Freevo Hardware Build Log

(Part of the BuildingFreevo set.)

2005-03-12: all the parts arrived yesterday, except for the hard drive. That means it's time for a test run to see if it can get into the BIOS screens at least.

So I connected up the lot -- mainboard, power board, 12V converter, monitor, keyboard, pvr-350 and riser, and plugged in the power. But nothing happened. :(

When I plug in the AC/DC converter, the keyboard lights light up for 1 second, then settle down to a gentle glow in all 3 lights and stay like that (ie. the normal "lights off" state for this keyboard). the display never powers on. ouch.

it could be the AC/DC converter, I think. next step is to cannibalize an AT/ATX PC case, and try with a proper PC PSU, and with speakers etc. so I can here any BIOS POST tones (if there are any, which I suspect there are); http://www.techadvice.com/tech/B/BeepCode.htm . It's a generic Phoenix AwardBIOS according to the screenshots in the manual, so that should be helpful...

http://www.phoenix.com/en/Customer+Services/BIOS/AwardBIOS/Award+Error+Codes.htm http://www.phoenix.com/en/Customer+Services/BIOS/BIOS+FAQ/default.htm#Q12

2005-03-14: got hold of an ATX case and PSU, and hooked that up. The only things connected are:

The PSU is connected to some existing peripherals in the case.

When the power button is pressed, the periphs come on (indicating the board is signalling the PSU to bring them online); the kbd lights blink once; but the CRT never receives any signal. In addition, there's no POST beeps through the speaker.

When I hold down the power button for 5 seconds, the PSU and peripherals power down. So the board can at least do that.

I've tried without the DDR DIMM in the slot -- the same thing happens, and there's no POST beeps in that case either. This leads me to think that it's not the RAM that's at fault.

Booting up and holding down a key on the keyboard doesn't produce the usual beeps caused by the input buffers filling.

Given all that, I think it's the mainboard that's at fault.

2005-03-15: the vendor (Ituner Networks Corp) agreed. mainboard RMA'd and shipped back via UPS Ground for replacement. Fingers crossed...

2005-03-22: "RMA will be processed tomorrow, you will receive an email with a new tracking number".

2005-03-23: "no sign of RMA; will check status and email you".

2005-03-24: now a week since RMA was delivered. "no problem found with the board; testing with 100W PSU and PW-200-M. but will ship a new board later today."

2005-04-09: back after a long absence -- replacement RMA was shipped and received by my housemate. Busy day....

(Just noticed; the board has "PW-120-M" printed on it, not 200-M! However from googling it looks likely that the board was simply renamed at some time to better reflect it's max output.)

Could it be the Coby 12VDC adapter?

THE AC-DC ADAPTER AND WATTAGE

Looking around at other shop sites, I notice that, where listed, the AC-DC 12V adaptors generally seem to provide 5A or 9A output. (e.g.: http://www.mini-box.com/s.nl/category.13/it.A/id.193/.f ) The Coby AC-DC adapter I'm using has a max current of 1000mA:

    MODEL: CA-44   Input: 110/220V 60/50Hz  Power: 18W
    Output: 1.5-3-4.5-6-7.5-9-12 VDC   Current: DC 1000mA max.

I think it's underpowered. The AC-DC 12V-5A adapter listed at mini.box.com above also notes a 60W max output, compared to 18W above.

A refresher course in Ohm's law seems to confirm this. I want something like 110W to 200W of power, at least over 60W, and I'm providing 12V input with 1000mA of current. Given Ohm's law via http://science.howstuffworks.com/question501.htm -- P = VI (power in watts = voltage in volts * current in amps) -- that's giving a puny 12W of current. it's not easy to find doco - the VIA manual doesn't list it, but the power calculator at http://resources.mini-box.com/online/powersimulator/powersimulator.html lists the EPIA-M 6000 mainboard's current draw as 16.2W.

OK, back to basics on this one. using the mini-box.com power simulator for EPIA boards with this hardware list:

that comes out to a draw of 52.2W and 4.35A (it lists Amps required! why didn't I find this page before?!) in idle and 57.9W/4.83A for "office apps" usage. It also suggests 2 power adapters; one of 5A and one 9.5A.

OK, all signs are pointing to requiring a 5A power adapter.

(worth noting for future use: if an additional hard drive is added, the ampage requirements go up over 5A, so a 5A power adapter may not be enough in that case. however right now I'm fine with that.)

That still doesn't explain what went wrong with the previous board; maybe it was bad to start with. Anyway, onwards.

BACK TO GETTING THIS THING UP

So back to the AT PSU until I can get to a shop for a 12VDC/5A adapter.

attached HD and booted; HD is found.

entered BIOS (with Del key) and set:

booted and waited for full POST mem check, which passed. reset the quick POST back to enabled.

next step: inserted the Hauppauge video card and PCI riser and powered on. no video output! same without riser. changed BIOS to "select display device: CRT" only.

still failed to produce video output. after repeated reboot attempts, I noticed that sometimes on a successful boot (without the card installed), one of the IDE devices attached to power would not power up immediately... I eventually managed to get the board to power up with the card inserted -- and again it was on one try when the IDE device was delaying it's power on. so once again it was a power issue.

There were several devices attached to the PSU -- a ZIP drive, a HD, a floppy drive and 2 CD-ROM drives. unplugging the ZIP, HD, and one of the CD-ROMs left the machine easily able to power up.

I attached the HD, and installed the PVR-350 on the PCI riser, booted, and it came up cleanly to attempting to boot off the (as yet unformatted) HD. last thing was the USB wifi card; I hooked that up just for completeness. success -- the hardware was now all installed!

Final BIOS tweak: AC Loss Auto restart: On. This ensures that using the soft power switch to power on the box is not required; the board will come up automatically once power is supplied.

Next step -- FreevoSoftwareBuildLog

FreevoHardwareBuildLog (last edited 2007-04-27 22:12:46 by proxy)