help! cap in charger shuts off power supply

[swatson144]

I finally broke down to build a ps out of an atx computer power supply. No problem with the setup on that. I pulled a nice dual fan 600W out of inventory and and kludged temporary connections with the 5v loaded and V-ok connected to 5v and all works well.

Problem is my triton has a large capacitor that the ps detects as a short and powers off.

I haven't done much expermenting with circuits in about 30yrs and have no idea where to start to build something that will address this problem. Basicaly something that will ramp up I until it goes full pass at about 10-12v.

I could probably cheat with a pot like 10k to 0 and then a switch to bypass it before I start to charge etc. I'd really appreciate a pointer to something a little less kludgy. I don't want anyone to design a circuit for me (though if you have one it'd be appreciated), just a pointer to a device that might be what I need. Or maybe I'm just making this hard.

I have been lurking here for a while and have built some rc switches, and micro screams. Made a living in the 70s doing stereo and TV repair. I can make simple circuit boards. eyes and uncle artor don't allow surface mount. I just got back to tinkering with electronics and am astonished at the circuits available in a 3 pin package :eek:

Thanks for all the ideas and sharing the fun folks. It's appreciated!

Any help would be appreciated. TIA

[micsaund]

Odd that you're having a problem with the Triton and a PC supply. I'm using a cheapie no-name AT supply with my Triton and it does not have any issues. Perhaps the newer-tech (fancier) 600W box you've got is more proactive about shutting-down in a "short" condition.

I suppose you could try putting an automotive light-bulb in series with the cap for a test, although that's pretty kludgey too. It might not affect the input voltage to the charger enough to cause any problems but slow keep the PS from detecting a "short". If that works, you could find something more permanent if needed. The other option is to maybe look for a supply that's not quite so "modern" and use that instead.

Mike

[Mr.RC-CAM]

What is the load resistor value that you have on your 5V output? If this resistor value is too large, the switcher supply can behave poorly.

[swatson144]

Yep the Atx is more active in short circuit protection.

I've tried 1-4 12v turn signal filaments in parralell with no change other than an increasing 12v output which is to be expected.

I really hadn't thought of it until mentioned by micsaund but maybe an easy fix would be loading the 12v a bit also.

Thinking in simple terms I didn't even try to power it up with the triton connected maybe that would make a difference. I was just wanting to make sure the PS came up and my voltages were right... nearing 11pm here and I'll just skip out and try that.

[swatson144]

I feel a pretty silly it was really that simple. Connect the triton to the ps and turn it on. It seems the ps charges the cap in the triton as it charges it's own or expects the main board caps to charge and doesn't shut down. :o

I'll butcher and groom it and make it pretty for use on saturday. It should be a pretty nice project as it is for gaming machines. gold anodized with blue LEDs in the 2nd fan. It claims 20A on the 12v line (right). Pretty pimpy and a lot cheaper than one of the server class PS I was going to pull from inventory if it didn't pan out.

Thanks for the replies! I forgot about KISS in the last few decades.

[micsaund]

Oh - hahaha yeah, gotta have the load present at power-up ;) I always have my charger(s) connected when I flip the switch on my supply after, back in the 386 days, I cooked an AT supply by turning it on without any load.

Glad you got it working! I just wish someone would figure-out how to not use the 5v load resistor as that's just excess heat (mine gets pretty toasty). Oh well - things work as they are so I'm happy :D

Mike

[swatson144]

I removed all the extra wires and mounted a turn signal bulb inside using both filiments in par'l as a 5v load. The bulb is right behind the black fan and doesn't even get warm. If I'd have thought to angle a little more to the right you could have seen the orangish dead flashlight glow from inside.

Only snag finishing is there turned out to be another orange wire that was the 3.3v test wire and the beast wouldn't power up until I found it and connected it to another orange 3.3v wire (it was connected at the straight atx connector). It took a couple of minutes to figure that one. :) It's the only Orange that goes to a different pad on the board.

I'm not interested in any of the other voltages so I just did 12v at 12.27V. I can always hook them up later if desired. The triton is happy.

Synapse Green to Black, grey to red, lone orange to any other orange, bunch of yellow to +12v, bunch of black to ground.

It really wasn't a load thing as my other chargers were ok. It's that big friggin arc making on connection cap in the triton that kicked it out :) Took me a long time to get used to that every time I hooked it up to the car battery under my work bench. ;)

[AndyOne]

LOL What colour is Bule? brown perhaps!

Andy.