Corsair RGB LED hub: Melted Wires

Hey,

 

Wondering whether anyone else has ran into this issue...

 

Today I moved my partners system into a 570x upon starting up the SATA power cable on the hub started to smoke and melt. It had 6 fans connected to it.

 

I then connected just two fans to a new hub and a different SATA cable and also smoked and melted.

 

Anyone else had this issue?

 

http://i.imgur.com/6o09c3C.jpg?1

Wow. Never saw this happen. Would need to see the rest of the wiring to guess. When you say 6 fans were connected do you mean the LED wire from the fans?

 

What type fans were connected?

 

What were the fans' power wires plugged in to?

 

For 2 of them to do this something is very wrong.

Hi there! This appears to be an issue with the PSU. I, and several of us here at CORSAIR, have built many rigs that included these, none of them have burned out. However, I've personally built just as many and I've seen this happen several times, it always points out to the PSU cable and/or the PSU itself. Seeing as you connected two different units using two different cables, it's definitely the PSU itself as it appears to have malfunctioned. If it is a CORSAIR power supply, please contact our support and we'll get you sorted! Edited August 7, 2017 by Corsair Blake

Hi there! This appears to be an issue with the PSU. I, and several of us here at CORSAIR, have built many rigs that included these, none of them have burned out. However, I've personally built just as many and I've seen this happen several times, it always points out to the PSU cable and/or the PSU itself. Seeing as you connected two different units using two different cables, it's definitely the PSU itself as it appears to have malfunctioned. If it is a CORSAIR power supply, please contact our support and we'll get you sorted!

 

Thanks for your response, yeah I'm thinking it is the PSU too - which is indeed a corsair ax850. Strange though as the PSU is acting normal with everything else, apart from trying to send the control hub straight to hell.

 

I'll get in contact with support and take it from there.

 

Thanks for the help

The melted wires are going to be from over-current. Actually probably -not- a PSU issue, since current is a PULL (draw), not a PUSH (That would be voltage).

 

Either one of the fans had a problem or something was connected incorrectly, since it's relatively unlikely for two separate hubs to have internal shorts

You have a ground fault somewhere. Check the fan cables to check and see if any aren't frayed, pinched or exposed to the bare metal on the case. Then check the fans individually in case one of them has shorted internally.

 

Looking at the hubs it most likely the fan plugged into the #2 socket(hub on left), with hub on right it's hard to tell since #4-6 are pretty burnt with #5 being the worst of the 3.

Edited August 10, 2017 by Zatarot

You have a ground fault somewhere. Check the fan cables to check and see if any aren't frayed, pinched or exposed to the bare metal on the case. Then check the fans individually in case one of them has shorted internally.

 

Looking at the hubs it most likely the fan plugged into the #2 socket(hub on left), with hub on right it's hard to tell since #4-6 are pretty burnt with #5 being the worst of the 3.

 

The connections on either hub don't have any visible burning in person - it is just the SATA wires that have burnt/melted.

 

The melted wires are going to be from over-current. Actually probably -not- a PSU issue, since current is a PULL (draw), not a PUSH (That would be voltage).

 

Either one of the fans had a problem or something was connected incorrectly, since it's relatively unlikely for two separate hubs to have internal shorts

 

I'm doing a full tear down this weekend and will be testing everything out of the case and with a different PSU so will update to let people know.

Edited August 10, 2017 by Krowzer

What are those black marks on the pins?

Looking more closely at the image, it looks like only one side of the wires was really melted, which does lend credence to a ground fault. That would be power running from the +5V (the melted wire) to the case or somewhere else that it should not be going. The lack of melting on the ground wire of the pair shows that there wasn't an overcurrent situation going on there and the power was taking a different return path.

 

Be very cautious... The wire you see there is not the only thing that can be damaged by overcurrent. A different PSU without correcting the underlying cause could easily change to being a dead PSU. Examine the fan wires very carefully (Completely unmount the fans and check for pinches against the case as well), and make sure everything is plugged in correctly from the LED cables on the fans to the hub.

Right.... so an update and the testing I've done so far as I'm still waiting for a response to my support ticket a week later.

 

I added a new EVGA PSU and control hub to the system and did a full tear down and check of all wires. There was no sign of burning or melting on any other wires and no grounding issues could be found. 'Great' I thought and now have a fully working system.

 

However, I decided to retest the two melted/burnt out control hubs with the old corsair ax850 PSU which was installed outside of the case. Upon testing the same issue came back with the wires of the control hub started to over heat when ever the fan RGB connecter was plugged in. No over heating of the wires happened when it was just the 3pin connected.

 

So I thought 'well this happened on two control hubs, so must be the ax850 PSU....' well I happen to have a second ax850 PSU. So tested the second ax850 PSU and the same over heating/melting of the control hub wires happened again. Different SATA power cables were tested again.

 

So in summary my testing leads me to believe there is some issue between Corsair ax850 PSU's and the control hub. For reference the serial numbers for both PSU's are (R1112BA1620407 Rev: A1W & R1112BA1620408 Rev: A1W.

 

So yeah, any other ax850 owners using the control hub successfully?

Edited August 13, 2017 by Krowzer

Let's dig into this a little more deeply then... *Digs out the opened-up hub* Hmmm...

 

Looking at the picture...

Please confirm:

Only one of the two wires involved in the SATA connector is melted on the outside.

It is the BLACK-only wire, not the wire with the white blocks "printed" on it. If you are looking at the top of the hub and the SATA cable is pointed away from you (like the alignment in the picture), this would be the wire on the LEFT coming out of the housing of the hub.

 

If that is true, I have excellent news:

You don't have a ground fault going on. That -is- the ground wire back to the SATA connector and if only that is melted, and the +5V is fine, then it's not over-pulling current on +5.

 

Please confirm or deny that's the situation, and if that is the case, let us know:

What did you have plugged into the ports of the hub? Describe the things and the connector plugged in for each. Take a picture of the thing and connector if you want to be extra-thorough. For example, what kind of fans? Which connector of the two on the fan? Did you plug strips into the hub? Etc.

 

The only way it's likely to be a problem with the PSU specifically is if the PSU is running +12V to that instead of ground, which would fry anything and everything connected to the SATA connector, and since you're not indicating that everything was fried. That sort of thing happens when the wrong cable is used for a modular PSU (for example, a cable for a different brand of PSU).

 

There are a few other things that could be wrong, for example, a loose connector inside the wiring. Though less likely to occur on two separate hubs, if they were both acquired from the same place, they could both be from the same batch. Heat travels well along the metal of the wire. However it doesn't travel perfectly and would cause more burning/melting closer to the fault.

Let's dig into this a little more deeply then... *Digs out the opened-up hub* Hmmm...

 

Looking at the picture...

Please confirm:

Only one of the two wires involved in the SATA connector is melted on the outside.

It is the BLACK-only wire, not the wire with the white blocks "printed" on it. If you are looking at the top of the hub and the SATA cable is pointed away from you (like the alignment in the picture), this would be the wire on the LEFT coming out of the housing of the hub.

 

If that is true, I have excellent news:

You don't have a ground fault going on. That -is- the ground wire back to the SATA connector and if only that is melted, and the +5V is fine, then it's not over-pulling current on +5.

 

Please confirm or deny that's the situation, and if that is the case, let us know:

What did you have plugged into the ports of the hub? Describe the things and the connector plugged in for each. Take a picture of the thing and connector if you want to be extra-thorough. For example, what kind of fans? Which connector of the two on the fan? Did you plug strips into the hub? Etc.

 

The only way it's likely to be a problem with the PSU specifically is if the PSU is running +12V to that instead of ground, which would fry anything and everything connected to the SATA connector, and since you're not indicating that everything was fried. That sort of thing happens when the wrong cable is used for a modular PSU (for example, a cable for a different brand of PSU).

 

There are a few other things that could be wrong, for example, a loose connector inside the wiring. Though less likely to occur on two separate hubs, if they were both acquired from the same place, they could both be from the same batch. Heat travels well along the metal of the wire. However it doesn't travel perfectly and would cause more burning/melting closer to the fault.

 

The hubs could be bad but one hub came in an 570x and one in fan bundle so unlikely these are the same batch but could be wrong. Correct, only the black wires are melted on both hubs - not the black and grey.

 

All cables are for the ax850's and are all labeled as such so no incorrect wires were used.

 

Nothing was fried other than the hub sata cable which on the first setup had a HDD and SSD connnected.

 

Close up: http://i.imgur.com/nYKHj8f.jpg?1

http://i.imgur.com/I9UAr9w.jpg?1

Edited August 13, 2017 by Krowzer

Definitely the ground wire got far too hot, and the +5V is not damaged (That's the intact wire right next to it with the white markings)

 

Electricity doesn't do stuff unless it has somewhere to go. So the ground wire had to be carrying too much current. The current comes "from" a voltage source, goes through all the things, and then "to" ground/common (not getting into electron flow physics).

 

The original claim of a ground fault would be related to "+5V is getting to the hub and then escaping to a ground without doing enough, so the +5V wire is melted". However the ground wire is melted, not the +5V as we initially assumed.

 

If the PSU were pushing +12V on the ground, that would cause -7V to flow to +5V (It's all about differences) and to any ground fault (which doesn't exist based on your examination).

 

But the problem is that only the ground wire is melted. So either that wire itself (the physical object) is defective in such a way that the current caused normal use to melt it, which as you point out is unlikely for both of them to be faulty, or there's another source of current somewhere that is then reaching ground through the hub and causing the overcurrent. I'd expect the SATA connector from the PSU to have slightly more tolerance that it's not required to have, and unless the +5V and ground are very, very different at some location, they would both melt if overcurrent were going through both of them.

 

In a default setup (Hub connected to SATA, connected to fan LED connections, and connected to the three-button controller) the only source of power hitting the rails in the hub is from the SATA connector. Your OP description indicates a very quick smoke and melting, so it's not, for example, pulling just a little over-amps. With only two fans' LEDs connected, that would normally be an absolute maximum of 1.44A and the wires should be specced to handle at least 4.5A.

 

We really run into "If the power current that melted the ground came entirely in through the other wire that is not melted, it would be melted too." unless the two wires are Very Different and both hubs are faulty, which you have pointed out accurately is unlikely.

 

Therefore it comes down to "Where is the rest of the current coming from?", which is why I asked what's connected.

 

The alternative to something being connected that is pushing extra current (which could also be a design flaw depending on the setup) is that you missed a wire break that was, for example, causing a short from the 12V fan power to the LED wire somehow.

 

Another potential flaw (here, one in the PSU for you) is in the event that a powered external controller (LNP for example) is connected to the hub, which would be pushing 5V signal in the data line of the LEDs, which also runs to ground. Normally the signal is relatively low current though (a few mA) but in the event that the ground from the other connection is faulty (Not grounded) that would generate extra current to the hub ground line. (Quite a lot more than expected for just two fans' LED connections though). In any case, there's a very big difference if there is no other source of power hitting that ground than if there is.

 

So how exactly were what things connected to the hub?

Definitely the ground wire got far too hot, and the +5V is not damaged (That's the intact wire right next to it with the white markings)

 

Electricity doesn't do stuff unless it has somewhere to go. So the ground wire had to be carrying too much current. The current comes "from" a voltage source, goes through all the things, and then "to" ground/common (not getting into electron flow physics).

 

The original claim of a ground fault would be related to "+5V is getting to the hub and then escaping to a ground without doing enough, so the +5V wire is melted". However the ground wire is melted, not the +5V as we initially assumed.

 

If the PSU were pushing +12V on the ground, that would cause -7V to flow to +5V (It's all about differences) and to any ground fault (which doesn't exist based on your examination).

 

But the problem is that only the ground wire is melted. So either that wire itself (the physical object) is defective in such a way that the current caused normal use to melt it, which as you point out is unlikely for both of them to be faulty, or there's another source of current somewhere that is then reaching ground through the hub and causing the overcurrent. I'd expect the SATA connector from the PSU to have slightly more tolerance that it's not required to have, and unless the +5V and ground are very, very different at some location, they would both melt if overcurrent were going through both of them.

 

In a default setup (Hub connected to SATA, connected to fan LED connections, and connected to the three-button controller) the only source of power hitting the rails in the hub is from the SATA connector. Your OP description indicates a very quick smoke and melting, so it's not, for example, pulling just a little over-amps. With only two fans' LEDs connected, that would normally be an absolute maximum of 1.44A and the wires should be specced to handle at least 4.5A.

 

We really run into "If the power current that melted the ground came entirely in through the other wire that is not melted, it would be melted too." unless the two wires are Very Different and both hubs are faulty, which you have pointed out accurately is unlikely.

 

Therefore it comes down to "Where is the rest of the current coming from?", which is why I asked what's connected.

 

The alternative to something being connected that is pushing extra current (which could also be a design flaw depending on the setup) is that you missed a wire break that was, for example, causing a short from the 12V fan power to the LED wire somehow.

 

Another potential flaw (here, one in the PSU for you) is in the event that a powered external controller (LNP for example) is connected to the hub, which would be pushing 5V signal in the data line of the LEDs, which also runs to ground. Normally the signal is relatively low current though (a few mA) but in the event that the ground from the other connection is faulty (Not grounded) that would generate extra current to the hub ground line. (Quite a lot more than expected for just two fans' LED connections though). In any case, there's a very big difference if there is no other source of power hitting that ground than if there is.

 

So how exactly were what things connected to the hub?

 

The wire starts to overheat with a single SP120 RGB fan attached to it via the RGB lead. When the RGB lead isn't connected to the hub but the fan is given power via just the 3pin it doesn't over heat. So basically whenever the RGB hub draws power it goes into melt down.

If you had absolutely nothing but:

- SATA to hub

- SP120 RGB LED connection to hub (Max .25A for the LEDs and controller chip combined)

(in which configuration the fan would be making no lights whatsoever and would normally have very minimal current draw to power the controller chip in the fan)

 

 

 

and it overheated the wire, then you have a shorted fan or a shorted hub or two defective hubs, or if your PSU is defective, then it's incorrect voltage by a substantial amount and the fan LEDs are fried and shorting.

 

Anybody who is trying to help you relies on your eyes for information and we can only have findings based on what we know and are told.

 

So do you have a shorted fan, a shorted or defective hub, or something else connected?

Old Thread I Know

 

I'm sure that is from using a sata power cable from a different brand/model PSU

they're not all wired the same .. so it melts the cable & can fry what it's feeding to

 

I had this happen with Cablemod UV/RGB LEDs .. not the wires though .. something melted the silicon on the UV LEDs