Corsair RM550X and statnge juju with Phanteks DLED controller

DWD1961

[H]ard|Gawd
Joined
Nov 30, 2019
Messages
1,314
This is with my computer completely shut down.

After plugging the Phatek's controller inton a SATA plug, it seems it is causing my PSU to turn on and not shut off after the system is shut down, and keeps power going to a couple of case fans though the SATA connection too. If I unplug the Phanteks' controller, the PSU immediately goes off. Plug the Phanteks' controller back in, and the PSU fan start spinning and my two radiator fans start spinning slowly. This is with the computer off.

Any ideas?
 
Does the computer boot up when you plug in the controller, or just the fans start spinning and nothing else happens?
 
Does the computer boot up when you plug in the controller, or just the fans start spinning and nothing else happens?

Yeah I should have stated that. It was late last night, so thanks for bringing it up.

First, the Silverstone fan hub is using it's own SATA port and cable, and the Phanteks DLED hub is using its own SATA port and cable.

My rig starts fine and everything seems to run okay, including the DLED strips I plugged in just to test that idea. After shutting the computer down, two fans plugged into the Silverstone SATA powered fan hub and the PSU's fan were still spinning. When I unplugged the Phanteks DLED controller, the fans plugged into the Silverstone controller and the PSU's fan stopped spinning. It seems like the PSU is coming on in a very low voltage way, and powering the SATA ports in a very low voltage manner, just enough to get the two fans plugged into the Fan hub and the PSU's fan spinning slowly, but not enough to spin the other fans plugged into the fan hub.

I talked with John Gerow, manager for Corsair's PSU, and he said there are no signal wires in the PSU SATA rails, only in the 24 pin.

I just now shut my rig off and plugged in an SATA drive into the same SATA power cable that the Phanteks controller was plugged into, and nothing came on, which is what you would expect.

Diagram:

SATA Port 01: Silverstone PWM fan hub ---- PWM signal cable --- fans plugged into the fan hub controller

SATA Port 02: Phanteks DLED controller -- DLED signal "Y" wire one side plugged in to the DLED controller -- 2 15" DLED light strips
(Plug this into the SATA cable when teh computer is off, and the PSU fan comes on along with two fans connected to the Silverstone Fan Hub start to spin, but at a reduced RPM. Unplug the DLED controller, and the PSU fan goes off and so do the two other fans.)

Motherboard DLED riser: Y cable connecting the AIO DLEDs and one side connected to the Phanteks DLED controller.

How can plugging in the Phanteks controller turn on the second SATA port that it isn't even connected to?
 
The SATA ports on the PSU are not actually separate. They terminate to the same point inside the PSU, so power being backfed into one port will show up on the other.

It seems like +5vsb power (which is always available when a PSU is plugged in and switched on) is being fed into the Phanteks controller somehow, and then backfeeding through the SATA +12v rail. Does the problem go away when you unplug the motherboard from the Phanteks hub while leaving the SATA plug connected?
 
The SATA ports on the PSU are not actually separate. They terminate to the same point inside the PSU, so power being backfed into one port will show up on the other.

It seems like +5vsb power (which is always available when a PSU is plugged in and switched on) is being fed into the Phanteks controller somehow, and then backfeeding through the SATA +12v rail. Does the problem go away when you unplug the motherboard from the Phanteks hub while leaving the SATA plug connected?
I think it does. I'll do that just to make sure, but I'm almost certain I did do that. However, if the 5V supply is being fed into the controller, then back to teh other SATA port, then that would indicate a problem on the controller side, right?
 
The SATA ports on the PSU are not actually separate. They terminate to the same point inside the PSU, so power being backfed into one port will show up on the other.

It seems like +5vsb power (which is always available when a PSU is plugged in and switched on) is being fed into the Phanteks controller somehow, and then backfeeding through the SATA +12v rail. Does the problem go away when you unplug the motherboard from the Phanteks hub while leaving the SATA plug connected?

Got a reply back from John, the manager of Corsair's power supply department. I told him your thinking on the problem, and he said this:

"The flaw with that logic is that +5VSB is only present on the 24-pin (a single wire). Not on the SATA."

I did the test you requested:

Indeed, when I have the controller plugged into the SATA port alone, without the DLED motherboard riser plugged in, it does not spin up the PSU fan or other fans.

When I have the controller plugged into the SATA port AND the DLED riser, it spins up the PSU fan and two other fans operating from my fan hub controller, which is itself using a different cable and SATA port.

Again, to be clear, when I unplug the DLED wire from the Phantek controller, and leave the SATA power cable plugged in, the fans do not run.

I did not test it with ONLY the DLED riser wire plugged in and the SATA cable unplugged.
 
He’s looking at it from the wrong direction. Yes, The PSU only feeds +5vsb into the motherboard’s 24-pin connector, but the motherboard then feeds it into whatever circuits it wants powered up while the rest of them system is off (e.g keeping a NiC powered for WOL) Some newer boards are doing that for RGB effects when shutdown. The Phanteks designers don’t seem to have accounted for that and didn’t isolate the +5v from the MB from the +5v from the SATA.
 
He’s looking at it from the wrong direction. Yes, The PSU only feeds +5vsb into the motherboard’s 24-pin connector, but the motherboard then feeds it into whatever circuits it wants powered up while the rest of them system is off (e.g keeping a NiC powered for WOL) Some newer boards are doing that for RGB effects when shutdown. The Phanteks designers don’t seem to have accounted for that and didn’t isolate the +5v from the MB from the +5v from the SATA.
Right that was my assumption too. John wasn't considering the motherboard, only the PSU unit. He's only interested if the PSU is faulty, which it seems it is not. His comment means that the MB is definitely involved, as I tested and the results at least tentatively support. That is, without the MB connection, the fans do not power up.

I have my BIOS RGB disabled. Is it possible the MB keeps it on anyway in the background? I just turned off the option for any LEDs to be on when the computer is in any sleep state, and then actually turned off the LED in the fusion part of the bios.

So, if you are correct, the Phanteks controller is allowing voltage to run back through it from the DLED riser, then back into the SATA cable and PSU and then out to teh othert SATA cable to the fans, and even allowing the PSU units fan to spin up? That can't be good in any sense.
 
The only thing that's stumping me about this is that fans usually run on +12v, and the review for the RM750x shows a 12v fan, and the fact that they run slowly indicates they are being undervolted, but how in the world is that +5v backfeed jumping over to the +12v rails.
 
The only thing that's stumping me about this is that fans usually run on +12v, and the review for the RM750x shows a 12v fan, and the fact that they run slowly indicates they are being undervolted, but how in the world is that +5v backfeed jumping over to the +12v rails.

This from Phanteks' technical support: "It looks like your issue is with motherboard compatibility. Our Digital starter kits work with ASUS and MSI motherboards. Gigabyte and ASRock boards seem to do their motherboards a bit differently where it may have issues like the ones you are experiencing."

And, good question about that 5V/12V rail.
Well, what seems to be happening is that the MB is supplying 5V's of power to the controller through the SATA connector and the controller is passing that to the SATA rail. But why would the PSU fan be turning? And, I noticed a couple times the PSU started to come on and then stopped. Anyway, this is a Phanteks' problem. No telling what bad juju they have going on in their when it comes to the Gigabyte system.

Johnny over at Johnny Guru said the MB is giving the PSU the signal to turn on, but didn't offer anything else.

How do case and radiator 12V fans work on a SATA 5V rail?

How can I get this shit to work? I need a controller that can work with my board.
 
Last edited:
I don’t think the PSU is being told to turn on. Otherwise you’d be booting up.

Look in the BIOS settings for a power management setting of some sort that reduces the power draw in standby and shutdown. (If you can tell which motherboard you have, I can try to help you find the setting if it exists).

If you still want to use the Phanteks hub, you’ll have to find a way to isolate the SATA power from the motherboard. Otherwise you have to find a compatible hub.
 
I don’t think the PSU is being told to turn on. Otherwise you’d be booting up.

Look in the BIOS settings for a power management setting of some sort that reduces the power draw in standby and shutdown. (If you can tell which motherboard you have, I can try to help you find the setting if it exists).

If you still want to use the Phanteks hub, you’ll have to find a way to isolate the SATA power from the motherboard. Otherwise you have to find a compatible hub.
The power settings in the BIOS for sleep stats are disabled. In other words, I disabled them being on. I've also disabled the BIOS level fusion DLED lights.

The Motherboard is Gigabyte B450 Aorus Pro wifi ITX: https://www.gigabyte.com/Motherboard/B450-I-AORUS-PRO-WIFI-rev-10/sp#sp

I doubt there is anything in there that will fix it or the tech dude at Phanteks would have mentioned it. I'm hoping Gigabyte can fix it with a BIOS update at some point.
 
I don’t think the PSU is being told to turn on. Otherwise you’d be booting up.

Look in the BIOS settings for a power management setting of some sort that reduces the power draw in standby and shutdown. (If you can tell which motherboard you have, I can try to help you find the setting if it exists).

If you still want to use the Phanteks hub, you’ll have to find a way to isolate the SATA power from the motherboard. Otherwise you have to find a compatible hub.

Have a question for you. Fans are all 12V, so how can the SATA connection power them when they are rated for 5Vs?
 
I swear you have more problems with your computer than anyone else in this forum.
 
Have a question for you. Fans are all 12V, so how can the SATA connection power them when they are rated for 5Vs?

Older PSUs derived the 5v rail from the 12v, so there's a network of components that when powered normally would isolate the two rails, but when fed backwards (e.g 5v without 12v) would allow voltage to show up on other rails. I would think newer power supplies wouldn't have this problem... but I guess it might still be possible. You'd need to ask your Corsair contact how the rails are regulated and what happens when a powered-down PSU is backfed voltage on one of the lower rails.

As for your BIOS, try enabling one, the other, or both of these settings to see if the problem goes away

1587646351726.png
 
I swear you have more problems with your computer than anyone else in this forum.
Tell me about it. I thought it was my imagination. I should have bought the damn ASUS board, and I wouldn't have this issue. PArt of it is becasue I want to do things a specific way. I'm not ok, for instance, buying any fan with it's own controller, tossing the controller into the box or the remote into the drawer, and calling it good. I want everything to work in concert. I'm not doing this jsut to see how many remotes, pieces of 3rd party software, and color I can get. This is a specific project with a idea I had in mind that I want to try and complete. So, yeah, if I were happy jsut tossing in a bunch of shit and turning it all one and saying, "WOW, cool" then I'd not have any issues (e.g., I could have used the Phanteks external controller and avoided plugging the Phanteks controller into the DLED port--problem solved--and for most people, that would ahve been just fine.)

CEC was already disabled but I checked in when I first posted to make sure it was disabled. I'll check the ERP but I think it is disabled. I already disabled power loading and tested it, but no change. I'll check erp now.
 
Older PSUs derived the 5v rail from the 12v, so there's a network of components that when powered normally would isolate the two rails, but when fed backwards (e.g 5v without 12v) would allow voltage to show up on other rails. I would think newer power supplies wouldn't have this problem... but I guess it might still be possible. You'd need to ask your Corsair contact how the rails are regulated and what happens when a powered-down PSU is backfed voltage on one of the lower rails.

As for your BIOS, try enabling one, the other, or both of these settings to see if the problem goes away

View attachment 239861
It was already disabled, along with power loading and CEC.
 
It was already disabled, along with power loading and CEC.
Enable them. They’re supposed to change the power off behavior of the system to reduce energy consumption, which means turning off +5vsb power to extraneous devices like keyboards mice, rgb headers, etc.
 
Enable them. They’re supposed to change the power off behavior of the system to reduce energy consumption, which means turning off +5vsb power to extraneous devices like keyboards mice, rgb headers, etc.
Enabled and did not make a difference. Think I will send the Phateks' stuff back today. Tech was right--incompatible with Giagbyte MBs.
 
Last edited:
I just tried a new generic DLED hub. Same problem as the Phanteks. At this point, it looks like Gigabyte has limited their motherboards to the on board DLED riser, which are 2A each. If you need more amps to run your DLEDs than you have risers, seems yer fucked. In my case, I have only 1 5V DLED riser, so my entire DLED system is limited to 2A. Stellar job implementing your LED power, Gigabyte--incompatible with all controllers.
 
The only thing that's stumping me about this is that fans usually run on +12v, and the review for the RM750x shows a 12v fan, and the fact that they run slowly indicates they are being undervolted, but how in the world is that +5v backfeed jumping over to the +12v rails.
Finally got around to posting your question for John at Corsair:

This is his answer:

"+5V and +5VSB are two different rails. It's the +5VSB that's "always on". That is a single wire on the 24-pin connector.

It is true that some boards use +5VSB for RGB while the PC is in standby. But you're explaining +5V going into a +12V circuit.

Now, I will say this.. IF there is +5VSB being routed back through the +12V via some crazy poor circuit design, it could cause the fan to spin. The RMx doesn't have a PWM fan. "+12V" is a fan's maximum voltage input. The speed of a DC fan is varied by varying the DC voltage. So if you gave the fan +5V, the fan would simply spin at a very slow speed.

But I'd be worried about other damage you're creating. If you're feeding +5V back to the +12V fan in the PSU, what else are you feeding +5VSB to that doesn't use that voltage? I would say that long term, you're going to kill something."

End Quote

So how do I shut off 5V SB?
 
Finally got around to posting your question for John at Corsair:

This is his answer:

"+5V and +5VSB are two different rails. It's the +5VSB that's "always on". That is a single wire on the 24-pin connector.

It is true that some boards use +5VSB for RGB while the PC is in standby. But you're explaining +5V going into a +12V circuit.

Now, I will say this.. IF there is +5VSB being routed back through the +12V via some crazy poor circuit design, it could cause the fan to spin. The RMx doesn't have a PWM fan. "+12V" is a fan's maximum voltage input. The speed of a DC fan is varied by varying the DC voltage. So if you gave the fan +5V, the fan would simply spin at a very slow speed.

But I'd be worried about other damage you're creating. If you're feeding +5V back to the +12V fan in the PSU, what else are you feeding +5VSB to that doesn't use that voltage? I would say that long term, you're going to kill something."

End Quote

So how do I shut off 5V SB?

Yeah, +5VSB is what I was meaning in that post.

As for how to shut it off, you’d have to unplug the PSU from the wall.
 
Yeah, +5VSB is what I was meaning in that post.

As for how to shut it off, you’d have to unplug the PSU from the wall.
LOL. Fuck yeah. Good job Giga-ignorance. I wonder if I could just cut the SB wire?

What does he mean by bad circuit design--the motherboard or the controller?
 
LOL. Fuck yeah. Good job Giga-ignorance. I wonder if I could just cut the SB wire?

What does he mean by bad circuit design--the motherboard or the controller?
If you cut that wire you won’t be able to turn on your system.

controller probably for not isolating the motherboard supplied power from the SATA supplied power
 
If you cut that wire you won’t be able to turn on your system.

controller probably for not isolating the motherboard supplied power from the SATA supplied power
Well, Phanteks doesn't do it either. I'm beginning to think it's more complicated than that becasue it looks like no DLED controller will work with Gig boards. This is ridiculous.

What if I cut the +5v lead on the DLED connection going into the controller, since it has its own power? You would think that the controller would do that if it didn't need it.
 
Well, Phanteks doesn't do it either. I'm beginning to think it's more complicated than that becasue it looks like no DLED controller will work with Gig boards. This is ridiculous.

What if I cut the +5v lead on the DLED connection going into the controller, since it has its own power? You would think that the controller would do that if it didn't need it.
Worth a shot, though I probably wouldn't cut it. Instead remove the pin from the block, fold it back to the cable, and tape it over.
 
Worth a shot, though I probably wouldn't cut it. Instead remove the pin from the block, fold it back to the cable, and tape it over.
You would think if that is all that needed to be done, the power pin would be disabled on the controller. I mean, why leave a powered pin when you are powering with SATA? Maybe to give it more power than the SATA 4.5Amps?

How do you remove the pin from the male or female version of this:

511vI%2B-2XJL._AC_SL1100_.jpg

I just checked a package I got and it has the male pins in it. I guess I can destroy one of them to knock out the power cord and see if that works.
 
Last edited:
Worth a shot, though I probably wouldn't cut it. Instead remove the pin from the block, fold it back to the cable, and tape it over.
I can't believe it! It worked. Can you imagine that? Incredible. I just plugged the data pin into the controller and done. Works fine. No feedback.
 
Back
Top