ATX psu turns off when powering zvs flyback for sgtc / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

ATX psu turns off when powering zvs flyback for sgtc

1 2 next

Move Thread

LAN_403

KUN

Fri Mar 02 2018, 09:57PM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

Hello i connceted 2 atx psu in series, to get 24 v 8 a, no dummy load. When i cunnect them to flyback it works great and gives really nice arcs without turning off. But when i connect the output of my flyback to mmc cap bank and spargap to drive tc, it only works for secound and then powersupplies turn off(i guess shourt circuit protection) then i have to wait like 30 sec and plug it again and same thing happens. So obiusely my caps draw to much current and then trigger the psu short protection. My question is how do i aviod that or fool my psu so that it gives constant 8 amps, do i need a load on psu output or what? pls help. My zvs is standard with IRFP250N and flyback is just from old tv. My tc is 2.5 cm and 25 cm tall. Thanks for help!

AndreiRS

Fri Mar 02 2018, 11:20PM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

I think your circuit is drawing a lot more than 8A then. Do you have how to measure that? Try to put a toaster or hair dryer in series with your PSU just for testing purpose. That is not short circuit protection, but over current. If it was short circuit, it would not even turn on. If there is any chance that your circuit is throwing back too much RF into the PSU, that could be the cause too. But I'm not sure.

KUN

Sat Mar 03 2018, 01:17AM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

yeah, if it is RF can i use diodes to somehow prevent those signals entering the psu? It is probbly drawing more than 8 amps. I noticed that if i put load on one power supply(a 12 ohm resistor), that psu contines to work wlile the other one shouts down. But if i put the same load on the other one it still make only one work. And no i only have multimeter and i dont think it is precise.

AndreiRS

Sat Mar 03 2018, 01:37AM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

Some PSUs need a load on the 5V line to work properly. But I have never found how much should be that load. If the problem is RF... Hmmm I think non polarized capacitors across the output would filter it better, poliester capacitors. Better wait for someone else's opinion. Usually multimeters are precise enough. But a lot of them only go up to 10A.

KUN

Sat Mar 03 2018, 11:10AM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

Ok i will try the load thing

Alex Yuan

Mon Mar 05 2018, 01:44AM

Registered Member #9614 Joined: Mon Jan 14 2013, 10:00PM
Location:
Posts: 44

Your power supply is turning on, so no problem with the 5v rail needing to be loaded. It could be an issue of overcurrent or noise. I have used computer power supplies for ham radios, and some of them shut off when I transmit (just rf getting into the psu). Just had to keep trying power supplies that werenâ€™t as sensitive. Iâ€™d be careful measuring the current with a meter, it could be well over 10A. (At least temporarily until the power supply shuts off). Either way, looks like youâ€™d need a different power supppy.

GrantX

Mon Mar 05 2018, 11:06AM

Registered Member #4074 Joined: Mon Aug 29 2011, 06:58AM
Location: Australia
Posts: 335

Alex Yuan wrote ...

Your power supply is turning on, so no problem with the 5v rail needing to be loaded. It could be an issue of overcurrent or noise. I have used computer power supplies for ham radios, and some of them shut off when I transmit (just rf getting into the psu). Just had to keep trying power supplies that werenâ€™t as sensitive. Iâ€™d be careful measuring the current with a meter, it could be well over 10A. (At least temporarily until the power supply shuts off). Either way, looks like youâ€™d need a different power supppy.

Yeah, I agree with this. The 5V rail is used to turn the supply on, if it's working without the Tesla Coil connected then that's not the issue. Also, if you can draw an arc off the ZVS-driven flyback and sustain it without the supplies turning off, then it's unlikely to be overcurrent shutdown. My guess is RF interference/noise from the Tesla Coil.

Are there long lengths of unshielded wires coming from the ATX supplies (which will act as an antenna)? Is there any kind of filtering between the flyback and primary circuit or between the ZVS and ATX supplies? The noise could be coming from either capacitive coupling between wiring, or conducted noise reflected back through the ZVS and flyback into the ATX supplies (or a fun combination of both). These situations can be difficult to solve without a lot of trial and error or a lot of extra components.

You could build filters for the DC outputs (after neatening the ATX wiring) and a HV filter similar to a "Terry filter" for the output of the flyback, but this becomes expensive and is not a guaranteed fix. Another solution is to switch to an iron-core transformer with rectifiers and electrolytic capacitors for your DC supply, this will be almost completely RF-immune, but far more expensive.

The cheap solution is just trial and error, by neatening wiring and trying to shield the low voltage components with grounded metal enclosures. If all that fails, you'll definitely need to try a different supply.

KUN

Mon Mar 05 2018, 03:45PM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

Ok, so there are no long unshielded wires coming from psu, yes i can draw arcs just fine without it connected to tc. I guess that i need to try those filters or something. There is no filtering anywhere as till now i didnt realy need it. So if we assume its RF, can you provide some schematics for those filters low voltage and high voltage side. Also i dont know if its important but both tesla secundary, and atx psu are connected to the same earth ground, dont know if thats relevant, and one psu is unearthed from case so it could work normal in series(only output side). The interesting thing is that if i connect my zvs to only one psu(no matter which one) it works normal - the whole tc. I thought about transformer as psu but it is quite expensive, but i could try to make my own as i have some iron core lying around.

Also before these psu i used laptop psu, and they were also acting the same, turning on and off at a constant rate. But as i made the spar gap smaller at some point it worked normal, but couldnt run it if the spark gap was bigger, and that was a problem, i thought i could solve it with more amps psu but no. Also even now i can occasionally run it with small enough gap, but as it is really small gap, tc doesnt work good.

KUN

Sat Mar 10 2018, 11:57PM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

Ok so, i think i know the reason for psu shutting down. When i use only one psu, no problem, it can provide enough amps and thats why it stays on, when i have two, they cant give enough current so the supply that can dettect overcurrent faster turns off faster, so then zvs only gets 12 v, and cant draw much more amps so it stays on. So how do i solve this. I searched how to remove over current protection from atx psu and found some chip pin grounding tips, i tryed them but as i had different chip i messed up one of the supplies, it died so i took it into peaces. Now i have only 1 atx and one laptop psu that is capable of giving decent output(19 v, 5 a), i guess it is enough for my coil as it generates almost 25 cm arcs with those supplies. So how do i hack my atx psu that i have left to give me 12 v, 8 a allways without shutting down, so that i can short it and that it still stays on? Now when i connect that atx with laptop one, if i make spar gap really small it works but if i make it as it should be the atx psu turns off, and the coil continues to operate with input of laptop psu. I am slowly desining the transformer for my needs but what can i do until then??

teravolt

Mon Mar 12 2018, 04:27PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

When you are done with ATX power supplies try a 30v 10amp transformer, bridge, and filter cap. the ATX supply
may not be able to handle surges. After all you would not want that kind of power on your mother board

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob