Looking for a creative solution: USB bus still powered after CPU is off - Page 5

dieselBenz · 05-26-2004, 03:55 AM

Anyone tried cutting the +5 standby wire? Does the epia MB need this to boot? I figure, if the 90W opus has a jumper to disable 5V standby, it probably does just the same... ?

Scouse Monkey · 05-26-2004, 05:20 AM

Quote: Originally Posted by rando

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

.......................

My god, i suggested something electrical that works - thats novel! I think i was better at electronics before i did my engineering degree!

I was thinking about using two relays so the ACC line and opus powered there own but thought that was just a waste of effort and power. One thing you could do is add the second relay as well (just powered by the opus accross the 12V input in parallel to the 1st as a backup - could be a small 5V relay running off the opus 5V line) but thats the aero engineer in me seeking redundancy so the opus supply is not cut if there is a problem with the main circuit.

I wondered if you would need a delay for the ACC line as the 12V power would come on just after but obviously not! Does it just need to be high for the Opus to turn on then?

rando · 05-26-2004, 09:14 AM

From what I can tell, the Opus turns on if (and only if) both ACC and +12v are high. It doesn't appear to care about the order. I believe, but did not measure, that the turn-on dealy is slightly longer when ACC goes high before +12v has stabalized. Maybe Kris from Opus can comment.

rando · 06-26-2004, 04:21 PM

Update .... this did eventually cause some problems. On occassion, the Opus failed to turn on. I assume this means Ricky was right. The Opus probably does need to see the +12V line go high (at least) a bit before ACC.

Quote: Originally Posted by dieselBenz

Anyone tried cutting the +5 standby wire? Does the epia MB need this to boot? I figure, if the 90W opus has a jumper to disable 5V standby, it probably does just the same... ?

Yes I did this. Without the +5V standby power, the M10k will pull the PSON signal low. This will cause the Opus to turn on. The Opus will then notice that ACC is low and turn back off. The pattern will continue (on, off, on, off, ...). To avoid this, you need to interrupt the PSON signal. What I did was put a DP relay on +5 standby and PSON. The relay is triggered as in ScouseMonkey's diagram. So far, this seems to work. I don't think this solution will support standby/resume mode.

Ricky327 · 06-26-2004, 04:26 PM

Ahh thanks for the confirmation on that.

I guess its time to insert a delay circuit of some sort...something similiar to coyotes autoturn on mods on lilliput may do the trick.

The ACC input is held low temporarily until the capacitor is charged up and triggering the OPUS to turn on.

UncleGeek · 06-26-2004, 09:53 PM

If I am following this thread correctly, why not abandon the +5v. for the USB from the mobo. Then use a powered USB hub and use the +5v. from a spare Molex connector from the Opus leaving the power connected to the PSU and eliminate the relays altogether?

Micco · 08-13-2004, 11:41 PM

Was there ever any final 'official' fix to this?

Did anyone ever build Scouse Monkey's circuit with a delay circuit on the ACC line to confirm?

Thanks!

rando · 08-13-2004, 11:47 PM

Yes, see post #64 above. No delay needed, just a Double Pole relay on the +5VSBY and PSON.

archimense · 08-14-2004, 03:35 AM

Quote: Originally Posted by rando

Yes, see post #64 above. No delay needed, just a Double Pole relay on the +5VSBY and PSON.

Would it be possible to get a diagram of this? I'm having a hard time understanding your description.

Paul_ · 08-20-2004, 05:04 PM

can't we just power the usb 5v line off the 5v+ from the molex plug?

RogerWilco · 08-20-2004, 10:39 PM

just use the cigar lighter to usb mod.

archimense · 08-21-2004, 05:29 PM

Quote: Originally Posted by rando

Update .... this did eventually cause some problems. On occassion, the Opus failed to turn on. I assume this means Ricky was right. The Opus probably does need to see the +12V line go high (at least) a bit before ACC.

Yes I did this. Without the +5V standby power, the M10k will pull the PSON signal low. This will cause the Opus to turn on. The Opus will then notice that ACC is low and turn back off. The pattern will continue (on, off, on, off, ...). To avoid this, you need to interrupt the PSON signal. What I did was put a DP relay on +5 standby and PSON. The relay is triggered as in ScouseMonkey's diagram. So far, this seems to work. I don't think this solution will support standby/resume mode.

So am I correct in saying that the only modification you did was:

Put a DPST relay on PSON and PSBY so that they were only connected when the ACC line was high or the computer was on(HDD power)?

RogerWilco · 08-21-2004, 08:17 PM

why mod your 200$ psu, and invalidate the warranty, when you could simply get a molex, run a switch, and bypass all of these problems ? Am i missing something, you can also buy extra cigar lighters, and run them descreatly (spelling) so they will be out of view. Sorry if im missing somethin.
Thanks!

archimense · 08-21-2004, 08:24 PM

Quote: Originally Posted by RogerWilco

why mod your 200$ psu, and invalidate the warranty, when you could simply get a molex, run a switch, and bypass all of these problems ? Am i missing something, you can also buy extra cigar lighters, and run them descreatly (spelling) so they will be out of view. Sorry if im missing somethin.
Thanks!

I'm not modding the PSU, I'm putting a relay inline with the ATX power cable.

Molex cables don't do power load balancing, handle startup spike, or prevent shorted usb devices from drawing too much power.

Not to mention this mod will allow the motherboard to run normally, as opposed to hacking in a seperate power supply for USB devices which doesn't provide all of the USB power protection.

Also the mod you are suggesting can only be done on USB ports that have a header on the motherboard. The EPIA M10000 has 6 ports, 4 with headers, and 2 on the back panel. Your mod would not allow you to shut down the 2 on the back panel, unless you want to disconnect and resolder the connectors on the back panel.

PS: Stop posting, let your computer rest....

RogerWilco · 08-21-2004, 08:29 PM

haha lol its had plenty of rest! for the last 8 hours almost ! i went to a family reunion, about an hour from my house. Anyways, i was unaware (seeing as ive never used one) that the molex connector was unregulated, i was under the impression that it was. Alright looks like i'll stray from ever using one in the near future. My bad there i misunderstood how you had run the relay, did you use a ATX12v extenison cable, and not directly cut the one from the psu ?
Thanks!