Thread: 1.7 ghz P4 with M2-ATX

I found some additional info about specfic graphics cards, which I detailed here.

Also, I found this in the FAQ,which I hadn't noticed before:

Component..................Requirement............ .Line(s) Used
AGP Video Card..............30 – 50W.................+3.3V
Average PCI Card............5 – 10W....................+5V
10/100 NIC.......................4W...................... +3.3V
Floppy Drive.....................5W...................... ..+5V
CD-ROM.......................10 – 25W...............+5V and +12V
DVD-ROM.....................10 – 25W...............+5V and +12V
CD-RW.........................10 – 25W...............+5V and +12V
7200rpm IDE Hard Drive....5 – 20W...............+5V and +12V
Case/CPU Fans..............3W (each).................+12V
Motherboard (w/o CPU or RAM).25 – 40W......+3.3V and +5V
RAM............................8W per 128MB............+3.3V OR +5v*
Pentium III Processor.........38W......................+5V
Pentium 4 Processor..........70W......................+12V
AMD Athlon Processor........70W......................+12V

So, it looks like that my power consumption would be:

• P4 1.7 ghz CPU (12V)----------------65W
• motherboard (mostly 5V, some 3.3V)--25W
• 512 MB RAM (3.3V) -----------------10W
• Laptop HDD (5V) --------------------5W
• pci cards (3.3V)--------------------30W (Radeon 7000)
• AGP cards (3.3V and 5V)------------none
• usb devices (5V)--------------------0.5W each x 2 = 1-2W
• cpu fan (12V) ----------------------3W

Total current by rail:

• 12V: ---65W + 3 W (fan)------------ = 68W / 12V = 6A
• 5V: --- 30W(R) + 5W(HDD) + 5W(USB)--- = 40W / 5V = 8A
• 3.3V: -- 10W(PC133) + ?(MB) ---------= 10W / 3.3V = 2-3A

M2-ATX specs:

• 12V: 8A
• 5V: 8A
• 3.3V: 8A

So, it looks like that I'm cutting it close using the Radeon on the PCI bus. It would likely be better to use the AGP version of the Radeon and take the power from both the 3.3V and 5V rails.

I'm using a M2 on a P4 1.7 system, but only with on-board graphics. I actually measured the current my system draws on each rail with an ammeter while it was at max load. I got 5A on 12V, 2-3A on 5V and <1A on 3.3V. That's with a 3.5" HDD. My USB stuff is powered from a POL. My laptop DVD/CDRW draws about 1.5A on the 5V rail, do you have one of those too? Also, why the Radeon? Can't you use onboard or some other low-level graphics card? If you're not doing any 3D then theres not much point using the Radeon, it will just waste power. Be warned that overloaded M2's will blow up and probably take out your PC too. They have no overload protection or overvoltage shutdown.

Originally Posted by deanthom

I'm using a M2 on a P4 1.7 system, but only with on-board graphics. I actually measured the current my system draws on each rail with an ammeter while it was at max load. I got 5A on 12V, 2-3A on 5V and <1A on 3.3V. That's with a 3.5" HDD. My USB stuff is powered from a POL. My laptop DVD/CDRW draws about 1.5A on the 5V rail, do you have one of those too? Also, why the Radeon? Can't you use onboard or some other low-level graphics card? If you're not doing any 3D then theres not much point using the Radeon, it will just waste power. Be warned that overloaded M2's will blow up and probably take out your PC too. They have no overload protection or overvoltage shutdown.

That is exactly the info I was looking for...thanks!

I am using the Radeon because I need a card with stable, good RCA signal for my PSOne LCD. I've found my PSOne to be sort of picky about the signals it will display....other graphics cards I've tried have resulted in too much flickering in the display or a completely unusable signal. I bought the Radeon because it was $29 and has a great driver that provides good control over the RCA and VGA out.

How did you measure the current draw while the system was operating? Did you measure the ATX harness wires for each rail wire individually, and then sum them to arrive at the total current (ex: measure each of the 12V+ wires individually and then sum the currents, then repeat for each of the 5V+ wires, etc.)?

If you could explain your current-testing methodology to me, then I can measure the current on the rails with a desktop power supply with and without the Radeon card before installing it in the car.

Finally (and I realize this is a long post), I am well aware of the issues with the M2, as an M2 (likely) destroyed my previous motherboard, processor, harddrive, and screen. If you search my username and "M2" you'll likely find the thread I started on the subject. If there is a reliable way to test current draw on a running system, I think that there should be a sticky/FAQ on the subject. Knowing the current draw of the system on the bench is a much better solution than estimating current draw, and everyone planning a system would be well advised to include an actual measurement in their installation process (especially if using the M2!)


Thanks!

No problem, glad to help out! It seems that systems generally draw less than what the calculator estimates, but as you say, you can never be sure! Unfortunately you must measure the current through all the rail's wires at once, you cannot sum individual ones because the current will not share evenly when you introduce the ammeter. So you either cut the wires and place your ammeter in series or do what I did and use a DC clamp meter. I stuffed all the wires for a particular rail into the jaws of the clamp meter and measured the current. You need a pretty good clamp meter to read low DC currents though!

Yeah, I've come to the conclusion that M2's are a bad design, but I guess that's why they are cheap. I wouldn't have bought one had I know, but you live and learn!

Cheers
Dean

Originally Posted by deanthom

No problem, glad to help out! It seems that systems generally draw less than what the calculator estimates, but as you say, you can never be sure! Unfortunately you must measure the current through all the rail's wires at once, you cannot sum individual ones because the current will not share evenly when you introduce the ammeter. So you either cut the wires and place your ammeter in series or do what I did and use a DC clamp meter. I stuffed all the wires for a particular rail into the jaws of the clamp meter and measured the current. You need a pretty good clamp meter to read low DC currents though!

Yeah, I've come to the conclusion that M2's are a bad design, but I guess that's why they are cheap. I wouldn't have bought one had I know, but you live and learn!

Cheers
Dean

I thought that current sharing might be an issue with individual measurement! Looks like I will be investing in a DC clamp meter.

I tried the "cut and combine" method on a desktop PSU and it didn't work well (although I didn't have a working MB at the time either).