Results 1 to 18 of 18
  1. #1
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375

    Homemade PSU help

    I am trying to build a PSU. I have it all put together but when i plug it it i get a large current spike on the 120v primary side of my transformer. i have removed all components exept the two rectifiers (their soldered in) and it still occurs. i lost my schematic and board design (due to format and reinstall) but here is a crude paint sketch of the way my board looks:
    PSUboard

    Can anyone tell me what i have done wrong?
    Last edited by Overload; 05-10-03 at 02:23 PM.
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  2. #2
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    additional note:

    with the transformer by itself there is the slightest little twitch on my ammeter, but when its plugged into my board it spikes well over 6 amps
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  3. #3
    Registered
    Join Date
    Apr 2003
    Location
    Netherlands
    Posts
    79
    ever tried powering it up with one of the 20 Vac disconected?
    Although I think it shouldn't be a problem in this case but I still get shivers down my spine when connecting a + to a -

    Might be prudent to check if your rectifiers are still working correctly.

  4. #4
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    yea i tried puting just 20v on each rectifier. it still did the same thing on each rectifier.

    i know what you mean about connecting a + to a - but Theoretically it should not be a problem. i even have a 12v supply (i didnt make it) with two separate 12v outputs. i just connected the + and - from each out put to give me +- 12v. it works fine edit: it does use discrete diodes instead of a rectifier if that makes any difference

    i suspected i may have burt it up soldering it to the board. i used a diode checker and campared it with one that has never been touched and they checked the same. anyone know the proper way to check a rectifier?
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  5. #5
    Member AMD Phreak's Avatar
    Join Date
    Apr 2003
    Location
    255.255.255.255
    Posts
    877
    You could try an o-scope and check for a sinusoidal waveform. That would me my first try. At least that is how we did it in labs at school.
    Embracing the goddess energy within yourselves will bring all of you to a new understanding and valuing of life; a vision that inspires you to live and love on planet earth.....

    If you don't like what I have to say, please direct your comments to /dev/null. Have a nice day :D
    -------------------
    MSI k7t
    AMD Athlon TBird 1100 @ 1291MHz
    Vcore 1.80v
    Verto GeForce4ti 4400 128MB
    Winblowz XP Pro
    Lian-Li PC60B

  6. #6
    Member larrymoencurly's Avatar
    Join Date
    Jul 2002
    Posts
    2,203
    Is it really OK to have those two AC inputs tied together if the plus and minus outputs share the same ground?

    I thought that if the transformer had only one secondary, the only way to get + and - voltages was with a center tap, and only one bridge would be needed.

  7. #7
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    My transformer is center taped with 20v on each side. and it can be done with 1 or 2 rectifiers. if you use just one and use the center tap as DC ground you get a plus and minus rail. if you dont use the center tap you get twice as much DC voltage but only + or -. using two rectifiers like i am trying to do gives a cleaner supply and allows higher current capacity, that is if you do it right:-P

    I wish i had an o-scope, but thay are expensive. however, any suggestions where i might get a good deal on one?
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  8. #8
    Senior Yellow Forum Rat
    pik4chu's Avatar
    Join Date
    Jan 2003
    Location
    Highlands Ranch, Colorado
    Posts
    4,587
    Originally posted by Overload
    .....
    I wish i had an o-scope, but thay are expensive. however, any suggestions where i might get a good deal on one?
    Cheapest Ive seen are on eBay..but still over a few hundreds bucks. But you could get lucky I guess.
    DFI X58-T3EH8 | I7-920 C0@3.0 (V8) | 6*2GB Corsair Dominator @1800-7-7-7-18 | eVGA 285 | 1.5 TB RAID 0 | Win7 64-bit (456Watts)
    Main Server (WIP): DFI X58-T3EH8 | I7-920 C0@3.6 | 3*2GB Corsair Dominator @1800-7-7-7-18 | 8*1.5TB in RAID 5 (283Watts)

    Folding User Stats
    Team 32 Countdown heatware

  9. #9
    Member pby5cat's Avatar
    Join Date
    May 2002
    Location
    california
    Posts
    605
    This artical on the front page could help http://overclockers.com/tips1054/index07.asp
    look on the last page for the scematic
    Intel Q6600 L727A
    Asus Maximus Formula
    2X2gb Gskill DDRII-1000
    2X HIS ATI HD3870 in Crossfire (835,1206)
    WD 150Gb Raptor and Maxtor 250Gb DiamondMax10
    2X Samsung DVD-RW DL SATA
    Silverstone OP850
    Thermalright Ultra 120 Xtreme (lapped)
    thats it, and proud of it.
    My Heatware

  10. #10
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    here is one of the referances i used. its about in the middle of the page, the top one from TR1

    psu
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  11. #11
    Member larrymoencurly's Avatar
    Join Date
    Jul 2002
    Posts
    2,203
    pby5cat, doesn't that schematic have the bridge rectifiers connected slightly differently, in parallel (for greater current capacity?) instead of stacked?

    What happens when you try your PSU with only one bridge installed at a time?

  12. #12
    Member bigben2k's Avatar
    Join Date
    Jun 2002
    Location
    Texas
    Posts
    474
    Your power spike could be from the caps that charge up, when you turn it all on. Otherwise, it's from the coils of the transformer.

    You can compensate for the former, but not the latter. You might want to take a look at an actual PSU schematic, or do what I did, and strip one apart
    Water Block Testing Alliance

  13. #13
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    How do i compensate for it.

    i am starting to suspect the type of transformer i am using. in the link i posted( and others i have looked at as well), the transformer used two separate windings for each rectifier. mine is a single winding with a center tap. other circuits i have looked at using a single rectifier use this type of transformer. so maybe you can only connect two rectifies in series if the windings at serparate. but i havent been able to figure out why this would be. can anyone explain it?
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  14. #14
    Member bigben2k's Avatar
    Join Date
    Jun 2002
    Location
    Texas
    Posts
    474
    To confirm that that's indeed the problem, try the circuit without the big caps (1500 uF): that should definitely identify the problem.

    How did you size those caps?

    What's the max amperage output of the transformer?
    Water Block Testing Alliance

  15. #15
    Member
    Join Date
    Jul 2002
    Location
    Indiana
    Posts
    1,517
    Originally posted by Overload
    so maybe you can only connect two rectifies in series if the windings at serparate. but i havent been able to figure out why this would be. can anyone explain it?
    During half the AC cycle, the diodes in the bridge rectifiers are shorting out the transformer secondary.



    Note the current path shown in red. This applies during one half of the AC cycle.

    To use that transformer you need to tie the center tap of the transformer to DC ground, and half wave rectify each polarity.

  16. #16
    Disabled star882's Avatar
    Join Date
    Jan 2003
    Posts
    536
    "To use that transformer you need to tie the center tap of the transformer to DC ground, and half wave rectify each polarity."
    You can have a design where you have two half-wave rectifiers for each output, one works on one half cycle, and the other works on the other half cycle.
    In other words, you get a full-wave rectifier.

  17. #17
    Member Overload's Avatar
    Join Date
    Dec 2000
    Posts
    375
    ahh, of course. my mistake is very clear now ( dont know why i didnt notice that before). thanks for all of the help everyone. time to start over and do it right this time!
    Mobile Athlon XP 2600+
    Thermalright SLK-900A w/vantec tornado @5v
    Abit NF7-S v2
    PQI PC3200 2x256
    Western Digital 120Gb 8Mb cache
    ATI 9800SE
    NEC 2500
    Toshiba SDM_1212 DVD<- Finally died, she was a faithful beasts:p... Now back from the dead in another box thanks to an "organ donor"!!!

  18. #18
    Member
    Join Date
    Jul 2002
    Location
    Indiana
    Posts
    1,517
    Originally posted by star882
    "To use that transformer you need to tie the center tap of the transformer to DC ground, and half wave rectify each polarity."
    You can have a design where you have two half-wave rectifiers for each output, one works on one half cycle, and the other works on the other half cycle.
    In other words, you get a full-wave rectifier.
    Yes, I didn't explain that very clearly.

    In other words, use the connection scheme shown for TR2 in Figure 2 here.

Posting Permissions

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