Cryptocoin Mining?

[IEC]

Bizarre problem today. So I move my miner from underneath my workbench to shelf next to it. The power is routed the same 'wall socket' --> 'ext. cord' --> 'kill-a-watt' --> 'power cord' --> 'TR-700 PSU'

I get it mining again and after 20 or so minutes I notice a burning plastic smell. We find the source, the ENTIRE POWER CORD going into the PSU is basically melting/smoking. Just the "sheath/cord", not the adapters @ either end. It's now a nice smooth shiny black where it didn't completely melt off!

So far, I've swapped power cords, PSUs, wall sockets (confirmed different circuits in the house), removed the ext. cord & kill-a-watt and, no matter what, the power cord always slowly heats up. After about a minute it's noticeably hotter.

So... it's my motherboard or the GPUs causing a short only in the power cord?!?!?! I guess I'll try each GPU individually and see what happens.... what the f%$k.

How much wattage are you running through that cord? Your regular PC power supply cord can't take much... you should use thicker cords for 1000W PSUs for example - and a completely different type of cord for 1300W+.

 

[nwo]

I noticed that even my TX 750 power cord is much thicker than the 600W and sub 600W power cords.

 

[Attic]

I've got a 280x that intermittently drops it's fan to under 1100rpm's and stays there. Fan monitoring software indicates that the fan changes state from ~1500rpm and shows operation at 40% fan speed but then something occurs that I can't pin down (random behavior it appears) and it goes to <1100rpms and shows 100% fan speed.

The card is on a powered 1x to 16x riser from a PCI-Ex 1x slot. The behavior has followed the card between two different systems. The behavior has also been observed, though still randomly, when plugged directly into a x16slot as the only card. I can get the card sub 70c with it's fan topping out at 1100rpms but I have to drop it to 680kh/s from 720 or so otherwise the fan won't spin up beyond 1100rpms with temperatures surpassing set temp limits. Been putting the thing on the back burner for awhile and every time I think it's sorted it nonetheless eventually reappears.

Any ideas? Possible autofan/cgminer/volt/bios related? It's an Asus 280x DCII Top. Another identical card is not showing same behavior in very similar/identical settings on the same system.

Edit: Well putting it out got me thinking so I tried --gpu-fan 25-80,25-80,25-80 in my configuration as opposed to --gpu-fan 25-80 and that has for now gotten the fan operating as intended. FWIW there are 3 gpus running on the config. Oddly I have confirmed the issue appears when running the card solo with --gpu-fan 25-80 in the past.

 

[T_Yamamoto]

I've got a 280x that intermittently drops it's fan to under 1100rpm's and stays there. Fan monitoring software indicates that the fan changes state from ~1500rpm and shows operation at 40% fan speed but then something occurs that I can't pin down (random behavior it appears) and it goes to <1100rpms and shows 100% fan speed.

The card is on a powered 1x to 16x riser from a PCI-Ex 1x slot. The behavior has followed the card between two different systems. The behavior has also been observed, though still randomly, when plugged directly into a x16slot as the only card. I can get the card sub 70c with it's fan topping out at 1100rpms but I have to drop it to 680kh/s from 720 or so otherwise the fan won't spin up beyond 1100rpms with temperatures surpassing set temp limits. Been putting the thing on the back burner for awhile and every time I think it's sorted it nonetheless eventually reappears.

Any ideas? Possible autofan/cgminer/volt/bios related? It's an Asus 280x DCII Top. Another identical card is not showing same behavior in very similar/identical settings on the same system.

Edit: Well putting it out got me thinking so I tried --gpu-fan 25-80,25-80,25-80 in my configuration as opposed to --gpu-fan 25-80 and that has for now gotten the fan operating as intended. FWIW there are 3 gpus running on the config. Oddly I have confirmed the issue appears when running the card solo with --gpu-fan 25-80 in the past.

Try using a different program.

Like try MSI Afterburner to play with the fan.

 

[Attic]

Try using a different program.

Like try MSI Afterburner to play with the fan.

I had tried this and using MSI afterburner during periods where the fan was behaving oddly would result in the fan spinning temporally up and then immediately back down in some cases and in other cases no change at all when putting fan to 100% and unchecking "auto". Still touch and go.

 

[Torn Mind]

less competition, lower difficulty, more coins for other pools/solo miners I guess. But it's too bad people can't do something more productive with their time I'm sure all it causes people to do is jump to other pools and really has little to no effect on the difficulty in the long run.

It can screw over a pool since miners who leave don't come back. It is technically productive for them, but at the expense of others.

 

[Torn Mind]

Bizarre problem today. So I move my miner from underneath my workbench to shelf next to it. The power is routed the same 'wall socket' --> 'ext. cord' --> 'kill-a-watt' --> 'power cord' --> 'TR-700 PSU'

I get it mining again and after 20 or so minutes I notice a burning plastic smell. We find the source, the ENTIRE POWER CORD going into the PSU is basically melting/smoking. Just the "sheath/cord", not the adapters @ either end. It's now a nice smooth shiny black where it didn't completely melt off!

So far, I've swapped power cords, PSUs, wall sockets (confirmed different circuits in the house), removed the ext. cord & kill-a-watt and, no matter what, the power cord always slowly heats up. After about a minute it's noticeably hotter.

So... it's my motherboard or the GPUs causing a short only in the power cord?!?!?! I guess I'll try each GPU individually and see what happens.... what the f%$k.

Simple matter is too big a wire gauge and/or too long a cable. Both factor contribute to the resistance of the cable. Bigger gauge means thinner wire. Too much resistance in the wire leads to too much heat in too small an area. You want 14 gauge at the least if you are pullung 1000+ watts and don't use too long a cable. Lower the gauge if the cord is longer, although 12 gauge cord aren't easy to find.

 

[Virge_]

Lower the gauge if the cord is longer, although 12 gauge cord aren't easy to find.

Amazon begs to differ!

 

[KlokWyze]

How much wattage are you running through that cord? Your regular PC power supply cord can't take much... you should use thicker cords for 1000W PSUs for example - and a completely different type of cord for 1300W+.

I noticed that even my TX 750 power cord is much thicker than the 600W and sub 600W power cords.

<700 watts continuously. Not a TON, but substantial.

So after more swapping components, etc. I've made a little headway and I THINK it's the motherboard.

After using both of the 7950's in the PCI Express slot 0 individually I still get the heating up of the power cable for both card (not as fast, not as much wattage). Here is the interesting part. I swap the power cable (from kill-a-watt --> PSU) with a much thicker one. It heats very little over a much longer period of time, BUT, I noticed a burning rubber/plastic smell again, but couldn't pin point it.

I leave it going for ~15 minutes and reinvestigate, I start touching and feeling around all the wires and such and the power cord is maybe very slightly warm, but one small segment of PSU --> GPU power connects are REALLY f%^king hot. I shot it off immediately. Keep in mind on the other PSU I'm using only built-in PSU --> GPU power connects, so there are no same components.

WTF is going on here? Could there be some voltage regulator or connection on the motherboard that is bad? My best idea is to just swap the mobo out and see if ANY of the wires heat up. Heading out to Microcenter 1st thing in the morning to pick one up.

Any electrical engineers here?!?!!?

 

[KlokWyze]

Simple matter is too big a wire gauge and/or too long a cable. Both factor contribute to the resistance of the cable. Bigger gauge means thinner wire. Too much resistance in the wire leads to too much heat in too small an area. You want 14 gauge at the least if you are pullung 1000+ watts and don't use too long a cable. Lower the gauge if the cord is longer, although 12 gauge cord aren't easy to find.

Then why would my PSU --> GPU wires be overheating? I was using all the same setup when this happened. Nothing changed. I simply turned it off after several weeks of mining, moved it and this happened like 20 or 30 minutes after it came back online.

 

[Torn Mind]

Then why would my PSU --> GPU wires be overheating? I was using all the same setup when this happened. Nothing changed. I simply turned it off after several weeks of mining, moved it and this happened like 20 or 30 minutes after it came back online.

Odd indeed, then. 700 watts is quite small, relatively speaking. It seems there is a fault somewhere in the pathway on the PCI-e connector and it is causing a huge current pull through the PSU. Is the 12 volt rail reading normally otherwise?

 

[IEC]

What PSU is it? I would think if it's drawing enough current to melt the cord the OCP should kick in... unless it's an el cheapo without proper OCP...

 

[Torn Mind]

What PSU is it? I would think if it's drawing enough current to melt the cord the OCP should kick in... unless it's an el cheapo without proper OCP...

I think it is some TR-700, according to his post with the pic. But Thermaltake has a lot of these TR units, and they all are not the same. So, the spec label and exact model number would help clarify a lot.

@KlokWyze
If your unit is out of warranty, you might as well just open up the unit and visually inspect ir.

 

[KlokWyze]

How do I measure the voltages on my PSU +12V rail? Does it have a built-in sensor that it shares with the motherboard, OS, etc.? Or are you saying to use a multimeter on it?

I've already swapped out the PSU with another and the SAME problem of the overheating cord is occuring. I also used verified different circuits in my house.

To humor you guys the PSUs are Thermaltake TR2-700W & a CoolMax VL-600B. Both PSUs are fairly new, both in warranty and I am very confident(99%) they are both good. I am 100% confident that both are not bad causing the exact same, very weird issue.

 

[AnonymouseUser]

How do I measure the voltages on my PSU +12V rail? Does it have a built-in sensor that it shares with the motherboard, OS, etc.? Or are you saying to use a multimeter on it?

Both sensors and multimeter, but multimeter is more reliable.

 

[Torn Mind]

How do I measure the voltages on my PSU +12V rail? Does it have a built-in sensor that it shares with the motherboard, OS, etc.? Or are you saying to use a multimeter on it?

I've already swapped out the PSU with another and the SAME problem of the overheating cord is occuring. I also used verified different circuits in my house.

To humor you guys the PSUs are Thermaltake TR2-700W & a CoolMax VL-600B. Both PSUs are fairly new, both in warranty and I am very confident(99%) they are both good. I am 100% confident that both are not bad causing the exact same, very weird issue.

Just the software reading. But since the problem persisted even with a different unit,, chances are the card is what is wrong. Do you still see a display when the monitor is connected to the card output? If you do, then this problem is very, very odd.

Coolmax units give me shivers since their past units failed testing at review sites, and quite horribly so. Since TT uses the same name, there are many TR2s out there.

 

[KlokWyze]

Both sensors and multimeter, but multimeter is more reliable.

Awesome will. Check that out in a moment. Need to RDP into the host, install, etc.

Just the software reading. But since the problem persisted even with a different unit,, chances are the card is what is wrong. Do you still see a display when the monitor is connected to the card output? If you do, then this problem is very, very odd.

Coolmax units give me shivers since their past units failed testing at review sites, and quite horribly so. Since TT uses the same name, there are many TR2s out there.

hehe... This is a dual GPU miner (2x Gigabyte WF3 7950's). I've already mined with each individual card in PCI-X slot 0 and the same problem persists.

The unit is ON right now, just not mining. When I start pushing more watts through it, by mining at least, the thinner power cords heat up rather quickly, 1 - 2 min. The thicker gauge cords don't really noticably heat much at all, but that's when I noticed a patch of cable between the GPU and the PSU were scorching hot.... think I'll try mining again and see if the PSU --> GPU cords heat up... like maybe the resistance on the power cord matters or something??????? Just throwing spaghetti at the wall here.

 

[T_Yamamoto]

Coolmax is trash but I've seen plenty of miners use them because its high power and low cost.

Gives me a shiver every time.

 

[KlokWyze]

So I guess this is normal? It doesn't fluctuate much at all when I start mining. This is a capture from it after I just mined with it for about 3 minutes. This time, only a small patch of cords in between PSU --> GPU molex connectors heat up pretty good.

Also, eletrical wires supplying power should NEVER heat up from the current right? That would mean they were failing at that level of current?

 

[KlokWyze]

Coolmax is trash but I've seen plenty of miners use them because its high power and low cost.

Gives me a shiver every time.

Thermaltake is good. I in fact have a TW2-430 in my main rig that is roughly 8 years old. That system is always on. I used to mine with it. Same PSU.

I got the Coolmax from a combo I bought. I actually ended up replacing it on that system and just had it laying around. Regardless, both PSUs are fine.... The power cord actually melted on the Thermaltake, but I am currently using the Coolmax.

 

[KlokWyze]

Unless anyone has anything to add. I'm just going to go to the store, buy a comparable motherboard just to test. If it works I'll leave it and RMA the Gigabyte GA-F2A88XM-D3H.

Hmmm... think I'll search for common problems, power issues, etc. on that mobo right about now.

 

[UNhooked]

So I guess this is normal? It doesn't fluctuate much at all when I start mining. This is a capture from it after I just mined with it for about 3 minutes. This time, only a small patch of cords in between PSU --> GPU molex connectors heat up pretty good.

Also, eletrical wires supplying power should NEVER heat up from the current right? That would mean they were failing at that level of current?

That PSU is failing. Just look at the voltages!!! 12V line is at 8V

That or that mobo is failing.

 

[Torn Mind]

So I guess this is normal? It doesn't fluctuate much at all when I start mining. This is a capture from it after I just mined with it for about 3 minutes. This time, only a small patch of cords in between PSU --> GPU molex connectors heat up pretty good.

Also, eletrical wires supplying power should NEVER heat up from the current right? That would mean they were failing at that level of current?

Something is really wrong there. Massive under-voltage on all of the rails. Way out of spec, yet your computer is still working.

 

[Torn Mind]

That PSU is failing. Just look at the voltages!!! 12V line is at 8V

That or that mobo is failing.

Since the other symptoms he experienced did not change with a PSU swap, I think it is the mobo or the card itself having a massive short circuit somewhere.

A rudimentary test of the PSU's being functional or not is to test the PSU while it is disconnected from the motherboard. One needs to short the grey "Power Good" wire to ground(black color) and then use a multimeter, with the black lead into any ground wire connector(black wire) and the red into the respective output wire(yellow, orange, red, etc)

 

[Torn Mind]

Thermaltake is good. I in fact have a TW2-430 in my main rig that is roughly 8 years old. That system is always on. I used to mine with it. Same PSU.

I got the Coolmax from a combo I bought. I actually ended up replacing it on that system and just had it laying around. Regardless, both PSUs are fine.... The power cord actually melted on the Thermaltake, but I am currently using the Coolmax.

Powering on a computer is something I would expect even the most crappy of PSUs to do. Computers can have high tolerance with regards to crappy power output so that the symptoms are not noticeable while it is in use, since one would usually blame on "instability", which in turn means blaming the OS. Otherwise, the flawed units don't show their colors until they die or suffer through transients they can't handle(brownout, etc).

I ran a P4 system the family received as a hand-me-down charity gift with a crap PSU(some case unit from Okia). Powered on and off with no issue, and the only noticeable bit of instability was when I closed Firefox once and the whole computer restarted.

TT makes okay functional units, though not the best. Coolmax, though, even if they cherry picked their units(don't know if they did or not), had their units die during review tests, tests the company should have known would kill their supplies if the PSUs were not up to snuff.