Differences
This shows you the differences between the selected revision and the current version of the page.
| hardware:autopoweron 2009/12/03 03:44 | hardware:autopoweron 2017/09/06 18:38 current | ||
|---|---|---|---|
| Line 5: | Line 5: | ||
| Here's a circuit using a NE556N dual timer IC that automatically turns the DNS-323 on whenever power is first applied or re-applied after power failure. This circuit is applicable to the DNS-323 hardware version B1: | Here's a circuit using a NE556N dual timer IC that automatically turns the DNS-323 on whenever power is first applied or re-applied after power failure. This circuit is applicable to the DNS-323 hardware version B1: | ||
| {{ hardware:delayed_on_pulse.png |Delayed turn-on circuit}} | {{ hardware:delayed_on_pulse.png |Delayed turn-on circuit}} | ||
| + | |||
| ==== For older DNS-323 hardware versions ==== | ==== For older DNS-323 hardware versions ==== | ||
| Line 46: | Line 47: | ||
| There is also another auto turn on circuit available, [[http://www.waggies.net/Ken/dns323/index.htm |this one using a hex inverter]]. | There is also another auto turn on circuit available, [[http://www.waggies.net/Ken/dns323/index.htm |this one using a hex inverter]]. | ||
| + | ====== Single capacitor solution (B1 hardware) ====== | ||
| + | **Background info:** The power on button is connected to pin 1 of HT48R05A-1 (U25) and has a weak 120k pull up. The H48R05A-1 is an 8 bit micro-controller from Holtek which conveniently handles very slow rise time inputs. | ||
| + | **Solution:** With such a weak pull up value, all you need is a (fairly large) capacitor to delay the rise time, which tricks the HT48R05A-1 in thinking the power button has just been released. | ||
| + | |||
| + | |||
| + | ===== How to do the mod ===== | ||
| + | |||
| + | - Disassemble your DNS-323 (4 screws on the front, 4 screws in the back, another set of 4 screws in the back, and you're done). | ||
| + | - {{ hardware:power_on_switch_bottom.jpg?300|}} Solder a 150uF (6.3V or higher) electrolytic capacitor between the ground (circled in black) and the power on pin (circled in red). | ||
| + | * **__NOTE: electrolytic capacitors are polarized!!!__** | ||
| + | * Make sure the negative is soldered to ground (the side of the capacitor will have a large stripe next to the pin). | ||
| + | * About the capacitor... I am pretty sure a lower value (ex.: 68uF) will also work, but I didn't try. I have a feeling high ESR capacitors are better since a 15uF __unpolarized__ capacitor (used for AC relays) worked fine yet it's 15 uF electrolytic equivalent didn't. | ||
| + | - Reassemble your DNS-323 and enjoy. | ||
| + | |||
| + | **Note:** The power switch still works to power off the unit, and to power it back on if it was powered down by the switch or by software shutdown.\\ | ||
| + | |||
| + | \\ **Limitations** | ||
| + | * Doesn't work for power glitches (anything below 3-4 seconds). | ||
| + | * {{ hardware:power_on_optional_diode.jpg?300|}}**To fix this limitation**, you can put (in theory since I didn't try it) a diode in parallel to the weak pull up (R209, which is near U25). When the 5V rail fails, the capacitor should discharge very quickly. The N side of the diode should be near the "R" of "R209". | ||
| + | * If you are serious about your NAS, it should be on a UPS (which filter glitches) and therefore you don't need to add a diode... | ||
| + | |||
| + | ~~CLEARFLOAT~~ | ||
| + | **Assembly pictures**\\ | ||
| + | I stole an electrolytic capacitor from a unused electronic circuit. The leads weren't long enough, so I had to use wire. The electrolytic capacitor you'll buy will likely have leads long enough so you won't need to go through this hassle. (But make sure that non-isolated leads don't touch front panel after assembly. The panel is actually made of (anodized) aluminum and there can easily be a short-circuit, leading to router powering down soon after start up - or worse.) | ||
| + | \\ {{hardware:power_on_capacitor.jpg?300|}} {{hardware:power_on_final_result.jpg?500|}} | ||
| + | ~~CLEARFLOAT~~ | ||
