| 23 May 2025 |
 emily | but I suppose if you are putting 10GBASE-T stuff in there then it's going to get toasty. | 12:02:49 |
| emily | thankfully I figured out I can just wire fibre :) | 12:03:02 |
 adamcstephens | Manufacturers only have to be comfortable with a design during the length ofa warranty | 12:04:40 |
| adamcstephens | * Manufacturers only have to be comfortable with a design during the length of a warranty | 12:04:48 |
 matthewcroughan @fosdem | doesn't have to be though | 12:07:01 |
| matthewcroughan @fosdem | as you found, there are ways to get below 50C on 10G | 12:07:08 |
| emily | I'm pretty sure even the good Broadcom 100 m 10GBASE-T modules are way hotter than fibre ones are gonna get. | 12:07:28 |
| matthewcroughan @fosdem | yeah, ofc, but 50C is fine | 12:07:37 |
| matthewcroughan @fosdem | the normal is much hotter | 12:07:46 |
| matthewcroughan @fosdem | normally it's 68+ | 12:07:49 |
| matthewcroughan @fosdem | 50C is an outlier, and I suspect it needs to be 50C or below to warrant a fanless design | 12:08:11 |
| matthewcroughan @fosdem | or at least safe for human hands | 12:08:25 |
| emily | sure I just mean if you're packing them into every port then sure, I'd worry about cooling. I wouldn't necessarily feel secure with the fanless TP-Link either though. | 12:08:34 |
| matthewcroughan @fosdem | Why would they sell a fanless 8 port SFP if it weren't tested | 12:08:59 |
| matthewcroughan @fosdem | * Why would they sell a fanless 8 port SFP if it weren't tested in that configuration | 12:09:08 |
| matthewcroughan @fosdem | it's not like you can add fans to it, it's designed to be fanless without the option of cooling it externally | 12:09:26 |
| adamcstephens | Because it meets their requirements for doing so, but that doesn't mean it will have as long of a life as one with a fan | 12:09:36 |
| matthewcroughan @fosdem | sure, lifespan, I'm concerned with human safety, not lifespan of switch | 12:09:52 |
| emily | probably they tested it not entirely full of copper modules | 12:10:17 |
| matthewcroughan @fosdem | I mean, would they not have to regulatorily? | 12:10:29 |
| emily | I mean… the actual risk we're talking about is it shutting off because of overheating, I presume. | 12:10:39 |
| matthewcroughan @fosdem | I find it hard to believe they could manufacture electronics like this without testing that edge case | 12:10:50 |
| matthewcroughan @fosdem | a friend in hackspace has to do way more stringent tests to get approved on his dainty little RFID products | 12:11:07 |
| emily | I assume it's not going to start a fire… but I also assume you want the switch to keep running | 12:11:22 |
| emily | a bunch of RJ45 SFP+ modules will specifically tell you to space them out for heat reasons fwiw | 12:11:29 |
| emily | but anyway I'm not saying it's going to be a problem. I'm just saying I don't think you can say the TP-Link is definitely fine fanless and the XikeStor definitely isn't | 12:11:53 |
| emily | if I was worrying about cooling I'd definitely also be investing in ones with the Broadcom chip! since last time you were fine with one that hit 70 C at 10G | 12:15:10 |
 magic_rb | In reply to @emilazy:matrix.org
magic_rb: is now a good time to ask about your antenna config? :) oh yes, ill send pic | 22:40:50 |
| 24 May 2025 |
|  fromtheeast710 joined the room. | 05:10:19 |
|  prince213 joined the room. | 12:34:18 |
