The "Everyday" show-off thread !

[enc0dex264]

Alright.. So i came home to a room full of burning smell a few days ago.. it seems my PSU i.e Antec HCG1000 paired with a RTX 4090 decided to burn itself. Funnily enough the system was still on when i came back home somehow.

I was using a Cablemod 12HPWR cable male to female 3 x 8PIN Female extenstion cable to run my 4090.

1) The extension cable had melted wires/burnt.
2) The point where PCIe cables go into the extension cable also burnt.
3) The PCIe Sockets on the PSU itself where the pcie cables get pwoer from also burnt.

Some pictures below:

So i decided to go buy a new PSU.. and got faced with another issue.. was installing MSI A1000G and saw that the ATX 3.0 pcie cable was way too short for the cabinet.. since in this cabinet the psu sits above the mobo and if i were to run the cable from below the gpu it would not make it.. and i didnt want to ruin the look of the cabinet by running the cable from above. It would ruin the whole point of having a glass cabinet with RGB fans etc.. although it would have saved me some money for sure.

Almost all PSU's have 600 to 650mm of pcie cables since they assume u would be running a traditional case where the PSU is below the mobo.

Did alot of googling and saw Corsair RM1000x cables were long enough to make it from below the gpu (750mm) and decided to go ahead with it.. and bought a vertical bracket + the corsair 12V HPWR cable to run with it.

Hopefully it doesnt burn itself again.. :'(

Trying to RMA the Antec Supply via a dealer now.. Will be selling the replacement to re-coup the new PSU's cost.

Here's what it looks like now.

I'm also very curious why it burnt? was it the PSU's fault or the extension cables fault?

The dealer suggested you need a 12V single rail supply for a 4090 since it draws to much current.. almost all older supplies with a few exceptions are quad rails and that was the reason of this incident.. just an FYI for other 4090 or similar card owners.. or it could totally have been a losse connection or the extensions fault.

 

[Alucard1729]

The dealer suggested you need a 12V single rail supply for a 4090 since it draws to much current

Antec HCG1000 is a single rail psu. This video was released recently for 12vhpwr cables.
Seems more like a cable issue, but might as well be a faulty PSU combined with the cable causing this.

There's also a wiki page documenting the reliability...

 

[vaibhavyagnik]

Alright.. So i came home to a room full of burning smell a few days ago.. it seems my PSU i.e Antec HCG1000 paired with a RTX 4090 decided to burn itself. Funnily enough the system was still on when i came back home somehow.

I was using a Cablemod 12HPWR cable male to female 3 x 8PIN Female extenstion cable to run my 4090.

1) The extension cable had melted wires/burnt.
2) The point where PCIe cables go into the extension cable also burnt.
3) The PCIe Sockets on the PSU itself where the pcie cables get pwoer from also burnt.

Some pictures below:


View attachment 189232
View attachment 189233
View attachment 189234

So i decided to go buy a new PSU.. and got faced with another issue.. was installing MSI A1000G and saw that the ATX 3.0 pcie cable was way too short for the cabinet.. since in this cabinet the psu sits above the mobo and if i were to run the cable from below the gpu it would not make it.. and i didnt want to ruin the look of the cabinet by running the cable from above. It would ruin the whole point of having a glass cabinet with RGB fans etc.. although it would have saved me some money for sure.

Almost all PSU's have 600 to 650mm of pcie cables since they assume u would be running a traditional case where the PSU is below the mobo.

Did alot of googling and saw Corsair RM1000x cables were long enough to make it from below the gpu (750mm) and decided to go ahead with it.. and bought a vertical bracket + the corsair 12V HPWR cable to run with it.

Hopefully it doesnt burn itself again.. :'(

Trying to RMA the Antec Supply via a dealer now.. Will be selling the replacement to re-coup the new PSU's cost.

Here's what it looks like now.

View attachment 189235

I'm also very curious why it burnt? was it the PSU's fault or the extension cables fault?

The dealer suggested you need a 12V single rail supply for a 4090 since it draws to much current.. almost all older supplies with a few exceptions are quad rails and that was the reason of this incident.. just an FYI for other 4090 or similar card owners.. or it could totally have been a losse connection or the extensions fault.

Heat total = (current)2*Resistance*time.
If only the connection between the PCIE cable and the extension cable had burnt, it can be contributed due to a loose connection, causing increased resistance, causing heat generation at the joint, causing melting. But, all the connectors in the circuit have burnt which leads me to think that the 4090 was drawing current which was more that what the 12V rail was designed for and for an extended period of time. Why did the overcurrent protection not kick in? - that is another question. Anyway it might be worth to buy one of those non contact type temperature measurement thermometers and keep measuring the temperature at all the junctions on the 12V rail. This might warn you early if there is a problem. I assume the wires will be made from XLPE or crossed linked poly ethylene which is rated at 90 Deg C. So if you see anything more that 90 degc, you should switch off your computer. Might also be worth it to have a habit of shutting down the computer when you are not around, unless it is running something.
When you returned home, was the computer underload or was it just idling?

 

[enc0dex264]

Heat total = (current)2*Resistance*time.
If only the connection between the PCIE cable and the extension cable had burnt, it can be contributed due to a loose connection, causing increased resistance, causing heat generation at the joint, causing melting. But, all the connectors in the circuit have burnt which leads me to think that the 4090 was drawing current which was more that what the 12V rail was designed for and for an extended period of time. Why did the overcurrent protection not kick in? - that is another question. Anyway it might be worth to buy one of those non contact type temperature measurement thermometers and keep measuring the temperature at all the junctions on the 12V rail. This might warn you early if there is a problem. I assume the wires will be made from XLPE or crossed linked poly ethylene which is rated at 90 Deg C. So if you see anything more that 90 degc, you should switch off your computer. Might also be worth it to have a habit of shutting down the computer when you are not around, unless it is running something.
When you returned home, was the computer underload or was it just idling?

When i had left the gpu was at a 100% mining some shitcoin .. but ive been mining for the past month or so..

So it seems it very well could have been a loose connection aswell cause just 2-3 days ago i had done some cleaning but if this was the case it would have only burnt at the gpu connector point.. but everything except for that point burnt.

 

[PunkX 75]

Alright.. So i came home to a room full of burning smell a few days ago.. it seems my PSU i.e Antec HCG1000 paired with a RTX 4090 decided to burn itself. Funnily enough the system was still on when i came back home somehow.

I was using a Cablemod 12HPWR cable male to female 3 x 8PIN Female extenstion cable to run my 4090.

1) The extension cable had melted wires/burnt.
2) The point where PCIe cables go into the extension cable also burnt.
3) The PCIe Sockets on the PSU itself where the pcie cables get pwoer from also burnt.

Some pictures below:


View attachment 189232
View attachment 189233
View attachment 189234

So i decided to go buy a new PSU.. and got faced with another issue.. was installing MSI A1000G and saw that the ATX 3.0 pcie cable was way too short for the cabinet.. since in this cabinet the psu sits above the mobo and if i were to run the cable from below the gpu it would not make it.. and i didnt want to ruin the look of the cabinet by running the cable from above. It would ruin the whole point of having a glass cabinet with RGB fans etc.. although it would have saved me some money for sure.

Almost all PSU's have 600 to 650mm of pcie cables since they assume u would be running a traditional case where the PSU is below the mobo.

Did alot of googling and saw Corsair RM1000x cables were long enough to make it from below the gpu (750mm) and decided to go ahead with it.. and bought a vertical bracket + the corsair 12V HPWR cable to run with it.

Hopefully it doesnt burn itself again.. :'(

Trying to RMA the Antec Supply via a dealer now.. Will be selling the replacement to re-coup the new PSU's cost.

Here's what it looks like now.

View attachment 189235

I'm also very curious why it burnt? was it the PSU's fault or the extension cables fault?

The dealer suggested you need a 12V single rail supply for a 4090 since it draws to much current.. almost all older supplies with a few exceptions are quad rails and that was the reason of this incident.. just an FYI for other 4090 or similar card owners.. or it could totally have been a losse connection or the extensions fault.

This may have also been because of Cablemod in general:

Cablemod issues Massive Warning not to use their 12VPWR...

www.overclock.net

 

[enc0dex264]

This may have also been because of Cablemod in general:

Cablemod issues Massive Warning not to use their 12VPWR...

www.overclock.net

AFAIK.. there angled connectors are faulty as its very easy to cause a loose connection because of a design flaw.. but the extension cables have had no reports of burning, the extension cables are pretty much the same as the connector that come inbox with the gpu itself.. just longer.

but yeah it very well could have been.. but then why did it take a full month of 100% load to burn it.. the timeframe leads me to believe it was probably a faulty PSU.

 

[vaibhavyagnik]

AFAIK.. there angled connectors are faulty as its very easy to cause a loose connection because of a design flaw.. but the extension cables have had no reports of burning, the extension cables are pretty much the same as the connector that come inbox with the gpu itself.. just longer.

but yeah it very well could have been.. but then why did it take a full month of 100% load to burn it.. the timeframe leads me to believe it was probably a faulty PSU.

I think the cable and connectors were running hot for quiet sometime and finally the heat buildup was enough to burn the insulation and connectors. My opinion is more towards long term overload rather than a faulty PSU.

 

[enc0dex264]

I think the cable and connectors were running hot for quiet sometime and finally the heat buildup was enough to burn the insulation and connectors. My opinion is more towards long term overload rather than a faulty PSU.

Well in that case.. Im going to run the corsair rm1000x in a similar manner but without any aftermarket cables.

its cold af and mining is helping me somewhat manage the cold.. guess we will find out in a few months or so.

 

[Tracer_Bullet]

When i had left the gpu was at a 100% mining some shitcoin .. but ive been mining for the past month or so..

So it seems it very well could have been a loose connection aswell cause just 2-3 days ago i had done some cleaning but if this was the case it would have only burnt at the gpu connector point.. but everything except for that point burnt.

Probably should update 4090 sale thread about mining ? anyway gl..

 

[enc0dex264]

Probably should update 4090 sale thread about mining ? anyway gl..

Not gonna sell anymore anyway. Will request mods to close it.

 

[nRiTeCh]

My Fast and Furious diorama..

Diorama by my friend..
LED by me: Glows in 9 colors adjustable manually. Disco mode, multi-rotate mode, and stable white, red, blue, green, yellow, pink & off white.

Will post with car scene in a few days..

 

[Black_Hawk]

My Fast and Furious diorama..

View attachment 189333

View attachment 189334

Diorama by my friend..
LED by me and: Glows in 9 colors adjustable manually. Disco mode, multi-rotate mode, and stable white, red, blue, green, yellow, pink & off white.

Will post with car scene in a few days..

This has to be one of coolest things I have seen being posted in this thread! The workmanship behind this and your passion is really amazing! Do post photos with the Hot Wheels models on display... will look lit!

 

[mav2000]

Well in that case.. Im going to run the corsair rm1000x in a similar manner but without any aftermarket cables.

its cold af and mining is helping me somewhat manage the cold.. guess we will find out in a few months or so.

But whats left to mine...?

 

[enc0dex264]

Alpheum is pretty good.. very speculative tho.. i mined some on my 4080 before i got the 4090 for a month or so.. and it 4xed in price (when i sold) and is currently priced at 2$ish.. i'm still mining at about 2$ per day right now and switching it to some alts i like more every week or so.

I think its only feasible to mine ALPH on 4000 series cards right now and that too the profitability is so low it doesnt make sense.. but if you are sepculative like i did when it was only 30 cents.. then u do make some good gains.

Even if you had mined KASPA after the ethereum merge at break even you would have stilll made decent money.. the real bull market is still yet to come.

I agree 15$ a week does nothing much really.. but i get free heating and also a chance for that 15$ to be 50-100 in a year or so when the real bull market begins.

edit: crypto is also a very deep pit.. for eg: i converted ALPH to a token called SHDW in January on a DEX called Jupiter (Solana based dex) and as a result qualified for a free WEN meme coin airdrop which i recieved yesterday worth 90$.. thats just free money for basically doing nothing.. the more you know

 

[rockyo27]

Yeah, but 40 series cards being pricey right now no one will buy them off for mining. As someone can get 30 series used cards at decent price so as to get good ROI. So it is all about ROI. If with used cards ROI is good then no point buying 40 series card untill it gives huge gain at decent price.

 

[enc0dex264]

Yeah, but 40 series cards being pricey right now no one will buy them off for mining. As someone can get 30 series used cards at decent price so as to get good ROI. So it is all about ROI. If with used cards ROI is good then no point buying 40 series card untill it gives huge gain at decent price.

I'm not recommending anyone buy 4000 series card for mining.

I just happened to have one for gaming which is sitting idle and i use it as a heater and use my knowledge in crypto to leverage it and make some money in the process.

 

[rockyo27]

I'm not recommending anyone buy 4000 series card for mining.

I just happened to have one for gaming which is sitting idle and i use it as a heater and use my knowledge in crypto to leverage it and make some money in the process.

Yeah. That can be done untill you are not paying the electricity bill from your pocket.

 

[nRiTeCh]

This has to be one of coolest things I have seen being posted in this thread! The workmanship behind this and your passion is really amazing! Do post photos with the Hot Wheels models on display... will look lit!

Sure man..will upload this weekend.. Stay tuned!

 

[anu0512]

Thinkbook 16+ and install the skin for protection. Plug two 980 Pro NVME

 

[blr_p]

Pardon my ignorance, but what is this used for?

Forgot to mention this.

Reminds me of the movie, Predator.

Works with centipedes & millipedes too.

Neat

@ze_cook