Thread: Opus 150 w/ Core 2 Duo? or atom 330?

I've got some findings to share that perhaps you can try to figure out if your power supply is even the culprit. I ran with the E8400 and then installed my E7200 and ran it at different clock frequencies with the last test being at minimum vcore that I can get (no undervolting by BIOS options).

Idle consists of no activity immediately after running some type of test off of the platter based HDD. Load runs INtel Burn Test and ATI Tool, both apps which cause the maximum power consumption and heat generation on my main PC. The graphics do not really impact the power consumption (<100mA), even when changing the processor priority of IBT. All devices still attached as above. All measurements taken from my shunt ammeter and clamp ammeter for consistency. Power taken off of my Stinger SPS60 power supply (13.8v output at <60A).

E8400
idle @2.00GHz 1.175v 1.8A
load @3.00GHz 1.225v 4.6A

E7200
idle @2.00GHz 1.15v 1.6A
load @3.13GHz 1.20v 4.4A

E7200
idle @1.60GHz 1.15v 1.6A
load @2.53GHz 1.20v 4.0A

E7200
idle @1.20GHz 1.15v 1.5A
load @1.90GHz 1.20v 3.5A

E7200
idle @1.20GHz 1.15v 1.5A
load @1.90GHz 1.15v 3.3A (RMClock)

You can save 1 amp, likely most or all on the 12v rail, by underclocking about 1/3 of the cpu's max frequency. Obviously this is on my system only but these figured should be similar enough across the board for percentages of frequency drops without voltage changes. The above numbers equate to ~3.6A MAX draw on 12v at 1.9GHz/1.15v if no other voltage rails were being used (3.3A input * 13.8v input = 45.54 * 0.95 efficiency / 12v output = 3.6A). I can't see how your X3 could be any worse.

It is nice to see a 45 degrees Celsius load temperature with the 1.9/1.15v though, with a low profile heatsink Too bad the motherboard is stupid and can't automatically control the fan, although it's PWM and I can control it in the BIOS by manually setting the speed.

Im gonna play with the frequencies today to see how well it runs. Knowing I may use addon cards later, it may be smart to go ahead and just get a PS thats a little better power efficient too.

Cards can use a bit more power, yes. USB on the other hand, my devices add up to under 500mA draw for the 5v with the heaviest load coming from the USB to RS232 adapters (3) at 90mA max each. GPS draws 50mA max and the rest are all simple signal connections such as controlling the touchscreen.

I'm fairly certain that your CPU draws from 12v as most motherboards do that when they have the separate 12v 4 pin connector. This still doesn't explain your issues once you attach a USB device. I just remembered that you said above that you tried it with a single USB device and with a normal PSU - still odd behavior unless the Opus has got a problem or the motherboard does in fact use 5v for the CPU (which not even the M4-ATX may deliver enough juice once you attach some USB devices but it should).

Have you tried booting the system with only the motherboard/CPU/RAM and USB attached, without any hard drives or other devices?

I don't see any harm coming from it, but if it were me I'd verify that the standard ATX power supply can put out 12v without any load on the other rails and attach it to the 12v 4 pin header on the motherboard, keeping the Opus on the 20/24 pin ATX connector. Give it a boot and there should be some 12v draw on the normal PSU, taking some of the load off of the Opus to help you find where the problem is coming from. The normal PSU can be powered by shorting out the green wire to ground (next pin over with a black wire) on its ATX connector.

As long as the M4-ATX is the solution to amperage problems, I don't see why it wouldn't work well in the above system. It's a spendy test though.

I found this "calculator" to be pretty accurate: http://web.aanet.com.au/SnooP/psucalc.php

They don't have your Rena core in there though. Also seems to show similar numbers from my findings yesterday about the power consumption with the vcores I was using. Obviously there are other power benefits by decreasing the front side bus, such as power savings in the northbridge, but the numbers in that calculator match up well with the numbers through the ammeter. My CPU should be sitting just over 21 Watts at full load at 1.9GHz/1.15v, 17-19 Watts if I had a newer revision motherboard that actually supports undervolting.

Keep in mind that stock the E8400 is estimated to use 33 Watts and the E7200 28 Watts, although BOTH have a 65W TDP. I'd expect your X3 uses comparable power, if not slightly less as it's only 2.2GHz.

I just wish people would stop going off of TDP - my quad core has been tested to be close to the 65w that others claim as their consumption for their dual core "65w TDP" processors. I've got it overclocked and undervolted, which in turn runs a couple degrees cooler than stock, both with the same TRUE heatsink. Only noise in that system is the hard drives.

Also notice how a Core2Duo at 1.0v and 1.0GHz (which would likely perform better than an Atom 330) falls into the same power consumption range of the Atom which DOES use power similar to its TDP rating. Atom is not for power savings as much as it is for price - $50 Atom vs $250 for Intel's "ULV" Core2Duo's, along with a motherboard that's simplistic in design so they are manufactured cheaper (excluding Zotac's expensive mb/cpu combos which are also much more complicated). Core2's have been getting similar battery life as Atoms, and in turn power consumption, for awhile in UMPC systems.

I still don't think your power supply model is at fault - likely just defective somewhere.

Hey guys so I picked up this power supply on ebay(the opus 150 watt) What proc/mobo do you recommend? I'd like fast boot times, but only need the computer for mp3 and gps.

id pick up an intel 45w tdp core2 chip, maybe even a dual core. im sure everyone else will tell you to go with atom, but if you have the power supply to run a real cpu- i say why not.