I think that some/many ATX supplies use feedback from the +5Vdc line to determine the regulation of the whole SMPS.
If you were to move the NFB tapping to the +12Vdc line and adjust the feedback ratio, you should be able to safely change the output regulation this small amount.
they do, most of the atx psu i tested, the 12 volt rails are at 11.xx volts when tested shorting pin 14 to any ground pin ground, but use that psu on a pc and the rails actually go above 12 volts...
thus my recommendation to load the 5 volt rail, some atx even have a load built inside the psu a small load resistor, but to load the 5 volt rail with 1 ampere requires a 4.7 ohm 10 watt resistor..
no need to tamper with the innards of the ATX psu, it works off the line without isolation, very dangerous, so i will not recommend that you do it...
thus my recommendation to load the 5 volt rail, some atx even have a load built inside the psu a small load resistor, but to load the 5 volt rail with 1 ampere requires a 4.7 ohm 10 watt resistor..
no need to tamper with the innards of the ATX psu, it works off the line without isolation, very dangerous, so i will not recommend that you do it...
ATX supplies don't work like that anymore. Ever since the all PC mainboards changed to use +12V to power the uProccesor. oh IDK 15 years ago? The newer PC supplies the 12V rail is the main feedback and they don't need pre-loading due to Intel sleep mode and other Green initiatives. Infact the newest ones have 5V and 3.3 DC/DC converters running off of the 12V rail.
Last edited:
The ATX uses a transformer to ISOLATE the output from the mains.
The metal enclosure around an ATX SMPS is directly connected to the Protective Earth.
There is nothing dangerous about that !
Isolated and an effective Safety Earth for the exposed metal parts.
It complies with ClassI, so we bolt them into our desktop computers and don't have to worry about electric shock.
Last edited:
Anyone who is familiar with offline switching supplies knows that half the board is fully charged at ~400Vdc and has no isolation. so yeah danger abounds.
As usual your words are sort of correct, but all wrong on any other practical level. The whole ball game changes when the OP talks about opening said "metal parts".
Last edited:
Infinia is right, obviously Andrew have not seen the innards of a psu to know that it works off the line without isolation, you can get zapped with the raw B+ of 300 volts dc and there in nothing between you and the main power lines...the thought scares me even though i have worked on a lot of those.....
Infinia, thanks for pointing out the 12 volt initiative by Intel, i almost forgot....
Infinia, thanks for pointing out the 12 volt initiative by Intel, i almost forgot....
this assume that you are merely using the psu on a pc, not opening it and tinkering with the innards...
and i remember how you reported posts dealing with mains without isolation, changed you mind now?
the topic is still taboo here...
The feedback can be tinkered on the secondary side... and then you can close the chassis back up.
There is no way for the secondary side to see 300+V unless you seriously screw it up and somehow connect the primary to the secondary. Which if you choose to use as your argument, is a fair one. But at this point of reasoning, all power supplies discussions including those using a 50Hz transformer should be taboo due to gross incompetence of the builder.
P.S. group-regulated PSUs still exist today. Not that I think cross-loading is a good way to increase the +12V.
There is no way for the secondary side to see 300+V unless you seriously screw it up and somehow connect the primary to the secondary. Which if you choose to use as your argument, is a fair one. But at this point of reasoning, all power supplies discussions including those using a 50Hz transformer should be taboo due to gross incompetence of the builder.
P.S. group-regulated PSUs still exist today. Not that I think cross-loading is a good way to increase the +12V.
Last edited:
Well, you *should*
Voltage is never absolute but relative to another point.
* If you touch (accidentally or on purpose) the "+380V" node with a fingertip (or elbow or wrist or any other body part) but nothing else in that supply, specially its chassis or supply ground, nothing happens.
* if you touch *any* point in the *hot/live* side of that supply, you die , for the very good reason that already, always, you are touching "another point" : ground/earth/the floor.
You think both situations are the same?
If you are "tinkering" with the secondary side, you are handling/touching/holding one way or another a single board, palm sized, where you will be soldering/desoldering/modding on the admittelly safe secondary, 1 inch away from half the board which is live/hot.
The potential for error is there.
No need to connect secondary to primary, fully exposed hot parts are there at your fingertips.
You are comparing apples to oranges.
Standard power supplies are:
1) very simple, you only have to pay attention to power switch contacts, fuse ones, and little else, and they are "all together" on a separate, easy to avoid place, wires carrying mains voltage into the transformer are insulated, while *all* components and tracks on an SMPS primary side are hot and exposed.
2) when building an amplifier, mains wiring is made with everything unplugged and never touched again, while when tinkering with the board as needed here, you are handling and touching it and to adjust it must be on.
Again:
You think both situations are the same?
Exactly. Thanks for that great example of the straw man logical fallacy JMFahey.
Take care with the hot side of a switching PSU.
There can be a lot of energy stored in the caps. The residual current breaker on your bench isn't going to stop that hurting.
If the OP wants more than his PC supply will give, step up switchers are available on ebay for the price of a beer. But I'm sure many attics still contain PC supplies where loading the 5V will boost the 12V a little. Or you can buy a 15V smps for the price of two beers.
There can be a lot of energy stored in the caps. The residual current breaker on your bench isn't going to stop that hurting.
If the OP wants more than his PC supply will give, step up switchers are available on ebay for the price of a beer. But I'm sure many attics still contain PC supplies where loading the 5V will boost the 12V a little. Or you can buy a 15V smps for the price of two beers.
I hate to see the final product incorporating hacked DC/DC boost boards and a lone ATX supply. Time to re-think the whole purpose of audio tube / glowing glass circuits. IMO half the sound is all the big iron and copper. Besides the sounds, the sight of a HV Rectifier tube is pretty awesome..
the feedback the controller is on the secondary side....
but the primary side of the chopper traffo feeds off the mains rectified
psu rectifier and caps and has no isolation from the mains...
and herein lies the issue....fraught with danger....
the new forum rules did not make the danger go away....it still is a safety concern...
the difference is, now we can talk about it...
12 volt initiative by Intel.......https://www.intel.com/content/dam/www/public/us/en/documents/guides/power-supply-design-guide.pdf
old Intel ATX psu design guide.....https://paginas.fe.up.pt/~asousa/pc-info/atxps09_atx_pc_pow_supply.pdf
old Intel ATX psu design guide.....https://paginas.fe.up.pt/~asousa/pc-info/atxps09_atx_pc_pow_supply.pdf
Switch Mode Supplies are everywhere.
There are many threads on modifying the Behringer DCX2496 crossover but not much mention of the Switch Mode power supply - It has exposed rectified mains voltages on the capacitor top - (A bit of tape helps here.)
Working on any device with high voltages is dangerous.
With the right information I don't see why the insides of an ATX PSU are different from almost any other consumer product.
There are many threads on modifying the Behringer DCX2496 crossover but not much mention of the Switch Mode power supply - It has exposed rectified mains voltages on the capacitor top - (A bit of tape helps here.)
Working on any device with high voltages is dangerous.
With the right information I don't see why the insides of an ATX PSU are different from almost any other consumer product.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.