My finding for a used, high quality PoE switch. 24/48-Port | Gigabit | PoE+ | 390w for around $100
- Thread starter elwood
- Start date
This. If it's configured with default settings, all of the 1-24 ports will communicate with each other. Definitely not the management or console port. If it's not working, they may be on separate VLANs or something. You'll need a console cable to reset things or you may be able to do it from the GUI if you can access it, not sure.
I had a long reply i been working on
Hell of a project you have going on, impressive! I plan on running nothing but CAT5e, i really hope i don't run into a situation where i need CAT6+, it seems like a pain to work with!You and I are walking the same path at different times my friend!I've spent most of this past summer running more dedicated CAT-6, CAT6A, CAT-7, CAT-8, and Fiber throughout the home. Truth be told almost every area, floor, zone had a ethernet drop when I built my custom home more than 13 years ago.
I honestly thought all of the preplanning and wiring I'd done was enough.
What changed all of that was obviously more IoT, and network related hardware not present in the past. As noted early on the biggest change came from wanting to centralize and power as many things I could via POE. So, that of course required multiple pulls of wire to the same location where there were only 2-4.
Now, almost every location has a minim of six keystone ports to twelve.
As it relates to your question about powering multiple devices in my personal environment there is enough ports to attach all of the current and future devices. I'm not a real fan of adding multiple pieces of hardware on a single line but if something only required power a simple Y splitter in whatever variety could be used. Given, the smallest switches are POE+ 500 watts there's no problem powering any device in the home.
Your questions about VLAN's . . .
Every isolated network uses all of the industry best practices of VLAN's, Port isolation, DHCP Snooping, Radius, most commonly known as port security & port based authentication (AAA). The problem is and reason for the four independent and isolated networks is there are just some IoT devices that don't operate well or at all when any of the above is enforced. As such, the easiest thing to do (even more time consuming and expansive) was to prop up another isolated network to handle only those chatty and IoT devices which run on a completely separate (ISP) Internet connection.
This area only had a dual port and has since been upgraded with twelve. Four dedicated optical fiber lines were run to this location.
In the garage and some other locations I installed dedicated POE+ (at) & POE++ (bt) ports for a big project in the works. The red ports are the dedicated POE++ (bt) ports for easy identification and use.
In the living room there was only a dual jack and have since installed two 6 port jacks. I don't ever foresee needing twelve POE jacks but there's plenty of space to install more if need be!What's missing in this photo is the heat shrink alphanumeric ID labels on each cable. As I ran out and had to go through the chore of re-terminating a perfectly working panel just to insure they were all labeled correctly!
I don't remember what I did or ate yesterday. For sure won't know what a blue, yellow, green, black, purple cable does later!
I tried to get an idle reading on my ICX 6430-24P, turns out the load calculation on my UPS is quite off as it was showing around 100w idle. Then, i found out that a Kill-A-Watt really hates modified sine wave. I plugged it into my UPS running on battery and it got super hot, smelled burnt, thought it was gonna let out the white smoke. It read 20w with nothing plugged in (PoE on/off without a load had no affect). After it cooled off, i tried it straight to shore power, it was reading 18-19w idle with a PF of 0.65 which seems quite low. May have fried my K-a-W!
Last edited:
Interesting about the kill-a-watt on MSW. I've run mine on MSW in the past and it not only didn't smoke but gave fairly accurate RMS readings for voltage, current, and power. For frequency, not so good. The 0.65 pf is in the realm of possibility. I've seen several switcher supplies with a pf around 0.7, and 0.65 isn't far from that.
This time fifteen years ago a pure sine wave UPS was extremely expensive in whatever size to purchase. Now, in 2021 almost every 1st ~ 3rd tier vendor offers a reasonable priced entry level pure sine wave UPS. Your personal experience would affirm the importance of using a pure sine wave vs modified sine wave (PWM) in this critical network system.
Nobody has to buy the largest UPS system on the market but one should be present to offer clean regulated power and protection to a really expensive investment already present!
If your area has reliable and clean power and don't see any sort of power issues a small 750 VA UPS is more than enough to ride out those 1-2 minute interruptions. If on the other hand your area sees frequent power issues you'll need to sit down pen to paper and determine what's reasonable to have in place.
Measure the actual power consumption from the entire system to calculate what size UPS needs to be in place. In the worst case scenario installing several smaller UPS's to meet a budget is a good route to follow too.
Regardless of what brand, make, model, of UPS always test, validate, and complete monthly / yearly power loss tests. Grab something to time how long the system is able to operate on UPS power. Validate the UPS system reacts, tracks, and sends the correct information to you / systems.
Don't count on what the UPS indicates as the available runtime. The only thing that matters is your real world testing and the timer in your hand. If the UPS says you have ten minutes but your tests shows five?!?
That's why you're validating and testing to know real world what you'll have when there's a grid down event. If the system can send email time how long it take for it to arrive on whatever system you have to receive the same. Ideally the system would be able to send email & sms but there are tons of options to mimic sms / push notifications.
If you program to gracefully shut down your system test and validate it does. If the UPS had the ability to load shed and disconnect less important loads on an outlet rinse & repeat the above process. Some of the most important aspects of testing and validation of any UPS is how does it operate and perform once power is restored!!
Many less expensive and those that use different topology to operate will either run fine once power is restored vs others will not.More specifically in a real world test you're literally draining the battery system. Some cheaper units will turn back on but will NOT restore power to the outlets until the battery is at some charged level.
You'll need to test and validate how your specific UPS will operate in this very common situation.
Regardless of all of the above batteries are consumables as such need to be replaced on a timed interval. This means you need to set aside finances to pay for this ongoing maintenance cost. The average life of a 1st tier battery cell is 3-8 years depending upon the quality of the same. Batteries are like sneakers and wear out even when they are sitting idle for their entire service life.
Lastly, don't count on the UPS's monthly load test to validate the battery is fine. Only a real world load test will validate how the system and cells are performing. One final point is you must find a balance as to how you conduct your monthly, yearly load tests as every cell has a limited cycle life and DOD (Depth of Discharge).
Meaning if the cell is rated for 400 cycles at 80% discharge you don't want to shorten the service life by load testing the UPS every month to 80%! As the battery will literally only offer you a year and half of operational run time!
