Foscam IP Camera powered from bench supply

[Zargon]

Hi all,
I am trying to prototype a battery back up switcher for my Foscam FI8910 Ip camera and I've run into a very basic problem with powering the camera from anything other than the supplied 5V 2A DC plug pack.  When it boots up the red led lights up then after a few seconds the green one lights and it goes through a pan/tilt calibration run.  Half way through that the green led starts to flash which indicates network connectivity (which I've confirmed I can access over the network via wifi).  So every thing is great using the plug pack as it should be.

But hooked up to my bench supply, the red led comes on and the green comes on just after but then just goes back to red as though it's reset.  The pan/tilt run never occurs and there is no network connectivity. 

I've tested the plug pack and it's 5.1V unloaded.  I set the bench supply to 5.1V and the current limit to 2A so I should be replicating the plug pack parameters but it just won't work.  The bench supply has a 50mV ripple spec where as I measured the plug pack ripple to be maybe 4mV unloaded.  I don't know if that may be an issue?

I don't have another 5v 2A plug pack to test with it though I did try a 6V 500mA and got the start of the pan/tilt happening but it stopped after a second or so and the device seems to reboot.

I've also tried running the bench supply through a 7805 connected to the camera but still no dice.

I can't understand why this is not working and it's going to keep me up at night.  I'm hoping it really something simple that I'm missing here.  Can anyone help?

Thanks

[Shock]

Run it on batteries that will tell you if it's the ripple.

[HackedFridgeMagnet]

How many amps does your PSU put out?

Is it a straight DC jack connector?

[amyk]

I set the bench supply to 5.1V and the current limit to 2A

That will give a very different current-limiting characteristic than the original supply - a hard limit, preventing even momentary peaks of high current, whereas the original supply is rated for a maximum of 2A presumably continuous, but it may be able to provide several times that in short bursts. That's why the 6V/500mA (usually it's a bad idea to give +20% overvoltage unless you know what the input circuitry is like...) appeared to work better - no hard current limiter either, although it was likely getting overloaded and caused the camera to reset when it drew enough current to drop the voltage below the low-voltage cutout.

[fluxcapacitor]

Heres a post on the foscam forums .Upping the amps seems to do the trick

http://foscam.us/forum/fi8905-dc-power-requirements-wifi-issues-t1105.html

[amyk]

Heres a post on the foscam forums .Upping the amps seems to do the trick

http://foscam.us/forum/fi8905-dc-power-requirements-wifi-issues-t1105.html

Measuring the actual current waveform would distinguish between really needing a higher current PSU, or if a big decoupling cap across the input is enough.

[Zargon]

Thanks everyone for your comments and help.  The problem was most certainly the current limiting.  That is now permanently etched into my memory banks as something to watch out for.  It didn't occur to me that a 2A rated supply was a continuous rating and peak can be higher but of course it's pretty obvious in hindsight

My supply is a 15A 1-30V Manson unit.  I raised the current limit to 10A and it still didn't work but then I realised I was using the 5A limited banana sockets at the front.  Moved to the full range connecters at the back and it works like a charm.

So now back to my backup switching circuit design.

Thanks again.

[amyk]

>5A for a device that draws 2A nominal is a rather big peak, but given that it has a motor in it, I'm not so surprised; start currents for motors are several times their running current.