Looking for a creative solution: USB bus still powered after CPU is off - Page 4

Ricky327 · 04-15-2004, 03:10 PM

Just in case anyone is looking to do this with an inverter, I wired up a relay on the HDD output accross the on/off switch of my inverter to do basically the same thing. I also put a relay accross it which was triggered by the ignition in the on position. My old MB automatically booted when power was supplied, so this worked well as an auto on feature.

Good post...Im sure alot of inverter ppl will follow the idea

Riff · 05-19-2004, 05:13 PM

Okay, I'm going to need to do the diode/relay mod. What kind of diode and relay will I need (links to radioshack models would be very helpful)? Would these work:

Diode

Relay

I'm a total electronics noob, so any noob advice appreciated! Thanks for your help.

NoPistonPC · 05-20-2004, 01:17 AM

Does anyone with an EPIA motherboard have this problem? I'm surprised this issue hasn't come up before. It seems like a significant one...

gospeed.racer · 05-20-2004, 01:28 AM

I have a epia M10000 and my solution for the usb hub staying powered up is this. And, keep in mind, mine is internal so accessing the usb wire was much easier than if it were external, but the same theory. All of my USB's appear to have constant power without this relay setup.

Cut the usb's red wire that runs to the motherboard, and connect it to the switched portion of the relay. Normally open contacts

Wire the relay coil to one of the molex connectors...5v or 12v depending on which relay you got. Mine was 5v.

When you turn on the pc, it fires up the buss and thus turns on the hub. Mine works great, never had a single problem. I have not had to install a cap or diode either. Your mileage may vary!

NoPistonPC · 05-20-2004, 04:14 PM

gospeed, that's too bad, i was hoping i wouldn't have to worry about this.

Do any of the standalone shutdown controllers have this problem? I will be using a regulator-bypassed ITPS in conjuction with an Arise 810H PSU. I posted my setup on this thread: http://www.mp3car.com/vbulletin/show...295#post206295

I know the ITPS pulses the reset button on the mobo on shutdown, but doesn't it also completely cut power to the 12V line going to the PSU (after the 45 second delay)? If so, I wouldn't have to worry about the USB current draw issue, correct?

rando · 05-20-2004, 04:59 PM

OK, so there are a couple of issues with Ricky's mod.

First, his circuit (initially) powers the opus from the car's ign/acc line. That may be fine on some cars, but in general, is a bad idea. You don't want to try to draw (upto) 15A from your ignition line. Since it's just used to bootstrap the Opus, it probably only draws a few amps max. Just the same, that's a heavy load as far as the ACC line is concerned. That's why most of us have run a custom circuit to supply the 12v line on the opus.

Second (and probably worse), once the Opus turns on, the 12v line and ign lines in your car will be shorted. The 12v line is probalby a high current line with large guage wire. The ignition circuit (being a low current circuit) is almost certainly using 18-24guage wire. Ignition circuits are usually protected by 15A or less (7A in my truck) fuses. If an ACC line shorts anywhere in your car, it is liable to release magic black smoke since the ACC fuse will no longer protect it. THIS IS BAD. VERY VERY BAD.

There is an easy solution. Use two relays. The switch side of the relays are in parallel and switch the 12v line. Hook up one of the coils through a 12v output of the opus. Hook up the other coil to the ACC line. Throw in a couple of diodes to block EMF if you like.

Instead of the second relay, you could also try throwing in a couple of diodes to assure that the ACC and 12v lines stay isolated. This would have the annoying side effect of lowering your 12v line input voltage and thus the opus' low batter protection circuit would then kick-in sooner.

Good luck.

Ricky327 · 05-22-2004, 09:11 AM

Rondo,

Glad you found something wrong with the mods before someone smoked up

As I said its not tested and I didnt even gave much thought into it.

Im not so sure how your mods are all wired up, please draw the diagram for us

Not that I need it, Im just curious how it works...maybe someone can use it.

Pokey · 05-22-2004, 09:54 AM

I kinda had the same thing happen to me, I had one computer that WOULD NOT power off, i was sitting there with the power cord in my hand, and that stupid LED on the mother board was still lit. Long story short, power was coming from another computer via firewire powering the +5v on that computer.

Anyway, most cars now have current being drawn from them when the key is off.. ie car's internal computer, clocks, etc. I would measure how much is being drawn via your USB, it might not be that much. You could also take the hot (or ground) wire off the battery and measure the total current draw of your vehicle while the key is off. If your battery is being killed by this little amount of current, you might want to check and see if your battery might be getting old.

Scouse Monkey · 05-22-2004, 07:04 PM

I was wondering if this would work:

I modified Rickys drawing (hope you don't mind) to save time (see pic). I set the relay on the OPUS 12V in to be powered by the IGN ine and the OPUS 12V output (eg HDD power connecter) so whn your IGN turns on it not only switches on the power to the OPUS but also triggers it into booting via the IGN line into the OPUS. When the IGN is turned off the OPUS continues to power the relay thus self sustaining itself while the drop in the IGN triggers the shut down. When the shut down turns off the OPUS 12V output it then kills its own power.

Not really familiar with the way the OPUS works and electronics are not my strong point so PLEASE COMMENT AS TO IF THIS WILL WORK. I put the diodes into the relay inputs to isolate them - think they are the right way round.

Andy

rando · 05-25-2004, 01:35 AM

Scouse Monkey's circuit looks good to me. You can also use 2 relays connected as below.

Connect ACC line directly to Opus ACC input
ACC--------------OPUSACC

Connect +12V battery to Opus through relay switch. Switch is controlled by ACC line.
ACC-----COIL1------GND
+12V----SW1------OPUS+12VIN

Connect +12V battery to Opus through relay switch. Switch is controlled by OPUS +12V output.
OPUS+12OUT-------COIL2---------GND
+12V----SW2------OPUS+12VIN

Ricky327 · 05-25-2004, 08:32 AM

Scouse Monkey,

Yes your circuit look good as it eliminate having to use 2 relays. It look like its the best solutions so far.

Rando,

I can see how your circuit working now, thanks for the diagram.

Both circuit should work but as I mentioned on the original diagram a delay is most likley needed when triggerring the OPUS ACC. On both circuit the OPUS ACC get the +12V first before the OPUS +12V.

Ill say a proper operation should be :

+12V to OPUS +12V supply
1 second or so delay
+12V to OPUS ACC input

I dont know how much current the OPUS ACC input takes in and what voltage it triggers at. I cant see it taking so much current in so a resistor and capacitor should work fine to delay the ACC trigger.

Ill like to see a final circuit working, Sorry I cant help out on testing as I dont have an OPUS.

Keep the ideas comming in...this what keep the forum going

rando · 05-25-2004, 12:53 PM

Are you sure the delay is required for the Opus? I'm going to wire mine up with the 2 relay solution today. I'll let you know if it still works.

Ricky327 · 05-25-2004, 07:28 PM

Are you sure the delay is required for the Opus? I'm going to wire mine up with the 2 relay solution today. I'll let you know if it still works.

No, Im not sure but its something I expect. But please do try it without any delays and see how it goes

rando · 05-26-2004, 01:25 AM

OK, I wired it up with 1 relay and two diodes ... basically the same as Scouse Monkey's circuit minus the diode across the relay. With the key off, the Opus is completely powerless. With the key turned on, the Opus comes on and then fires up the CPU after a few seconds. When I shut the key off again, the Opus signals the CPU to shutdown. Once the CPU turns off, the Opus turns COMPLETELY off again.

At least in this case, the Opus doesn't appear to have any problem with the ACC line going high before the +12v line.

As for the diode across the relay, I don't think it's needed since both lines feeding the relay are already protected by their own diodes. The EMF generated by the coil field collapsing has nowhere to go. Then again, I may be totally wrong about that.

Ricky327 · 05-26-2004, 03:22 AM

Nice one,

I guesss the modified circuit diagram can be redrawn and finalised everything for the benefits of other user.

On the diagram its probably easier to tap the +12V output of the OPUS from the HDD/CDROM supply. Not so sure about the diode across the relay either in this case. Putting an extra diode wont hurt, its better to be safe than sory until someone can confirm it is safe without it.