I have recently aquired an i3-530. The system draws similar/less power than my old sempron system which my picopsu 120, with dell 150W brick, has been powering for some time without issue, but it will not boot, and the hard drive power cuts out. The i3 system works fine with an ATX PSU.

Any ideas???

your new system might be drawing less power overall, but if it's asking for more of its power from the 5v rail than your picopsu can handle that might be the issue.

What components were you using previously and what have they been replaced with?

Some stupid ones: is the 1156 mobo an ITX one? Does your pico have the P412V connector?

At what point does it fail? Can you get the system to POST with the bare minimum of components (no HDD connected, only one stick of RAM etc)? Is it possible for you to measure the rails using a multimeter?

I'm running an i3 530 on a Gigabyte H55N-USB3 with two drives (SSD - Vertex 2 + 320 Gig Notebook drive) plus a Compro E900F on a 120W PicoPSU connected to a no-name 90W brick from eBay and the system powers up, no issues.

Although mostly in my sig already.. heres the full details;

Previously powered: Sempron 3000+, slightly OC'd, 2x512mb RAM, biostart t-force 6100 onboard gfx, sound card, wifi card, 2x 3.5" HDD.
Currently failing: i3 530, onchip graphics, Gigabyte H55M-D2H, 1x 3.5" HDD, 2x2GB RAM.

And yes i have the P412V connector. I was wondering if it was the 20pin connector in the 24pin ATX connector that could have been the problem although it hasn't been before and doesn't seem to be for others.

What would be pulling much power from the 5v line? IIRC the CPU power is mostly derrived from the 12V line right?
It seems to pull about the same from the wall with the ATX psu as the sempron system did with the PicoPSU.. < 50W idle, depending on how many volts / GHz i give the processor / processor core.
Unfortunately I don't have a multimeter right now (just a Kill-a-watt), although i have been meaning to get one for some time..

I will try without HDD and 1xram and report back.

OK.. with an undervolt and underclock i managed to get it to boot. It worked long enough so that i could log in, get prime95 and speedfan up but soon messed up such that only the mouse would move, nothing else responded. Drawing 59W from the wall at this point. The HDD was skipping (ie, powering down very briefly).
On default settings, no go, can't even get into bios. HDD skips almost right away. I thought it might have been the molex->sata power converter which was used but no, using that with the ATX PSU now and all is fine.

According to SPCR the pico-120 can deliver 6A on the 5V line.. that would be over half of my total, load, power draw as tested, and yet it still messed up.

A multimeter so you can measure voltages at different places would help. You could measure the +12v coming from the brick and compare with the +12v going to the HDD for example.
It could be the brick, it could also be the picoPSU or the cabling. Hard so say. Do you have another brick you could try?

Using another brick is a possibility, but i'd prefer not to as it may require some cutting.. but, i don't see how it could be the brick, as it powers a greater total load on the old system, and it has been very reliable in doing so??

Other than a small amount used by the HDD, what uses the 5v line on a modern motherboard???

It might be worth trying a different hard drive. 7200 rpm notebook drives use a lot less power (approx. 5W for spin-up) than their 3.5" counterparts (~20W for spin-up).

My picoPSU 120W could spin up four 3.5" HDDs without any problems, and I was thinking about adding a fifth. Something is wrong if a single drive has problems spinning up.

I agree that something with the power supply (either the Picopsu or the brick) seems to be not right. However, the next best thing after checking with a different brick (which seems to be a bit difficult for the OP) is to reduce power consumption of the system. Replacing a 3.5" drive with a notebook drive is the easiest to start with.

Another 3.5" HDD would also work. Only reason to replace the HDD is to rule out a faulty HDD causing the problem.

I've been meaning to jump in on this discussion as I've been having similar odd issues with PSUs. I picked up a Zacate system just recently and was powering it with the 160W PSU from the ITX case it was in. Worked fine for a while but then I would get just a short burst of power before it would shut down. Tried my 120W picoPSU (65W 19V brick) and it worked fine so I assumed it was the cheap PSU. I since picked up a 60W 12V DC brick to use with my picoPSU 150XT. I had the same problem there. Worked for a little while and then it won't turn on. The 12V from the brick seems steady. My kill-a-watt reports about 7W (instead of the 2W normally) and I get a little squeal from the picoPSU. I've been wondering if it's maybe something with the 5VSB.

I've seem a similar problem trying to power an old PIII PC. Works fine on the 120W pico but not the 150XT. I was assuming it was related to the 150XT being designed for newer system with more 12V power where the PIII runs off the 5V rail. I need to do more investigation still.

Why would you not try a cheap 300 watt PSU to see if it makes a difference. Then experiment to see if the problem is deficient power output from the Pico PSU or that maybe it just broke at an inopportune time.

Could it be that the Pico PSU was just damaged somehow during the switchover?

Without a multimeter we will never know.

You can vet a second Pico PSU and try that. Don't know what a multimeter costs. Depending on what it costs this might not be an unreasonable approach to diagnosis.

A cheap one is 10$ or something like that. There are of course also much more expensive ones as well but a cheap one will do. Sorry for off topic, but a multimeter is really a must for all kind of stuff like this. Without the ability to measure the rails coming out of the PSU we are down to guessing, and replacing stuff until it works.

OK.. i return with multimeter.

I get a reading of 12.15v from the output of the brick (no load)... let me know where else to measure/how ?

edit: with the sempron system getting 12.04v / 5.06v from a molex.. off to try with i3.

mmkay so with the i3 it measured around 12v /5v.. for a fraction of a second and then down to 0. Auto restart and repeat. bios corruption, yay. luckily it recovered from the back up and all is ok. I could under-clock ahead of use and try again, however i'm fairly sure this is what will happen but slower.

what next?

Undervolt and underclock again and get into the OS if you can, but do not stress test the system. Check the voltage (+5v and +12v) using a molex. Anything weird like low or unstable values? Try stressing the system a bit, monitoring the voltages.

Urgh this thing is ****ing me off.

Settings;
QPI:133/1.05v; CPU 1.2/0.8v; GPU 400mhz/1v

idle
5v @ 4.99 / 12v @ 11.87

p95 11.82/ 4.94
---------
up to 2ghz / 0.9v

idle 11.88v/ 4.97
p95 11.79 - 4.92
---------

2.4ghz/ 1v
p95 11.77/4.93

hmm going really well at this point.. so i tried stock settings...
it faillllled. voltages were ~ 11.88 / 4.93 (?). Dropping momentarily as the HDD span up/down again.
back to 2.4ghz.. faill again. Not even as far as the windows logo on startup mostly. despite the fact i had i was just priming fine at the same settings (short test, but still). Tried again many times, same. Monitor outputting solid primary colours sometimes on crash.. sometimes flickering between them. I would suggest it was due to it being a cold start up.. but then a few times it got to the 'start windows as normal' option.. and crashed after selecting that, when the hdd was fully spun up.

I'm hoping this has given you some ideas...

Running out of ideas over here, sorry.. Nothing wrong with those numbers, I guess you could also measure the voltages using the normal PSU. Perhaps the +12v is a little low, but it's well within spec. Perhaps you could also measure the +3.3v rail.
Or start looking for another solution. Thinking about a Winmate 130W for my next build myself.

lol, i thought the multimeter wasn't going to solve much (I wanted one anyway though, so nothing lost).

I plugged in the kill-a-watt so a couple of things to add.. the hdd spin downs are notably not around the peak power loads (ie during spinup).. at 2ghz i'm hitting around 65W max on startup, with an idle of around 38W, from the wall. It just refused to start (once) like this.
(it's pretty annoying when the board 'fixes' my underclock whhen it doesn't boot by returning to stock values and thus screwing things up much more!)

How would i measure the 3.3v line?

Gremlins. It's gotta be Gremlins.

Might be the board. Do you have access to a different 1156 board to test with? I had to RMA my H55N-USB3 due to a non-working audio chip. The replacement board is ok now. I wouldn't be surprised if a different board works well with the rest of your components.

Unfortunately, no.

I have got i3 530 +gigabyte h55n-usb3 + 2,5 inch 7200 RPM drive and successfully tried to power it with :

Pico80
Pico90
Pico 160XT

all using 80W 6,7A Ac/DC brick .

There is no chance that pico120 will not work with your config.

1156 mAtx will not use much more power than itx board .

Try other power brick and maybe change mobo.