Cold Cathode Causing Crap

[RyanM]

Bought a pair of the dual Cold Cathode UV lights off SVC (eXtreme Gamer (XG) brand) since they're so goddamn cheap.

I got the first set out of the box, hooked it up, and was amused to find that one of the cathode bulbs is dead. Out of the box. Yay.

The other one works fine enough...till you flip the included switch to try and turn it off. The system then loses video, and shuts down after a few moments. I'm guessing this inverter is cheap and causing a large enough power spike when activated/deactived to trip the Active PFC on my power supply.

I'm going to try hooking up the inverter that came with the other set, but if this is indeed the case, what's a workaround short of buying a non-Active PFC power supply? And has anyone else seen this problem with cheap cathodes, and what's a good brand, and where's a good place to find them?

SVC, FrozenCPU, and Newegg all carry ones that may not carry the XG brand name, but look identical down to the cathode housing, power inverter, and included switch type.

 

[gotensan01]

Well, I know nothing about CC lights but I was thinking about buying some. LOL I was going to buy the cheapest ones on svc but I'm glad you posted this. I'll go with a better brand name like cooler master. Sorry I can't help though.

 

[RyanM]

What's bugging me is damn near every other brand of light that's on Newegg, SVC, and FrozenCPU all look JUST like this "XG" brand. Logisys, Sunbeam, even Vantec - They all have the cubed plastic feet and identical specs. I'm hesitant to buy any of them.

I'm hoping some people on here can help figure out if the issue is the lights or not.

Until my power supply tester comes in the mail Friday, I've got no way of powering up another power supply to test this stuff out, short of hooking up a mobo on the table, and I really don't feel like pulling one out right now.

 

[RyanM]

Okay, I've managed to narrow this down.

Tested the identical inverter/tube on my computer. Even when put on the same rail as my 6800GT, I could flip the switch two dozen times with no ill effects.

Now, there are two differences between my system and the one I'm trying to get the cathode to work on, aside from the fact that I'm probably pulling twice the wattage as the other system:

1. It uses a power supply with Active PFC.
2. It's a Socket 754 system, while mine is a Socket A system.

Aside from that, there are no other factors that could possibly be effecting the cathode and my ability to switch it off.

I was ready to crucify the motherboard when a friend mentioned he's 100% positive he has an active PFC power supply and has no problem switching his lights on and off.

So, the ONLY possibility is that the Active PFC power supply, despite being rather expensive and supposedly quality made, sucks.

I can't think of a damn other thing that would cause this behavior.

 

[RyanM]

GAH.

Or maybe it is the motherboard.

Speedfan is reporting the CPU voltage fluctuates between 1.79v and 2.79v, and the 12v is supposedly hitting between 3v and 11.19v.

Now, there's no bloody way the computer could be running with voltages like that, and NO f*cking way a brand new $70 PS would put out voltages like that, short of it being DOA.

So now I'm thinking, mobo problems?

Ugh.

 

[Gnoad]

How bout this. You flick the switch and it shuts the power down to your vid card because anything below the molex pin the switch is connected to will be turned off. I did this without thinking once before. Make sure your cathode is plugged in at the very last molex connection.

 

[RyanM]

I've currently got the thing on it's very own rail. The video card in my system is hooked into the PS, but the video in the system I'm installing the cathode into is on the motherboard (I haven't unwrapped the Radeon for it yet.)

So that's not the problem. Thanks for the suggestion though.

 

[gotensan01]

Buy a PSU tester. I'm not sure how they work or if they'd help you in this situation but they are fairly cheap.

 

[RyanM]

Used to have one. Left it at a client's house.

New one's on the way, should arrive from Newegg tomorrow, along with a similar make/model power supply so I can see if it's the mobo or the PS that's tripping up here.

 

[Akira1224]

At the risk of sounding....well like me.... is all this testing and system down time worth a cold cathode? Personaly I would have RMA'd as soon as I saw that one of the lights wasn't working. I installed a cc in my rig and it was one I got from Circuit City...thats about as ghetto as you can get. I havent had any problems.

I would say send it back have them send you a replacement. That unit just may not be fully functional.


Either way though. My logic would just state that

IF(system is working ok w/o cc)ELSE(system crashes with use of CC)THEN(Do not use cc)

 

[AMDFanatic]

The cathode needs it's own rail from the PSU, just like your HDDs and your videocard. Never share a rail unless you need to. If you share a rail, use one off of a case fan or something that's not too critical. I ran out of rails in my case and I have my inverter sharing a rail with a fan connected to the fan controller. The only problem I have is my tachometer reading for one of the fans is higher because the voltage is higher.

 

[glenn beck]

got the exact uv fan, uv worked for like a day, had to send back for replacement, still waiting...

 

[RyanM]

Originally posted by: AMDFanatic
The cathode needs it's own rail from the PSU, just like your HDDs and your videocard. Never share a rail unless you need to. If you share a rail, use one off of a case fan or something that's not too critical. I ran out of rails in my case and I have my inverter sharing a rail with a fan connected to the fan controller. The only problem I have is my tachometer reading for one of the fans is higher because the voltage is higher.

As I've stated, in the system I'm installing this on, it has it's own rail, and killing the cathode or turning it on causes the system to take a shit. However, when hooked up to MY PC, ON the same rail as my 6800GT, on a 400w power supply that came with (I think) a Maxtop case, it works without flaw.

I'm thinking either motherboard or power supply.

I just got another power supply today that's identical to the one that is in this system, and my PS tester arrived along with it, so I'll be able to figure this out once and for all later on when I get a chance to play with it again.

 

[RyanM]

Okay, dupe power supply arrives. Plug in PS tester, plug into cathode....flip switch...over and over and over.....not a single LED on the tester so much as blinked.

Go over to computer, unplug ATX connector from mobo, plug into PS tester, flip cathode switch....not a single LED blinks.

It's the motherboard. I'm struggling to figure out how the motherboard could be causing the problem, but this cathode's been tested on two systems, 3 power supplies, one identical to the one on the system that's malfunctioning, and as long as the thing isn't hooked up to this mobo, it works.

I'll be returning the mobo promptly. I hope I can find that receipt....

 

[Akira1224]

lol...buy a cc....return the mobo. There is logic there somewhere...there has to be.

 

[RyanM]

It's quite simple logic:

Customer wants computer that glows.
Computer that doesn't glow doesn't meet customers requirement.

If returning the motherboard allows me to illuminate the inside of this f*cker, then it makes sense in every logical manner possible.

Lose the attitude.

 

[RyanM]

New mobo's on the bench. Proc's in. RAM's in. PS is plugged in. Cathode is on standby...

Boot....POST....*flicks switch a dozen times*



Works. Somehow, I have no f*cking idea how, but somehow, the other motherboard simply did NOT like cathodes being hotswitched. POS.

Problem cleared up. Yay.

 

[shilala]

Good work MachFive.
I can't imagine why it happened either.
I was going to suggest hooking it up to a different rail, then reversing the polarity.
May be the inverter was miswired.
If the psu's ground is floating, the miswired inverter could be going to ground.

 

[RyanM]

Well, I ruled that possibility out when I hooked it to my computer, then two other PS's via the PS tester. It worked fine on those systems, so that helped me narrow it down to either the PS on the computer or its motherboard. And unhooking the PS and powering it using the PS tester worked fine, so it had to be the mobo.

Damn, that was a pain in the butt to diagnose. I'd just *love* to know WHY the motherboard was at fault.