The SynJet Project - It sucks and blows

[jeshikat]

I'm only running these tests for short periods of time so I'm not too worried about frying a chip. A dummy CPU has the advantage of being more repeatable between tests but it's also less representative of what people would actually see. For instance, the CPU die is only a small area under the heatspreader.

I'll compile my notes tomorrow but for now it looks like the Corsair rad + SynJet isn't going to work for 65W+ CPUs. It may be fine on low-wattage chips but at that point it'd make more sense to use a normal low-profile heatsink.

 

[jeshikat]

Ok, don't have any pics for now but here's the results from my testing with the Corsair H5 SF radiator.

System specs:

• Intel i7-4690K (overclocked to 4.2GHz)
• ASRock Z97E-ITX/ac
• 8GB RAM
• Seasonic X-850
• Some SSD
• Windows 10

This is all just sitting in open air on top a table. Ambient temps measured with a thermocouple thermometer and CPU temps and RPM taken from the ASRock tuning utility. Measurements taken after 15 minutes. Load test is Prime95 Blend w/ 4 threads.

For the SynJet test, I had both the radiator and SynJet sitting on top an SSD box, with the SynJet positioned about 20mm away from the radiator.

And here's a sound clip comparing them:



Each section is 10 seconds long:
0:00-0:10 Noctua L9x65 20% ~820 RPM
0:10-0:20 Noctua L9x65 50% ~1560 RPM
0:20-0:30 Noctua L9x65 100% ~2550 RPM (forgot to measure at 75%)
0:30-0:40 Corsair H5 SF 20% ~1020 RPM
0:40-0:50 Corsair H5 SF 50% ~1200 RPM
0:50-1:00 Corsair H5 SF 75% ~1380 RPM
1:00-1:10 Corsair H5 SF 100% ~2130 RPM (watch out if you turned up the volume for the earlier clips, this one is loud!)
1:10-1:20 SynJet XFlow 42 100%

Edit: Actually, here's one pic showing the room at work where I was testing. This view is from my desk and you can't actually see it from here but the test rig is at the end of the table where the arrow is, about 9m (30ft) away.

So maybe I'm just overly sensitive but I can hear the Corsair idling at 20% from my desk if I turn my head. This is with all the computers other than mine and the test rig turned off and it's after hours so it's pretty quiet but it's not an anechoic chamber or anything.

 

[iFreilicht]

The Corsair is stupidly loud, and while the SynJet is audible, it has a much nicer tone to it than the L9x65. More humming than whizzing.

 

[Tek Everything]

Awesome job with the testing, The Synjet performance is disappointing, but not totally unexpected.

The H5 SF really does work well as a cooler. It would be perfect for people who use headphones/headsets ha.

 

[jeshikat]

I'm probably going to return the Corsair since I'll never use it in one of my builds. But I can still measure it and design a bracket if you're still interested in the SynJets for your build.

I think they could be useful for a low-power i3 setup but an i7 plus video card doesn't look feasible.

 

[Tek Everything]

I don't think it will work for me. I am most likely switching cases so I can use a more powerful GPU anyway. So sad... Maybe the H5 SF V.2 will be better.

 

[jeshikat]

No worries, I've been curious about these things so consider the cat killed

I still have some time before I have to return the H5 SF so if anyone wants me to check something with it please let me know. Also, if anyone wants me to test something with the SynJets I can try that as well.

And if you're in the USA and want to take these SynJets off my hands, PM me.

 

[EdZ]

Is the h5 SF block wide enough that you could stuck a 'sideways' ducted 120mm/140mm high static pressure fan perpendicular to it to force air in? Probably not very efficient, but it might still be quieter than the cheap blower.

 

[Tek Everything]

Is the h5 SF block wide enough that you could stuck a 'sideways' ducted 120mm/140mm high static pressure fan perpendicular to it to force air in? Probably not very efficient, but it might still be quieter than the cheap blower.

It is 120mm wide, so yes you could do that. Like you said though, I don't know how efficient it would be.

 

[iFreilicht]

Is the h5 SF block wide enough that you could stuck a 'sideways' ducted 120mm/140mm high static pressure fan perpendicular to it to force air in? Probably not very efficient, but it might still be quieter than the cheap blower.

It would be so funny if that actually worked better. The only thing you really have to worry about is making the duct completely tight so all of the airflow from the fan actually gets through the radiator.

 

[QinX]

I did a test with H2O-Micro with a duct system, to get the top clean and the exhaust out the back. Unless you have decently high duct with good static pressure you could do it. but is no where close to efficient. I'd look into a different higher quality blower fan and have 2 intakes for the fan like the ASUS GTX970 turbo/

PS: Do we know the manufacturer for the fan? Or is it a corsair designed fan?

 

[jeshikat]

I had actually tried that with the 92mm fan:

I don't have my notes right now but temps weren't great and if you have room for this you have room for a small tower cooler anyway.

 

[jeshikat]

I believe @Tek Everything mentioned the fan is a Sunon, don't remember the model #.

 

[Tek Everything]

 

[confusis]

After all the testing, did you figure out any use cases in the PC world where the synjet would work well?

 

[jeshikat]

Low-wattage CPUs, maybe low-wattage GPUs, and spot-cooling.

For that last option, I think it could possibly be used in combination with ducting for cooling M.2 drives that are on the back of motherboards.