Jump to content
Corsair Community

Possible HX1200i problems?


Recommended Posts

I have a decent system that was working fine for the past 4 months but went dead today and will not post. I am using a HX1200i (with iCUE and the USB cable, set to single rail usage) and do not have another PSU here at the house with the two auxiliary CPU EPS connectors (for the far top left of the motherboard). I have completely unplugged the HX1200i and jumpered the main 24 pin ATX connector by connecting pins 15 and 16, and then adding one 140mm fan to give it a small load. When I turn on the main power switch, the fan inside the HX1200i and the attached 140mm fan both spin for about one sec and then shut down. (direct current seems to be flowing and then stops?) When I try the fan self test on the back, pressing the button does make the fan inside the PSU turn as long as I hold the button down. When I connect the PSU back to the motherboard, I have the LEDS on my Corsair memory lighting up and staying lit (which seems to indicate that DC current is indeed flowing thru the motherboard to the memory - stepped down from the 12 V rail by the MB if I remember correctly), a CPU error LED coming on (for about 100 milliseconds) and no POST activity (the system does not load into the BIOS). I am working to ascertain whether it is CPU, Motherboard or PSU. I would think that jumping pin 15 and 16 with a 140mm fan would either allow the fan to spin (the PSU is working) or not (the PSU is dead), but I have this indeterminate gray result. (the caps still have some charge left?, ... the rectifier is partially damaged?) Do I have a dead PSU? Edited by jayhall0315
Link to comment
Share on other sites

Update - Dragged the multimeter and oscilloscope out of the garage and tested at the wall socket and the PSU power cable itself, ... both 120.8 V rms at 60 Hz fine. Then tested the 24 pin ATX connector, .... dead. No 12 V, ...no 5 V, ... and no 3.3 V. No 12 V rail and yet the memory is lit? Black magic? ... someone poured goat's blood on my rig and did a voodoo curse? .... completely mystified at this point. 5 VSB is fine but that is a separate circuit iir. Does anyone know if Corsair warranty allows me to remove the cover and do some checking? (I think the answer is no, but if anyone knows I would appreciate it.) Edited by jayhall0315
Link to comment
Share on other sites

I spent a fair amount of money (iirc ~$350) to buy this HX1200i and now, I will be down some weeks in a close to mission critical system as I fart with the RMA process. I can understand that possibly with a Wing Tech $35 PSU but a $350 Corsair unit, .... you have got to be #$%$#$! kidding me!

 

To say that I am angry considering the money I spent, ... would be a gross understatement.

 

I am not trying to be a dick but I do not have ~ 3 weeks to fool with the typical RMA process. There is a 10 or 12 year warranty on these units and if that is true, then 99.9% of them should last 10 years without failure, .... no questions asked.

Edited by jayhall0315
Link to comment
Share on other sites

Jonny Guru or other Corsair techs, what PS_ON# controller IC is used in the HX1200i? Also, what is up with the attached 140 mm fan spinning for only about 1.5 sec while the LEDs on the memory stay lit indefinitely? (You can give an exact electrical engineering answer using calculus, diff EQ, etc, ... if needed.) Edited by jayhall0315
Link to comment
Share on other sites

  • Corsair Employees
Does anyone know if Corsair warranty allows me to remove the cover and do some checking? (I think the answer is no, but if anyone knows I would appreciate it.)

 

Opening the PSU voids the warranty. Besides, it's very unlikely the failure is user serviceable.

 

You would need to go through the warranty process. And to save time, ask to do an advance replacement (where they ship you the PSU first).

 

As for using another PSU: The two CPU power connectors is not necessary 99% of the time. What CPU do you have? The only time you actually NEED two CPU power connectors is if you're using a 180W TDP or higher CPU like a Threadripper or Core i9 Extreme Edition.

Link to comment
Share on other sites

Thank you Jonny for mentioning the advance replacement. I am using the i9900K which is about ~ 150 W at 5.2 GHz, so perhaps not. Been several years since I was an undergrad and took Circuits, Power Systems, etc .... but shouldn't be too hard to get back up to speed for repairing a PSU, but I will RMA it. Any idea why the memory LEDs can stay lit while attached fans won't spin?
Link to comment
Share on other sites

Jonny, I ordered the advance replacement, and the good news is that it arrived six days later. I do appreciate that Corsair has this service, and I do appreciate that they charge the original MSRP for the product. Six days is certainly better than three or four weeks.

 

That said, the last week has given me time to read up on reviews of Corsair power supplies, and I have to say, I think the real failure rate of the HXi and AXi lines is definitely more than 1%, which should be unacceptable in 2020. Based on the rate of failures mentioned at Newegg, Amazon, etc .... it seems like the real failure rate is probably about ~ 3 to 4% when purchased new and rising to ~ 8% or more by three years out. This is a bit mystifying to me since the electrical engineering specs for the ATX 2.31 standard have been out for about 12 years now. I am considering forgoing the HX1200i return (and losing about $309) because I am genuinely curious as to a total breakdown of this damaged unit to see what is up. Since my unit worked fine for 2.3 years before failure, I can rule out damage from shipping or poor installation, and am left with real curiosity.

 

I guess Corsair would take none too kindly if we sent them back a fully dissected unit with possibly an exact answer for why the unit failed?

 

On another note, I've got to say, the whole iCUE monitoring stuff seems, ... not needed. It may be a great selling point for Corsair rep to advertise at electronics shows but in real life? The whole point of a good PSU is that you install it and forget it. We've got one AX1600i in the lab and while it can be adjusted through iCUE, most of it is stuff no end user (including extreme overclockers) would ever actually need to fool with. I remain unconvinced there is any real need for it.

 

Let me see how much free time I get this weekend, ....

Edited by jayhall0315
Link to comment
Share on other sites

Also Jonny, this is not specific to Corsair, but if an end user correctly mounts the PSU in the bottom of their case, with the fan facing upwards, then the internal fan of the PSU is pulling in hot air from the nearby graphics card to cool the PSU? Many cases made in the last ten years have grills or vents in the bottom which means the PSU should be mounted inverted (upside down) with the internal fan of the PSU pulling in cool air from under the case (unless the case sits directly on the floor with no offset or feet). Of course the critic here would say the caps are rated at 105 deg C but still, a 3 or 4 deg delta for free is worthy of consideration.
Link to comment
Share on other sites

it's not unusual to mount PSUs in the down position with fan facing upwards. People with pets or who place their computer on carpet (may they be flogged for this heresy) can do that to avoid clogging the internals with dust and dog hair.. or just to have air at all, in case of carpet.

 

For caps, in efficient PSUs like this one, temp is of no concern really.

rule of thumb is lifespan doubles for every 10° saved from the max rating.

 

take your average Nippon chemicon rated at 2000h at 105°

4000 at 95, 8000 at 85 etc...

 

if you make it run at 30 instead of 35, you're still well within the warranty period. They usually don't heat up anyway (lets be serious, we buy oversized)

 

As for failure rate, since i joined this forum i have seen lots of problems about low end PSUs and AXi. As low as Corsair has fallen in my personal view, HX/HXi still are workhorses afaik. but well, i speak from an experience of two PSUs. an old 850 that i sold and still works perfectly for my friend and a 1200i that has been flawless. Flukes happen ^^'

Link to comment
Share on other sites

  • Corsair Employees

That said, the last week has given me time to read up on reviews of Corsair power supplies, and I have to say, I think the real failure rate of the HXi and AXi lines is definitely more than 1%, which should be unacceptable in 2020. Based on the rate of failures mentioned at Newegg, Amazon, etc .... it seems like the real failure rate is probably about ~ 3 to 4% when purchased new and rising to ~ 8% or more by three years out. This is a bit mystifying to me since the electrical engineering specs for the ATX 2.31 standard have been out for about 12 years now. I am considering forgoing the HX1200i return (and losing about $309) because I am genuinely curious as to a total breakdown of this damaged unit to see what is up. Since my unit worked fine for 2.3 years before failure, I can rule out damage from shipping or poor installation, and am left with real curiosity.

 

Absolutely not. And I'll tell you why....

 

We sell probably about 1000+ of each model every week. Do you see 1000 new reviews every week? No. That's because less than 1% of users leave feedback if their parts work as expected. I'm guilty of it too. But if a part fails, the user is 99% guaranteed to leave negative feedback.

 

I really WISH 100%... or even 50%... of users would leave feedback good OR bad. But that's simply not the case.

 

On another note, I've got to say, the whole iCUE monitoring stuff seems, ... not needed. It may be a great selling point for Corsair rep to advertise at electronics shows but in real life? The whole point of a good PSU is that you install it and forget. We've got one AX1600i in the lab and while it can be adjusted through iCUE, most of it is stuff no end user (including extreme overclockers) would ever actually need to fool with. I remain unconvinced there is any real need for it.

 

Let me see how much free time I get this weekend, ....

 

Well... that's you.

 

For gamers, overclockers, etc. they like to see the power consumption. I find the logging the most useful aspect as you can start logging and play games or benchmark and then come back to the spreadsheet and see the spikes in power consumption, any drops in voltage, etc.

 

Also, a lot of people like being able to dial in the fan profile. With Zero RPM fan mode, the fan jumps to a relatively high RPM to bring the temps down fast. If you set your own minimum RPM, you can keep the PSU quieter for a longer period of time before it has to ramp up.

 

Also Jonny, this is not specific to Corsair, but if an end user correctly mounts the PSU in the bottom of their case, with the fan facing upwards, then the internal fan of the PSU is pulling in hot air from the nearby graphics card to cool the PSU? Many cases made in the last ten years have grills or vents in the bottom which means the PSU should be mounted inverted (upside down) with the internal fan of the PSU pulling in cool air from under the case (unless the case sits directly on the floor with no offset or feet). Of course the critic here would say the caps are rated at 105 deg C but still, a 3 or 4 deg delta for free is worthy of consideration.

 

Everyone's case is different and different graphics card coolers cool differently. A lot of our users have water cooling, so hot air circulating inside the case is not an issue.

 

Sucking in hot air shouldn't be an issue as a good PSU should be able to operate normally in 50°C temperatures. But because ambient temperature is one of the sensors used to determine the fan speed, it can make the fan spin faster.

 

it's not unusual to mount PSUs in the down position with fan facing upwards. People with pets or who place their computer on carpet (may they be flogged for this heresy) can do that to avoid clogging the internals with dust and dog hair.. or just to have air at all, in case of carpet.

 

For caps, in efficient PSUs like this one, temp is of no concern really.

rule of thumb is lifespan doubles for every 10° saved from the max rating.

 

take your average Nippon chemicon rated at 2000h at 105°

4000 at 95, 8000 at 85 etc...

 

if you make it run at 30 instead of 35, you're still well within the warranty period. They usually don't heat up anyway (lets be serious, we buy oversized)

 

As for failure rate, since i joined this forum i have seen lots of problems about low end PSUs and AXi. As low as Corsair has fallen in my personal view, HX/HXi still are workhorses afaik. but well, i speak from an experience of two PSUs. an old 850 that i sold and still works perfectly for my friend and a 1200i that has been flawless. Flukes happen ^^'

 

Unfortunately, PSUs are more than a box of capacitors. You have to look at the overall temperature rating of the PSU. The caps actually are some of the coolest running parts in a PSU. And when a PSU fails, it is RARELY because of a failed capacitor.

Even PSUs rated at 25°C can have 105°C caps in them.

 

For example: The main transformer can get as hot as 90°C under heavy load.

 

That's why if you get a PSU that's rated at 40°C or 30°C, etc., you have to take the de-rate from full power capability into account.

Edited by jonnyguru
Link to comment
Share on other sites

  • Corsair Employees

To my point (about the end user reviews):

 

The HX1200i has been on the market for 7 years. There's only 42 reviews on Newegg.

 

Do you honestly thing only 42 units have been sold over the 7 year period?

 

Amazon certainly has more reviews, because they "twist" multiple models into the same review pool... 1282 reviews total. You think Amazon has only sold 1282 units? More like 1282 units PER DAY.

Link to comment
Share on other sites

Hey Jonny, I agree with you that Amazon and Newegg, etc... only represent a small sampling but I also believe that the return rate for several models of Corsair PSUs is higher than 1%, which is not what it should be if they were designed correctly. (It should be less than 1%, ... it only rises to 1% to realistically account for shipping damage from carriers outside of Corsair's control.) I work mainly in pure research from the lab (I am an electrical engineer but I mainly work with mathematical models of human like neural networks. It has been 20 years since undergrad when classwork dealt with power supply basics, so a tad rusty recalling the exact specifics.) but I talked to one of our tech guys to ask about failures yesterday. He checked our purchasing logs this afternoon and got back to me, .... our lab has used 57 Corsair PSUs during the last 3 years with 11 of them failing. Myself personally, I have used 7 PSUs in the last ten years roughly, 1 Thermaltake, 3 EVGA and 3 Corsair. This Corsair HX1200i i am considering returning is dead and the 650 watt unit I used in my wife's rig started acting strangely after about 14 months back in roughly 2012. That is 2 out of 3 that have caused me problems. (My HX1000 is doing fine after 5 years.) If someone thinks that those failures might turn me into a Corsair hater, ... I'm not. I have probably spent $4000 roughly on Corsair products in just the last 4 years or so. Never had a Corsair Dominator or Vengeance dimm fail me yet, my ML 140 fans work great, my H150i is awesome, my 275R and 330 cases are not bad (could use a little extra room on the back side and in the top for AIO watercoolers), ..... the PSUs, .... eh, not so much. To spend $311 (I found the receipt) for an HX1200i and watch it fail, ..... very pissed off. As LeDoyen said, flukes happen and it is not the end of the world. It is just that I can replace a bad fan in 5 minutes. Taking apart an entire system to remove all the wiring and then rewiring it back is a major PITA. I also use that particular rig to make calculated buys for specific financial options, .... and if it is down, that is potentially money lost from pocket. Edited by jayhall0315
Link to comment
Share on other sites

I should have mentioned above Jonny that I have read many of your reviews and always appreciated their clarity and often depth. Despite the hassle of dealing with this dead PSU, If am glad to see that you are working with Corsair to improve things.

 

One thing I notice by looking at pics of other's rigs that may help Corsair out a bit, is enthusiasts who are using one PCIe 6 + 2 pin plus 6 + 2 pin cable to plug into both sockets of a high end card like the RTX 2080 Ti. If I recall correctly (and I might not), the wiring in a single cable is rated safely up to 288 watts but folks should be using two separate cables. That way the PSU can safely deliver ~ 288 to 500+ watts for an overclocked GPU. It is mentioned with one brief sentence in the instruction pamphlet to plug PCIe (PEG) cables into the GPU power sockets but I think some percentage of the failures Corsair is seeing, are folks using high-end GPUs to boot up into Windows just fine with things moving along swimmingly until they load into a game or 3DMark overclocked. It is bc some of these folks are using one cable I suspect.

 

I can also 100% confirm that the Power On Self Test is not accurate. Pressing the fan self test button on the back of my HX1200i does indeed spin the fan just fine on a completely dead unit :)

Edited by jayhall0315
Link to comment
Share on other sites

When i run my 2080 overclocked, it rarely peaks beyond 250W. Maybe it's a concern for TIs and some old AMD GPUs that drained stupid current.

I recently moved to two PCIE cables and the only effect was slightly less voltage losses. It went from like 11.7v at worst loads on one cable, to 11.9 with two cables.

Still better to connect them properly of course, but the PSU shouldn't care much about joule losses in cables, within reason.

It's the components that will misbehave if voltage drops too much, as we see with daisy chained 5V RGB regularely here (why my fans flicker when i set all 12 of them to full white??)

Link to comment
Share on other sites

  • Corsair Employees
Hey Jonny, I agree with you that Amazon and Newegg, etc... only represent a small sampling but I also believe that the return rate for several models of Corsair PSUs is higher than 1%, which is not what it should be if they were designed correctly. (It should be less than 1%, ... it only rises to 1% to realistically account for shipping damage from carriers outside of Corsair's control.)

 

I have no idea how you come to the conclusion that the failure rate is that high. It is LITERALLY much lower than 1% and that ACCOUNTS for shipping damage and often some user error.

 

I currently work for Corsair in the PSU department. I know the real numbers.

 

I work mainly in pure research from the lab (I am an electrical engineer but I mainly work with mathematical models of human like neural networks. It has been 20 years since undergrad when classwork dealt with power supply basics, so a tad rusty recalling the exact specifics.) but I talked to one of our tech guys to ask about failures yesterday. He checked our purchasing logs this afternoon and got back to me, .... our lab has used 57 Corsair PSUs during the last 3 years with 11 of them failing. Myself personally, I have used 7 PSUs in the last ten years roughly, 1 Thermaltake, 3 EVGA and 3 Corsair. This Corsair HX1200i i am considering returning is dead and the 650 watt unit I used in my wife's rig started acting strangely after about 14 months back in roughly 2012.

 

11 out of 57 PSUs failing is very unusual. I'd have to wonder what the application is if this number is real.

 

I wish they updated this article (they used to every year. Maybe they did and I didn't know about it?) but in 2017 the RETURN RATE (not failure rate, because their source doesn't actually test returns and puts everything in the same "bucket") and you'll see that the return rate is quite low: https://www.hardware.fr/articles/962-3/alimentations.html

 

Mind you, none of the product is more than 2 years old because EU law only requires their source to take product back from customers during the first two years (wish the U.S. had the same kind of consumer protection laws!)

 

Corsair would quickly go out of business if they had a 19% failure rate (11 out of 57) on a product they're only making 15 to 20 points of margin on.

 

And I have to ask, how does a PSU "act strangely"? Doesn't always power on when you press the power button on the case? Because that would be a motherboard issue, not a PSU issue. I'm truly curious.

Link to comment
Share on other sites

  • Corsair Employees

I can also 100% confirm that the Power On Self Test is not accurate. Pressing the fan self test button on the back of my HX1200i does indeed spin the fan just fine on a completely dead unit :)

 

 

Because it's not a power on self test. It's just a fan test. The +5VSB is always live on an ATX PSU. This provides power to the fan for self test. If they didn't use the +5VSB, they couldn't have a fan test that worked even if the PC was off.

Link to comment
Share on other sites

Because it's not a power on self test. It's just a fan test. The +5VSB is always live on an ATX PSU. This provides power to the fan for self test. If they didn't use the +5VSB, they couldn't have a fan test that worked even if the PC was off.

 

Yeah sure Jonny, you and I understand that, but that is what is written in the instructional HX1200i pamphlet. A newer user with less experience though is likely to be fooled if for example their unit was damaged during shipping but the power on self test is working. They would then naturally assume things are good to go and wire up everything only to be left in bad spirits wondering why the motherboard will not POST. I think Corsair should include a mandatory jumper block for the ATX 24 pin header to confirm everything is alright for new users.

 

As to our lab computers, it is a mix. ~ 30% are desktops, 30% are more for software development and ~40% are machines running Quadro cards for nn modeling. Our lab has had particularly horrible results with AX units. We will not use them anymore.

 

As to my wife's old computer, the Corsair unit was exhibiting random failures. It took several days to eventually narrow the problem down to the PSU (we initially thought it was a bad EVGA graphics card). We replaced it with an Antec unit and the desktop continued to work fine for three more years until we upgraded all the hardware.

Edited by jayhall0315
Link to comment
Share on other sites

Depending on the type of lab, there can be external causes for that too.

In factories, the mains supplies are usually well separated, but in labs we see some horrors sometimes. Like electronics frying because there's big inverters, motors, high power RF generators on the same mains used for more delicate electronics.

Link to comment
Share on other sites

  • Corsair Employees
Depending on the type of lab, there can be external causes for that too.

In factories, the mains supplies are usually well separated, but in labs we see some horrors sometimes. Like electronics frying because there's big inverters, motors, high power RF generators on the same mains used for more delicate electronics.

 

Excellent point.

 

My mains are pretty clean, but in the office we always use a Chroma power source regardless. And at home I use a gigantic rackmount Trip-Lite Isobar. :D

Link to comment
Share on other sites

  • Corsair Employees
Yeah sure Jonny, you and I understand that, but that is what is written in the instructional HX1200i pamphlet.

 

I'm looking at the pamphlet now.

 

Top of page 8.

 

It says it's a fan self-test multiple times. It says nothing about being a PSU test.

Link to comment
Share on other sites

Excellent point.

 

My mains are pretty clean, but in the office we always use a Chroma power source regardless. And at home I use a gigantic rackmount Trip-Lite Isobar. :D

 

Mine is behind a Furman PL8C i used for music (got rid of ground noises and pops.. our mains go as low as 190V in winter here.. 235 in summer.. it's a mess, super unstable and lots of spikes when heaters kick in and stop). Mains shenanigans kills things..

Link to comment
Share on other sites

  • Corsair Employees
Mine is behind a Furman PL8C i used for music (got rid of ground noises and pops.. our mains go as low as 190V in winter here.. 235 in summer.. it's a mess, super unstable and lots of spikes when heaters kick in and stop). Mains shenanigans kills things..

 

Sounds like you need a Tesla Powerwall after your meter. LOL!

Link to comment
Share on other sites

the whole neighborhood is overloading the local transformer. Used to have lots of power cuts in winter. What i need is a friggin tokamak!

That said, the HX850 didn't mind at all, and the 1200i couldn't care less either.

 

One thing i always wondered.. the mains voltage readout is a readout? or does it only display high voltage as 230 and low as 120?

Link to comment
Share on other sites

  • Corsair Employees

One thing i always wondered.. the mains voltage readout is a readout? or does it only display high voltage as 230 and low as 120?

 

The AXi shows actual input voltage. HXi and RMi only shows hi or low as 115V or 230V.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...