Personal computing discussed

Moderators: renee, Starfalcon

 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Enermax Liqtech TR4: potential issues

Thu Jun 21, 2018 3:36 pm

Hey everyone,

For the past few weeks, I've been "that guy" talking about the Enermax Liqtech TR4 on the news pages, which as far as I know, is the only AIO liquid-cooling solution with a custom-fitted coldplate specific for the TR4 (and SP3 EPYC) platform. As such, it got great reviews and great cooling performance. In fact, that's the reason why I bought myself a 280mm version of the Liqtech TR4, it seemed like the best choice for cooling.

Alas, I'm here to suggest that other Threadripper builders stay away from the TR4. There are some issues popping up in forums everywhere and the number of 1-star reviews on this unit continues to grow.

I started to investigate greater issues when my Threadripper 1950x at stock settings was at 68C on Prime95 tests. 68C is actually the "throttle temperature" on Threadripper, so this means I have zero headroom on my build. Further testing puts my idle temperatures around 44C, which is higher than expected. I replaced the thermal paste and checked for simple errors: but the temperatures remained the same. Most worrying: the pump / CPU block is quite hot after these tests, suggesting that maybe the liquid isn't flowing as well as it should. I'm still investigating what's going on with my unit, but its clearly not performing as well as it should right now.

But as I looked up other user's experiences, things look... bad. Very bad across the internet.

Alas, this is my first post on the forum, so I can't post links yet. I'll get those links in asap.

The "exploding radiator" is certainly a story that scares me. But even more scary is the consistent corrosion-related issues that are being reported in a number of channels. The story is all the same: after a few months, performance decreases to the point where the AIO cooler is completely unable to operate.

------------

Fortunately, it looks like the AIO cooler is relatively easy to maintain and swap out the fluid: youtu.be/0fobND-PZUg

But this is not the experience I wish on any beginner. AIO kits are supposed to make all the hard work already done for you. If you gotta buy fluid and refill the damn thing with proper anti-corrosion fluid for it to last more than 6 months... then its not really worthwhile as a beginner-friendly AIO form factor.

For now, it seems like the only legitimate options for Threadripper are to buy air-coolers (like CoolerMaster or Noctua's TR4-specific heatsinks), or to go $300+ full-custom liquid. I guess its reasonable to buy the Enermax Liqtech TR4 models if you're willing to swap out the fluid yourself (which would still be cheaper than full-custom liquid), but that's a hard sell IMO.
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Enermax Liqtech TR4: potential issues

Thu Jun 21, 2018 4:01 pm

Nostalgia isn't what it used to be.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Thu Jun 21, 2018 4:08 pm

Thank you very much just brew it!

Yeah, after I pushed this particular model so hard on the news articles, I feel like I need to "undo" the damage and make sure people out there know about this issue. 68C is certainly not normal, even the Noctua air-coolers are allegedly getting 63C at stock Prime95 tests.

At this point, I'm thinking I need to go full custom loop to reach the performance I desired.
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Enermax Liqtech TR4: potential issues

Thu Jun 21, 2018 4:12 pm

Hopefully it is a production/QA glitch which they will quickly rectify. Sucks for the people who already bought the problematic units though. Keep us posted on how your RMA goes (assuming you pursue one).
Nostalgia isn't what it used to be.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Thu Jun 21, 2018 4:42 pm

just brew it! wrote:
Hopefully it is a production/QA glitch which they will quickly rectify. Sucks for the people who already bought the problematic units though. Keep us posted on how your RMA goes (assuming you pursue one).


Alas, RMA is not a reasonable option for me. I can't afford any downtime as one project I'm working on has a deadline soon. The only paths forward are:

1. Buy a Noctua air cooler -- If I RMA, I'll need the air cooler to hold me over. But frankly, I don't give a damn about getting another Liqtech TR4 that would have the same issues, and that so many people online seem to have issues with. So at that point, I'd just stick with the air-cooler. Since Noctua air coolers cost $80 and the TR4-specific heatplate beats out the non-TR4 specific coolers (some tests show the Noctua U14S beating out 360mm Liquid !! due to the heat-plate issue) this is the standard option.

2. Buy a custom EKWB / Thermaltake / something custom cooler -- These also have a proper TR4-sized heatplate. But that will put me out $300 to $400 or so for all of the parts.

----------------

If I were to guess what is going on right now: I'm guessing the pump has failed. There are a ton of 1-star Newegg reviews about how the pump is unreliable on this system. Plus, with how hot the water-block was, its just more evidence for this suspicion.

Fortunately, the thing works, just poorly, for now. I've got some time to think things over and form a strategy for what I want to do.
 
DPete27
Grand Gerbil Poohbah
Posts: 3776
Joined: Wed Jan 26, 2011 12:50 pm
Location: Wisconsin, USA

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 9:34 am

Jeez. Watched that youtube video. Did they fill it with tap water and bleach?
Main: i5-3570K, ASRock Z77 Pro4-M, MSI RX480 8G, 500GB Crucial BX100, 2 TB Samsung EcoGreen F4, 16GB 1600MHz G.Skill @1.25V, EVGA 550-G2, Silverstone PS07B
HTPC: A8-5600K, MSI FM2-A75IA-E53, 4TB Seagate SSHD, 8GB 1866MHz G.Skill, Crosley D-25 Case Mod
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 10:54 am

I watercooled PCs almost two decades ago and I learned many lessons, but the most disgusting and catastrophic lesson I learned was this.

NEVER MIX METALS IN A WATER LOOP.

All of these AIOs with copper blocks and aluminium radiators make me cringe. I've had to repair PCs (read: replace most of the internals) after failing AIOs, There were a couple of H60s and an H100 - and I'm not saying that Corsair are sub-par, I'm saying that ALL of these mixed-metal AIOs are sub par and you're just likely to see more Corsairs than anything else because they outsell anything else.

The biggest problem with corrosion furring/gunking up your loop is that CPUs are capable of boiling the water, which causes enough pressure to either burst a hose, pop a hose loose, or in the case of plastic Asetek designs, blow the pump/block apart :(
Last edited by Chrispy_ on Fri Jun 22, 2018 11:05 am, edited 1 time in total.
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 11:01 am

DPete27 wrote:
Jeez. Watched that youtube video. Did they fill it with tap water and bleach?


Yeah, something like that. For a review site to get corrosion in a barely-used demo-unit is incredibly worrisome.

Chrispy_ wrote:
I watercooled PCs almost two decades ago and I learned many lessons, but the most disgusting and catastrophic lesson I learned was this.

NEVER MIX METALS IN A WATER LOOP.

All of these AIOs with copper blocks and aluminium radiators make me cringe. I've had to repair PCs (read: replace most of the internals) after failing AIOs, There were a couple of H60s and an H100 - and I'm not saying that Corsair are sub-par, I'm saying that ALL of these mixed-metal AIOs are sub par and you're just likely to see more Corsairs than anything else because they outsell anything else.


Well, EnerMax Liqtech TR4 is certainly copper-block / aluminum fins. But I thought there needs to be an electrical connection between the two metals for an issue to occur? Is the liquid conductive enough for the galvanic corrosion to happen? Galvanic corrosion needs electricity to flow between the two units. Its a big issue with mixed-metal radiators or mixed-metal blocks (Nickle/Copper barely has a voltage differential, so its okay).

But with just a pipe of liquid between the radiator and copper block, I find it unlikely that electricity would flow between the two.

Furthermore, the demo-unit showed corrosion on the screws in the Youtube video. Water itself is incredibly corrosive (see rust), so IMO its likely they just didn't use a proper liquid for this cooler.
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 11:07 am

Unless the water in the loop is very pure distilled water, it is conductive. Even distilled water will become conductive over time as trace amounts of the metal dissolves in it.
Nostalgia isn't what it used to be.
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 11:28 am

You don't need electricity to suffer from galvanic corrosion; When you put two metals in an electrolytic solution (like water) and both are electrically connected (the copper block is grounded to the IHS of the CPU, and the radiator is also grounded to the case via screws) ions move from the more electrochemically-reactive metal to the less electrochemically-reactive metal.

Image

Electrolysis is similar, but accelerates the process by massively increasing the flow rate of ions through a potential difference between anode and cathode metals.

Two metals in solution is the fundamental way most batteries work, so you're effectively creating (a very weak) battery cell with your water loop and the battery is depleted once all the electrochemical difference between the anode and cathode has been reduced to zero. Sadly your AIO will fail long before that happens in most cases, and the coolant additives are designed to slow down this irreversible process. The only way to truly avoid it would be to use a non-electrolytic fluid for cooling, like liquid polymer or something, and since polymers are good insulators they would be awful for cooling!
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 11:47 am

Chrispy_ wrote:
the copper block is grounded to the IHS of the CPU, and the radiator is also grounded to the case via screws


I can buy that as a sufficient electrical connection for galvanic corrosion.

Well, thanks for the analysis. I'm still in the middle of deciding custom loop vs air cooling. But you've only confirmed my belief that I need to replace this thing ASAP before it fails.
 
Aranarth
Graphmaster Gerbil
Posts: 1435
Joined: Tue Jan 17, 2006 6:56 am
Location: Big Rapids, Mich. (Est Time Zone)
Contact:

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 12:36 pm

My old dangerden system used a copper block and copper radiator.

That's one way to make sure you have no issues with galvanic corrosion.

(Edited B/C I cannot spell today!)
Main machine: Core I7 -2600K @ 4.0Ghz / 16 gig ram / Radeon RX 580 8gb / 500gb toshiba ssd / 5tb hd
Old machine: Core 2 quad Q6600 @ 3ghz / 8 gig ram / Radeon 7870 / 240 gb PNY ssd / 1tb HD
 
DPete27
Grand Gerbil Poohbah
Posts: 3776
Joined: Wed Jan 26, 2011 12:50 pm
Location: Wisconsin, USA

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 3:13 pm

What if just the circulation tubes through the rad are copper and the rad fins are still aluminum? Then it's essentially the same as any air cooler with heatpipes. Is it rare to see such a rad?
Main: i5-3570K, ASRock Z77 Pro4-M, MSI RX480 8G, 500GB Crucial BX100, 2 TB Samsung EcoGreen F4, 16GB 1600MHz G.Skill @1.25V, EVGA 550-G2, Silverstone PS07B
HTPC: A8-5600K, MSI FM2-A75IA-E53, 4TB Seagate SSHD, 8GB 1866MHz G.Skill, Crosley D-25 Case Mod
 
thecoldanddarkone
Minister of Gerbil Affairs
Posts: 2449
Joined: Wed Mar 26, 2003 4:35 pm

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 4:14 pm

DPete27 wrote:
What if just the circulation tubes through the rad are copper and the rad fins are still aluminum? Then it's essentially the same as any air cooler with heatpipes. Is it rare to see such a rad?


Fins can be aluminum as long as the flat tubes are copper. Basically as long as the water doesn't touch the aluminum.
I7 4930k, 32 GB Ballistix DDRL3@2133 , 1.2 TB Intel 750 AIC, 500 GB mx200, Sapphire R9 Fury, asus x79 ws, HP ZR24w, edifier s730
HP Pro x2 612- i5-4302Y, 8 gigs of memory, 256 ssd
 
jihadjoe
Gerbil Elite
Posts: 835
Joined: Mon Dec 06, 2010 11:34 am

Re: Enermax Liqtech TR4: potential issues

Fri Jun 22, 2018 9:55 pm

It might be possible to replace just the radiator instead of junking the whole thing.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Sat Jun 23, 2018 11:48 pm

jihadjoe wrote:
It might be possible to replace just the radiator instead of junking the whole thing.


Certainly not with this unit. Its a closed-loop all-in-one.

There are some expandable designs out there. But Enermax Liqtech TR4 is NOT one of them.
 
Shobai
Gerbil First Class
Posts: 165
Joined: Sat Sep 03, 2005 1:18 am

Re: Enermax Liqtech TR4: potential issues

Sun Jun 24, 2018 12:03 am

It's probably not the kind of stuff you want in a production machine, but I've successfully swapped the radiator on a friend's Intel-branded closed loop cooler where the coolant crystallised.

It was a fair bit of work cleaning out the gummy junk, to then determine whether the pump + coldplate was worth salvaging, but the swap itself was easy enough: determine tubing size, buy a couple of G1/4 barbs to suit, install barbs onto new copper radiator, fill with distilled water. That unit has been running with no issues for several years since.
 
jihadjoe
Gerbil Elite
Posts: 835
Joined: Mon Dec 06, 2010 11:34 am

Re: Enermax Liqtech TR4: potential issues

Sun Jun 24, 2018 4:25 pm

It's very doable. I've done the same to an old Corsair H70 and replaced the tiny radiator with an HWLabs 240.

At the end of the day an AIO is just an integrated block/pump and a radiator. How hard can it be?
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Tue Jun 26, 2018 4:16 pm

Minor update: I've decided on the Noctua U14S TR4. Its shipping and will be in sometime this week. I'll let yall know how my temperatures do with it.

I splurged a bit and got a 2nd Noctua A15 fan to go with it, so that I can run the air-cooler in a push-pull configuration. In addition to the 2x case fans (which are replacing the radiator+2x fans), the overall cost of the Noctua U14S is actually going to be roughly the same price as the Enermax Liqtech TR4.
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Mon Jul 02, 2018 10:33 pm

EDIT2: Okay, ignore what I said before, it was all stupid. Now that I correctly configured the Noctua fans to spin 90% at 60C and a "Critical" temperature of 65C (100%). (My first post wasn't configured correctly).

~5 minutes in, temperatures are 66.5C, all-cores 3.7GHz. At ~10 minutes in, temperatures are now 67.5C and clocks are beginning to drop to 3.5GHz (suggesting I've begun to throttle).

The Enermax Liqtech TR4 in comparison hits 68C rather quickly, and drops to 3.4GHz within 5 minutes.

So yeah, Noctua isn't perfect, but its certainly better than the Enermax Liqtech TR4 somehow. I'm going to blame the pump, because there really should be no reason for the Noctua to beat the liquid cooler. In any case, the cheaper price, the better case compatibility, and the safety (no liquid to spray inside my case) means that Air >>> Liquid in this round. I'm still looking at the nice temperatures people are getting from full-custom setups however (below 60C stock, below 70C with major overclocks) and I do feel a little bit jealous.

Anyway, I have a carpet and the computer is under my desk. So airflow around the case is not as ideal as it could be. But still, the air vs liquid comparison was held in consistent situations (same case, same carpet, same location under my desk), so its about as fair a comparison as anyone can hope for. EDIT: My ambient is also 76F (25C), so its +4C or so above what other testers ambient (~21C). That's probably the cause of the discrepancy.
 
Amiga500+
Gerbil
Posts: 76
Joined: Wed Sep 21, 2016 2:10 am

Re: Enermax Liqtech TR4: potential issues

Thu Oct 18, 2018 3:05 pm

I had a Liqtech TR4 280 mk1 fail on me - performance gradually degraded to the point I was getting about 1 GHz at best across any core! No leaks and no danger to the CPU thanks to AMD's thermal throttling.

Got a mk2 via RMA, installed it yesterday and straight back to work, been running flat out now for about 24 hrs - temperature is 64.75 degC - bouncing off the socket 180W power limit (haven't moved it from default) - at a clock speed of ~3.7 GHz across the 16 cores.


[Air circulation around the machine isn't great, tucked under a desk in a wee room - resting a thermometer on the upper surface of the box gives an air temp of around 36 degC.]
 
dragontamer5788
Gerbil Elite
Topic Author
Posts: 715
Joined: Mon May 06, 2013 8:39 am

Re: Enermax Liqtech TR4: potential issues

Mon Oct 22, 2018 11:14 am

Amiga500+ wrote:
I had a Liqtech TR4 280 mk1 fail on me - performance gradually degraded to the point I was getting about 1 GHz at best across any core! No leaks and no danger to the CPU thanks to AMD's thermal throttling.

Got a mk2 via RMA, installed it yesterday and straight back to work, been running flat out now for about 24 hrs - temperature is 64.75 degC - bouncing off the socket 180W power limit (haven't moved it from default) - at a clock speed of ~3.7 GHz across the 16 cores.


[Air circulation around the machine isn't great, tucked under a desk in a wee room - resting a thermometer on the upper surface of the box gives an air temp of around 36 degC.]


Its the "degraded" status that worries me. Even if I do a bunch of tests when I buy one, there's no guarantee that it won't get worse over time. Hopefully the Mk2 works out for you! For me personally, the Noctua is sufficient, even if the temperatures are probably higher than a properly functioning liquid cooler.

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On