Argon one V2 Vs ice tower cooler Test. Part 1

preview_player
Показать описание
I wasn’t expecting that!
Save any harsh judgments until you have seen part 2. up tomorrow

Thanks to LilRedDog for the Super thanks. 👍🏻
Thanks to Vil Lev for a Super thanks

As an Amazon Associate I earn from qualifying purchases

Amazon US & other Countries Links

My Amazon UK store

Amazon UK Links

My videos

Great Raspberry Pi 4 Cooling solution

Raspberry Pi 4 NEWS playlist

All of my Pi videos are here (500 plus)
Raspberry Pi 4 Raspbian and more
Рекомендации по теме
Комментарии
Автор

While you are doing tests something important you should check! About a year ago I posted about a flaw in the Argon case that got overlooked. The Argon has great passive cooling. However there is a design flaw in when the magnetic lid is on it creates a "low air pressure" zone under the lid. This causes the the fan to literally suck back in the hot air it is blowing out into a loop! If the magnetic lid is off then most of the heat escapes as it rises and no low pressure develops to pull the heated air back into the unit. With the lid off I could max overclock fine for emulation. But with the magnetic lid on my Argon it always overheated after a long run when the who case heated up over time! I did notify the Argon builders. But they did not think it was an issue at the time.

gene
Автор

My first suspect would be the Python script that controls the Argon One fan. I think it is using a different control algorithm than the ice tower cooler. I think there is hysteresis built into the Ice Tower cooling control. If you keep using the Argon One Script I'd go to the custom option and put in the temperatures and speeds that you really want. Ideally you could use the same fan software on both systems.

bmanske
Автор

Very interesting results, I can't wait for Part 2. Also props to @LilRedDog for the donation towards more testing.

circleofowls
Автор

Love all the Content, I have the Argon One with built in SSD, just love it for its neat and tidy footprint
I run Twister and never looked back 👍

frecko
Автор

I was immediately thinking it must be either the CPU clock or the proportion of video resources allocated from the CPU in the raspi config. Either way - thanks for being amazing and taking the time out of your day to run these tests for all of us!!

italianbasegard
Автор

That’s the same reason why I stopped using the Argon One cases and changed to an open case. Whenever it got too hot it would slow down to a crawl and if I set the minimum frequency to the stock 1500 it would crash the Pi. My Overclock settings are
arm_freq=2100
arm_freq_min=1500
over_voltage=6

Using argon one m.2

KaffeDad
Автор

Possible reasons for different timings:
- processor type B vs. type C
- differences in the storage drives & linkages
- possible overhead from the Argon fan control software, if it uses one of the four cores, leaving only three to run the stress tests.

MarcusPHagen
Автор

Force_Turbo=1 to force the 2100 MHz all the time on both machines. I think it may point you in the right direction. I think it's the SSD causing the Pi to throttle down to the 600 MHz. But, it's hard to say with out more testing.

spuds
Автор

Idk if the Argon one case trounces on all other cases but I doubt I will ever replace mine. They did a great job with the size and design.

Mr.Smith
Автор

Hi Lee, I'm wondering if the differnces you are seeing is to do with the way the Pi's are powered. The Argon case supplies power to the Pi via the GPIO pins at a constant rate, via the 5v pin. Whereas the other pi is plugged dricetly into your power outlet, and is able to pull what it needs straight from the wall.

markfgns
Автор

There's something wrong with the power circuitry in the argon case, If I had the chance to buy the Argon case again knowing what I know now I wouldn't touch the thing

Verightess
Автор

I would think to compare cooling, I would leave the fan on constantly.

stevesweb
Автор

When I'm using my argon one, it always takes quite awhile for that hunk of metal to get up to temperature so I think thats something to look out for. It may be the pi's running cold now but that's cause it hasent heated up yet. But if its not running full 2100, i can see it staying cool

roli
Автор

Begin with swapping what is different, starting with the SSD's, which seem to be the one common difference between the testing. It is such a shame that since the introduction of the Pi4, so many videos have had to deal with the lack of cooling in the fist place, something that Ebon Upton failed to mention was a necessary thing for most.

srvuk
Автор

Hello,
I think that the Argon driver for the fan regulation is modifying the PI speed. Disabling it for the comparison could be interesting.

MosheMiParis
Автор

Looking forward to the answer.

I would swap the Pis leaving everything else the same as my first test. Swapping the SSDs would be the next test, but the images also need to be swapped, so the only difference is the device.

Candidate issues are in CPU governor policy, and interaction between the fan control software and the other activity. That difference in the frequency and power is likely caused by the source of the difference in performance and, therefore, heat.

Another interesting test would be to see what happens with the fans present, but disconnected (or just with no control configured). This will give you some idea of the thermal inertia available.

YeOldeTraveller
Автор

I'd just love to find one on stock

bones
Автор

I think the argon is for a non tinkerer whereas the open cooler is for someone who does more os swapped and gpio etc

dakshrokadia_MTB
Автор

so I have 4 raspberry pi servers in my room in Argon M2 cases and set up to start fans at 50C. During the hot times in the summer I could hear the fans spinning. 10 months out of 12 they are never on - Pi is below 50C. Btw I wanted to replace the fan with Noctua - unfortunately they don't make 30mm fans ;-(
Your case is newer then mines: I don't have access to microSD cards ;-(

zyghom
Автор

Fan failed on my Argon 2 - darned hard to get a reply from the company. I wasn't asking for a replacement- just advice on compatible model. I had to cut wires on a replacement because it had different connector. Darned thing runs continuously.

stephenhookings
visit shbcf.ru