Unfortunately no one can be told what fun_plug is - you have to see it for yourself.
You are not logged in.
Does the DNS-323 support WOL (Wake on Lan)?
Offline
one would think so if it has the ability to stop spinning the disks when not active
Offline
@masterg - no it doesn't
@fickle - when the DNS-323 spins the disks down the logic is still powered
Offline
Thanks for the reply. If anyone can figure out a way to WOL, do share with the group.
Offline
fordem wrote:
@masterg - no it doesn't
I am curious as to why it doesn't. Is it because of the 88F1111 chip that doesn't support it or the main process, or is it just a software problem? (No driver, or not supported by the kernel?)
Offline
As far as I can tell the 88F1111 chipset supports some form of WOL - there is however, some doubt as to whether the current driver supports it
Offline
fordem wrote:
As far as I can tell the 88F1111 chipset supports some form of WOL - there is however, some doubt as to whether the current driver supports it
I've been digging around in the 2.6.12.6 kernel code recently but haven't really looked for WOL yet. What do you know about the driver? Is it a module or is it built into the kernel?
Offline
It appears that the 88E1111 PHY circuit has no power on its supply rails when the device is off. Thus I think that WOL is impossible, unless I missed some small converter.
Offline
bluesmoke wrote:
It appears that the 88E1111 PHY circuit has no power on its supply rails when the device is off. Thus I think that WOL is impossible, unless I missed some small converter.
I wondered about that - it shouldn't be too hard to fix for someone so inclined. There is some level of power supplied to the logic even with unit switched off, because the front panel button is a "soft' switch, something has to detect the contact closure.
What I will say at this point is holding the button down and applying power does nothing, but applying power, waiting 1 second and holding the button for 1 second will power it up and that's the basis of a small timing circuit I'm developing to get "auto power on"
Offline
I think it works like this: The power from the power brick (+5V and +12V) is cut using two PMOS load switches to the buck converters. They are in turn controlled by a small latching circuit that is powered directly from the incoming power and also connected to the front panel button and main processor (for turn-off (my speculation)). Thus, only this latching circuit has constant power (and the RTC, but that comes from the battery) I suspect that the 88E1111 shares its power rails with the main processor and most of the other circuits with the same voltage. Unfortunately, it also appeares that the alarm output from the RTC is not connected to this latching circuit, this could otherwise be used to implement a timed turn on.
Offline
fordem wrote:
I wondered about that - it shouldn't be too hard to fix for someone so inclined. There is some level of power supplied to the logic even with unit switched off, because the front panel button is a "soft' switch, something has to detect the contact closure.
What I will say at this point is holding the button down and applying power does nothing, but applying power, waiting 1 second and holding the button for 1 second will power it up and that's the basis of a small timing circuit I'm developing to get "auto power on"
Interesting, an auto powerup circuit is also on my todo list, although if i had WOL i wouldn't have to worry about an auto powerup circuit. Looks like i'm subscribing to this thread
Edit:
I'm still curious to find out which chip the main power button is actually connected to. If the button can power up the unit then obviously the chip the button is connected to must have power when the device is off. I'm yet to pull out a multimeter, but i speculate that the processor has power and sits in some kind of low power state with or without the processor clock running? or does the button just enable power to be supplied to the processor via some other chip, maybe that latch?
Also would it be possible for someone to post or PM me the link to this 'other documentation' that was discussed in one of the WOL threads.
Last edited by orbitaudio (2008-02-07 23:31:37)
Offline
If WOL was available, I'd simply have the same server that shuts the DNS-323 down, send the magic packet - but alas - that is not to be, so sometime in the not too distant future .....
Offline
The power button is connected to the circuit surrounding and including the SO14 (74AHC74D=dual D-flipflops) that is adjacent to the HDD0-connector. This circuit is powered at all times (if power brick is connected).
Last edited by bluesmoke (2008-02-08 09:13:10)
Offline
orbitaudio wrote:
Also would it be possible for someone to post or PM me the link to this 'other documentation' that was discussed in one of the WOL threads.
Offline
Thanks fordem - that's the document I was referring to in the other thread, notice it's v8.41.2.3 where as the documentation with the dns-323 kernel driver is for version v6.23.
Offline
bluesmoke wrote:
It appears that the 88E1111 PHY circuit has no power on its supply rails when the device is off. Thus I think that WOL is impossible, unless I missed some small converter.
Correct: there needs to be some components added to support WOL ..
so the exiting devices: no go ..
or you have to do some engineering yourself ;-)
Offline
Well - just in case anyone is interested ...
The first draft of the auto power on circuit was still born - it was designed on the expectation that the power button would be a momentary contact, normally open switch and would pull a logic signal low - it doesn't. it is a momentary contact normally open switch, but one side is connected to the +12v rail so closing the contacts pulls something high.
Somewhat unusual, especially in today's world of low voltage, low power processors - but hopefully nothing more than a minor set back - in theory, simply inverting the circuit and changing the transistor polarity should solve it.
Offline
fordem wrote:
Well - just in case anyone is interested ...
I am *very* interested in auto power on ...
The first draft of the auto power on circuit was still born - it was designed on the expectation that the power button would be a momentary contact, normally open switch and would pull a logic signal low - it doesn't. it is a momentary contact normally open switch, but one side is connected to the +12v rail so closing the contacts pulls something high.
I am wondering if there is a software solution to do this problem. Given whatever happens when power is applied, the system is waiting for the button to be depressed. At what level is this done? If it is kernel, then perhaps this could be patched and would be a relatively safe hack.
If this happens before the kernel gets control, that is the loader level then perhaps if that is open source it too could be patched although at a greater risk compared to kernel patch for testing purposes.
Jaya
Offline
i'm wondering how far you got with hacking this fordem (and others)?
sorry for the stupid question but what happens when you shutdown dns-323 manually? does it turn back on instantly?
i'm asking because the only time my dns-323 shuts down is when windows computer connected to ups sends a shutdown command. but after the power outtage i would still like access the nas. problem is i don't think there are any solutions to this besides wol (auto start wont help because there is still some battery life left in the ups - it'll re-start, drain the battery and shutdown incorrectly).
dang d-link why no wol
Offline
I never found the time to get beyond the sketch on an envelope stage, but someone else did develop a circuit using a hex inverter and a couple of caps.
http://dns323.kood.org/forum/t1602-Auto … up%3F.html
Offline