400..450v be? how much current you pull from those caps?
    also supercaps often have low max temps, like 60-75c max ... and sometimes limited life. not suitable.


    i'm wondering if having a custom 24pin atx cable would help with the lack of space on sfx psu.
    like, use awg14 or awg16 wires, but fewer (2 pairs for 12v, 2 for 5v, etc), have a tiny box a few inches from the 24pin connector which has maybe -12v regulator inside, maybe 3.3v 5-10a regulator inside and have those 23 wires come out of that thing.

Similar Threads

  1. Future of PSU's and efficiency ...
    By myststars in forum PC Power Supply Discussion
    Replies: 4
    Last Post: 01-07-2017, 06:52 AM
  2. Yor: Hunter From the Future
    By Lemmy in forum Off Topic
    Replies: 1
    Last Post: 06-27-2010, 10:15 AM
  3. One big PSU or two smaller ones?
    By ADK in forum PC Power Supply Discussion
    Replies: 17
    Last Post: 04-21-2010, 08:35 PM
  4. Truck of the future
    By Tech Bori in forum Off Topic
    Replies: 4
    Last Post: 07-29-2007, 11:46 AM
  5. New System PSU Reconmendation (To Future proof, or not to future proof)
    By amcphee in forum PC Power Supply Discussion
    Replies: 9
    Last Post: 05-14-2007, 06:45 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
  •