BFG EX-1000 1000 Watt Power Supply

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
BFG Tech has done a very good job at commanding respect in the computer power supply market in the last year. It has stepped up its game and along with that it is now entering the tremendously demanding 1 kilowatt PSU market. 1000 watts of power is not easy to deliver while doing a good job of it.

The BFG EX-1000 carries on in the same fine tradition of the ES-800 and represents an evolutionary but not revolutionary step forward for this design. The EX-1000 bumps up the capacity by 200W and in the process increases the efficiency as well, further strengthening a key feature of this platform.
 
80 to 90C under full load tho? Yikes. I have the es-800 in a spare pc and it seems to be fine. Time will tell if those temperatures are a long term problem (which they could be)
 
80 to 90C under full load tho? Yikes. I have the es-800 in a spare pc and it seems to be fine. Time will tell if those temperatures are a long term problem (which they could be)

We are hoping to hear back from BFG with an official response, but I have heard that there is a new revision for this unit to address many of these issues.

Part of the temperature issue may be the airflow from the unit as you can see in the pictures there are side vents. With most high power units with large fans that move large volumes of air I have to trim the temperature up on the incubator to resettle on 45C at each load step as the fan RPM's increase moving more air with the higher loads and internal PSU temperature. With this unit it was quite the opposite as I was constantly having to trim the set point down at each load step as it seemed the unit was dumping heat back into the incubator rather than expelling it. Nothing was blocking the rear of the unit so I can only guess that airflow was hitting something inside the unit and be directed out of those vents raising the incubators temperature.
 
Hmm. I really wouldnt want that heat exhausting into my case... They should ditch the side vents and try and streamline the airflow out the back.
 
I personally will never buy another BFG power supply. I had a BFG 800 watt and it fried within 6 months. Garbage. I had to steal the P/S and video card from my wife's computer and have her use our laptop until I can afford a new one so I can get my 4870 X2 back in order; since her power supply can't power the beast.
 
I personally will never buy another BFG power supply. I had a BFG 800 watt and it fried within 6 months. Garbage. I had to steal the P/S and video card from my wife's computer and have her use our laptop until I can afford a new one so I can get my 4870 X2 back in order; since her power supply can't power the beast.

People need to learn...when you are buying electronics, there will be some that go bad, it is just the nature of it. How many other computer components have you sworn off because you got a DOA, or it went bad within 6 months?

Guess what...companies also change. You probably bought one of the old TOPOWER units...now they have better units. Why didn't you RMA the BFG to begin with?
 
The "Discuss" link at the end of the article points to the Power Supplies subforum instead of this thread; you might want to change that.

Apparently BFG has added some additional filtering to newer revisions of the EX series, so I'm curious about whether or not that has helped the +12V output quality substantially. As it stands, they still have pretty decent overall performance and very good efficiency at lower loads. They're also quite reasonably-priced for units in their wattage ranges, so that's also a big advantage.
 
What amazes me is that the unit survived at those temperatures.

Not for long, I'd guess, since the main caps in the unit are rated at 85C :)

Having a flawed ('leaky') airflow through the unit would indeed explain why the exhaust temperature keeps jumping up like that, with steps of 7 to even 20C :eek: I would wager a guess that they didn't test this unit at 45C, but at 25C or so and thought it was okay. I wonder what the effect of some tape over those side vents would be... Paul? :D

That said, I must say I really don't like the DC output quality. Any unit which goes over 50% tolerance on any rail is questionable in my book, let alone if it hits 100% tolerance on most rails at around 75%.
 
Another Timely Indepth Review, and one worth reading. I was especially interested in your description of the Soldering Joints, and they are usually a Source of Failure, and the Fact that they withstood such High Temps is good news at that. Those kinds of Observations help make Informed Buying Decisions, and I applaud the fine Work you do at Ferreting out a PSU's Strengths and Weaknesses.
 
You're not really loading it much with that setup, so it's bound to run fine.

Let's hope Paul's unit was a dud then.


Yeah, the only reason I upgraded was for the modular cables. I don't even need half the output. I HAVE loaded it with much more, but I just don't have enough stuff to stress it.
 
I would wager a guess that they didn't test this unit at 45C, but at 25C or so and thought it was okay.

These units are tested and rated at 40C continuous power. Burn in could be at a higher temp than that.

I wonder what the effect of some tape over those side vents would be... Paul? :D

It might help, but my unit is currently in storage so I can't give it a shot at the moment.
 
I would wager a guess that they didn't test this unit at 45C, but at 25C or so and thought it was okay.

You would lose that wager. ;)

The units are ALL tested at 40°C. NOT 25°C. We make mention of this in our copy and on the box itself, but looking at the review it seems that Paul didn't mention what the PSU is rated at by us in the review... only that he ran it at 45°C (which the PSU should still tolerate at < 100% load because it's operating temperature is upwards of 50°C).

You can look at this summary of the EX-1000 @ 40°C we did here: http://www.jongerow.com/BFGpower/EX1000_40°C.doc
 
Okay, so it could reach 100% load at 45C intake, but pretty much doubled the intake temperature for the exhaust, even exceeding the temperature rating of the caps inside the unit with the 100V input test. I really hope it was just a fluke with this particular unit and not characteristic of this PSU in general.

In the document you linked to, what exactly is meant with the MOS and the 'bridge'?
 
Okay, so it could reach 100% load at 45C intake, but pretty much doubled the intake temperature for the exhaust, even exceeding the temperature rating of the caps inside the unit with the 100V input test. I really hope it was just a fluke with this particular unit and not characteristic of this PSU in general.

Not really a fluke. Then again, in an actual PC chassis, you have cool air intake and hot air exhaust. It's not typical for a power supply to be the only thing moving air inside the PC, never mind 45°C air.

In the document you linked to, what exactly is meant with the MOS and the 'bridge'?

"MOS" for Metal Oxide Semi-Conductor, "bridge" for bridge rectifier. About the hottest components inside the PSU.
 
Not really a fluke. Then again, in an actual PC chassis, you have cool air intake and hot air exhaust. It's not typical for a power supply to be the only thing moving air inside the PC, never mind 45°C air.



"MOS" for Metal Oxide Semi-Conductor, "bridge" for bridge rectifier. About the hottest components inside the PSU.

Thanks for clarifying :)

The thing I find interesting is that even with 45C intake air, the Corsair HX1000 barely raised the exhaust temperature: http://www.hardocp.com/article/2008/04/21/corsair_hx1000w_psu_review/4

As Paul suggested, might the side vents of the EX-1000 be the reason for this difference? The efficiency numbers for the BFG unit are significantly higher than those of the Corsair, so it seems strange that the more efficient unit would run so much warmer, in the case of the 100V, 100% load test 33C warmer (57 vs 90).
 
As Paul suggested, might the side vents of the EX-1000 be the reason for this difference? The efficiency numbers for the BFG unit are significantly higher than those of the Corsair, so it seems strange that the more efficient unit would run so much warmer, in the case of the 100V, 100% load test 33C warmer (57 vs 90).

I'm sure it Paul is correct. And you are right in your assumption as well: A more efficient power supply WOULD NOT run hotter than a less efficient one. But if the heat isn't immediately evacuated and/or is allowed to dump back into the chassis, then the hot air is going circulate back into the power supply if it has nowhere else to go.

In a normal PC, it has somewhere else to go. In Paul's test rig, it doesn't. It goes back into the incubator warming up the air in there, requiring Paul to trim down the temperature of the incubator.
 
Taping the vents would increase the downforce and make the car turn better at high speed......ooops, that's NASCAR..........:cool:

The vents undoubtedly work fine in a case but not in an incubator.:D
 
Probably not considering Oklahoma Wolf's review showed similar numbers.

Worse, even: http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story3&reid=151

120V input, 980 Watt load gives 85C exhaust with 39C intake (!!), and his hot box doesn't seem to circulate the exhaust. The 12V3 rail also went out of spec with DC ripple with this test (127 mV), though the modular connection seems to be responsible for this. Wolf added some heavier caps to two 12V rails and had them at ~60 mV ripple, half of what they were before.
 
Well $200 is cheap for a 1000 Watt psu. But is it reliable over the long haul and what about the heat issue?
 
Just bought one of these after my PCP&C 750w Silencer exploded...

The BFG seems very nice so far.

JonGerow:

Can I order more molex cables from BFG? My server needs more molex than available so I'm running a bunch of Y cables, but I'd prefer to spread the load across more of the modular connectors.
 
Can I order more molex cables from BFG? My server needs more molex than available so I'm running a bunch of Y cables, but I'd prefer to spread the load across more of the modular connectors.

Not normally, but PM me and I'm sure I can dig some up for you.
 
Sent PM earlier.

Thanks ahead of time BTW.

Also just a quick shout out to BFG warranty, I had an old 550w unit or something like that which failed, I no longer had any paperwork but they swapped it out with an entirely new unit.

So far I'm very pleased with the EX-1000, it runs much cooler than my old Silencer was running though I have a feeling it was defective from the get go.
 
I just checked my spec sheet and the Best Buy listing is correct. There's only supposed to be 6 Molex (3 fixed and 3 modular). The listing on bfgtech.com looks to be wrong because they copied and pasted from the EX-1200 which DOES have 9 Molex.

But that's cool. I'll get some extra cables out to you Monday just the same.
 
I just picked one up a couple weeks ago went back with my receipt and they gave me the difference back :) , but..... I'm thinking about going back yet again and returning it all together. The high temp issue really does worry me as I have the fan side down only way for me to get the 8 pin to reach without stretching the cable across the MB.

@JonGerrow Please tell me the issues this PSU has have been fixed since its' first appearance in June-July?? A new revision? If so, is there a way to confirm?
 
I just picked one up a couple weeks ago went back with my receipt and they gave me the difference back :) , but..... I'm thinking about going back yet again and returning it all together. The high temp issue really does worry me as I have the fan side down only way for me to get the 8 pin to reach without stretching the cable across the MB.

@JonGerrow Please tell me the issues this PSU has have been fixed since its' first appearance in June-July?? A new revision? If so, is there a way to confirm?

Like I already said in this thread... there is no temperature issue. In the reviews, the PSU is the only exhaust, so it has to exhaust it's heat plus the heat generated by the incubator. The PSU itself isn't generating additional heat. If it were, it wouldn't be as efficient as it is. These PSU's are rated to run continuous at an ambient of 40°C and are made to operate in ambients upwards of 50°C. We've had no failures due to "overheating" and are confident in having a lifetime warranty on this product.

The fact that you have the PSU on the bottom of the case with the fan pointed down is actually a GOOD thing (as long as you're not completely blocking the intake fan) as you're not sucking hot air up off the CPU.
 
Like I already said in this thread... there is no temperature issue. In the reviews, the PSU is the only exhaust, so it has to exhaust it's heat plus the heat generated by the incubator. The PSU itself isn't generating additional heat. If it were, it wouldn't be as efficient as it is. These PSU's are rated to run continuous at an ambient of 40°C and are made to operate in ambients upwards of 50°C. We've had no failures due to "overheating" and are confident in having a lifetime warranty on this product.

Allow me to quote myself:
Worse, even: http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story3&reid=151

120V input, 980 Watt load gives 85C exhaust with 39C intake (!!), and his hot box doesn't seem to circulate the exhaust. The 12V3 rail also went out of spec with DC ripple with this test (127 mV), though the modular connection seems to be responsible for this. Wolf added some heavier caps to two 12V rails and had them at ~60 mV ripple, half of what they were before.

What part of this doesn't scream 'temperature problem'? I find the ripple issue to be highly disturbing as well. Did this get fixed yet?

Props for standing behind the product, though.
 
The ripple problem's been addressed in production units by using larger caps.

Go back one page in that same review and look at the cold tests. The temperature delta is much lower on every test. It's not as if the power supply runs hotter at hotter ambients. If it did, the efficiency numbers would be drastically different. It's because the PSU can't push all of the heat that's being pumped into it.

It's not as if the heat is not coming from any part of the PSU. Any heat that comes off of any PSU is from the power conversion itself. The "problem" in the tests actually come from the fan not pushing enough of the hot air that the PSU is sitting in out of the back of the unit. So this heat builds up inside the unit until only very hot air can be measured coming out of the back. Sure, this problem can be addressed with a faster fan, but the tests aren't realistic enough to justify ramping the fan up on this unit. Power supplies aren't typically the only exhaust in the system and constantly sucking in 40°C plus air. As it is, Andyson initially designed this unit to not spin a fan at all until internal ambient temperatures hit 50°C. I personally didn't like that idea, but do you think they would engineer a PSU that put out that much heat to run fanless until it hit the upper echelon of it's operating temperature?
 
Is this better than Corsair 1000HX?
The original models weren't. I haven't seen a good review of the revised version yet, so we have no way of knowing whether or not ripple suppression has been improved to the level of the HX1000.
 
Back
Top