In principle there are the ground issue and an inrush issue. I don’t want to speculate more about the grounding than what others have already done, but I’m speculating about inrush. Both the UPS and the data center have likely very high inrush currents in initial charging banks of capacitors, or in the case of the UPS, a battery bank. You can fall a victim of the short, 20 to 50 millisecond, up to 100 A inrush that trips protections, maybe in the UPS. Just a thought - not knowing about the internals of any of your items in the installation.
I have thought of this as a possibility,, and it might still be.... and that was why I did the resistant load of the pool heater,, to smoothen it out,, it is just not a good long therms solution. :-/
Morten, how do you supply power to your home data center? Are both of your racks connected to a single breaker in your home’s breaker box, or are they connected to multiple breakers? If they’re connected to a single breaker, what is the amperage capacity of that breaker? If you were to turn on all of your servers, would this capacity be sufficient? Can you one day prepare a video on the topic of supplying power to a home data center?
I paused at 3:15, you made a huge mistake there, when you use an oscilloscope to measure line voltage you should never connect the oscilloscope ground lead to anything that is not ground or earth. If you connect the cable in reverse you would short circuit the main power and potentially damage your probe and oscilloscope
I should add that if the power station is only running on solar than it should be floating and shouldn't be an issue but in the last video you had it connected to both main and solar so it's very dangerous
Hello Morton. The hotswap psu modules have a 12 volt rail. You could try making an adapter out of a defective power supply and running it directly from 12 volts. A DC DC UPS in between and you're done. Then you don't need to convert DC-AC up to AC-DC down.
What does the UPS log say? It may have been an undervoltage, or an overvoltage. You can change the upper and lower limits the UPS will accept before it starts correcting. You can also turn down the sensitivity of the UPS. My APC UPS will take just about any power at all and provide clean power to the server.
I commented in your las ups video about the allpowers giving an isolated output this is still your issue here you took a big risk connecting your oscilloscope like that to your mains as the ground or neutral or minus whatever you call it in Denmark on your oscilloscope is connected to directly to the earth pin on oscilloscope mains input so you could have shorted mains to earth through your oscilloscope and destroyed your scope or at least poped the breaker you only survived it because like I said previously your alpowers is isolated output and since one of your terminals was connected to ground through the scope your ups worked. try powering up ups from scratch without scope connected. " DON'T EVER CONNECT YOUR SCOPE TO ANY MAIN WITHOUT AN ISOLATED PROBE OR BATTERY POWERED SCOPE"
Grounding going into the UPS is not grounding the unit itself as power going out is usually isolated so you would need to find a grounding point on the unit itself or the power cable going INTO the Allpowers...
@ 15:43, Isn't this the UPS's periodic self-test process, which briefly switches to batteries to test their condition and availability before switching back to the mains?
@@MyPlayHouse you could use an earthing plug on a regular house socket to connect the ups chassis (or sometimes they have a grounding screw at the back) with ground, this way your UPS and servers are still properly grounded to your house while power is coming from the power station
@@MyPlayHouse assuming all sockets have continuity with each other the most simple way to bond neutral and Earth is to use a male plug with a wire from the left pin which is neutral to the center pin which is Earth to create neutral to Earth Bond if you don't understand what exactly i say just search neutral to Earth Bond plug or Generator Neutral Bonding Plug and you going to find a solution btw after doing that be careful how measuring with the output with the oscilloscope I hope I helped
Morten, I believe the Eaton UPSes we use at work have a sensitivity button that changes what the UPS tolerates. Does your UPS have a setting like this?
100W of solar-power. Nice. Yesterday we got a little more at the company i work for: 380kW solar-power from big panel-arrays at the roofs. And additional 100kW from our 5MWh battery-container with 1MW inverter. While this test-run the whole company took about 700kW power 😎
you should do more testing and perhaps read up and test other models to get a system that will work optimally - almost there, i think you want more watts coming in but the power is almost less of an issue than ventilation and humidity control - you may want redundant and slightly overlapping systems #failover
i caught the monitor on the first reset, and it appears that the sine wave, right prior to the ups cutting off, appears to be faster before the power cut out, maybe even double the freq, then after resetting the sine wave appears to return to the normal hz,
Is the wave form monitor youre using not considered a load by the allpower? That could be keeping it's outputs "live" when the ups complains and shuts off the input,then tries to accept it again? That would be my thought?
The oscilloscope only uses a few watts,, not enough to be a real load,, and I was powering it from the grid,, as I would not be able to see stuff if the R2500 was powering on and off the AC..
Maybe the UPS is too picky about the input power. Can you change the sensitivity in the UPS Software? On my Rack-UPS I can set how "clean" the incoming power has to be to be accepted. Just a suggestion, I hope it works!
In principle there are the ground issue and an inrush issue. I don’t want to speculate more about the grounding than what others have already done, but I’m speculating about inrush. Both the UPS and the data center have likely very high inrush currents in initial charging banks of capacitors, or in the case of the UPS, a battery bank. You can fall a victim of the short, 20 to 50 millisecond, up to 100 A inrush that trips protections, maybe in the UPS. Just a thought - not knowing about the internals of any of your items in the installation.
I have thought of this as a possibility,, and it might still be.... and that was why I did the resistant load of the pool heater,, to smoothen it out,, it is just not a good long therms solution. :-/
Morten, how do you supply power to your home data center? Are both of your racks connected to a single breaker in your home’s breaker box, or are they connected to multiple breakers? If they’re connected to a single breaker, what is the amperage capacity of that breaker? If you were to turn on all of your servers, would this capacity be sufficient? Can you one day prepare a video on the topic of supplying power to a home data center?
One braker,,, did you miss that all of the stuff I am running is only 320W to 460W ??
I paused at 3:15, you made a huge mistake there, when you use an oscilloscope to measure line voltage you should never connect the oscilloscope ground lead to anything that is not ground or earth. If you connect the cable in reverse you would short circuit the main power and potentially damage your probe and oscilloscope
I should add that if the power station is only running on solar than it should be floating and shouldn't be an issue but in the last video you had it connected to both main and solar so it's very dangerous
Guess I was lucky...
Hello Morton. The hotswap psu modules have a 12 volt rail. You could try making an adapter out of a defective power supply and running it directly from 12 volts. A DC DC UPS in between and you're done. Then you don't need to convert DC-AC up to AC-DC down.
12volt server :-) never thought of that,, I do think the server will refuse to start with out all the control signals from the PSU.
What does the UPS log say? It may have been an undervoltage, or an overvoltage. You can change the upper and lower limits the UPS will accept before it starts correcting. You can also turn down the sensitivity of the UPS.
My APC UPS will take just about any power at all and provide clean power to the server.
Someone else suggested me to look there,, and it(the log) was absolutely useless :-(
I commented in your las ups video about the allpowers giving an isolated output this is still your issue here you took a big risk connecting your oscilloscope like that to your mains as the ground or neutral or minus whatever you call it in Denmark on your oscilloscope is connected to directly to the earth pin on oscilloscope mains input so you could have shorted mains to earth through your oscilloscope and destroyed your scope or at least poped the breaker you only survived it because like I said previously your alpowers is isolated output and since one of your terminals was connected to ground through the scope your ups worked. try powering up ups from scratch without scope connected.
" DON'T EVER CONNECT YOUR SCOPE TO ANY MAIN WITHOUT AN ISOLATED PROBE OR BATTERY POWERED SCOPE"
Ahh it was not total luck,, well I guess I could have put in the plug upside down,,and that could have been bad...
Grounding going into the UPS is not grounding the unit itself as power going out is usually isolated so you would need to find a grounding point on the unit itself or the power cable going INTO the Allpowers...
Not sure if this is posable..
@ 15:43, Isn't this the UPS's periodic self-test process, which briefly switches to batteries to test their condition and availability before switching back to the mains?
It does that onces a week,, It did it the day after I did this video.
the power station has a floating ground you need a neutral to Earth Bond on the output
How would one do that..
@@MyPlayHouse you could use an earthing plug on a regular house socket to connect the ups chassis (or sometimes they have a grounding screw at the back) with ground, this way your UPS and servers are still properly grounded to your house while power is coming from the power station
@@MyPlayHouse assuming all sockets have continuity with each other the most simple way to bond neutral and Earth is to use a male plug with a wire from the left pin which is neutral to the center pin which is Earth to create neutral to Earth Bond if you don't understand what exactly i say just search neutral to Earth Bond plug or Generator Neutral Bonding Plug and you going to find a solution btw after doing that be careful how measuring with the output with the oscilloscope I hope I helped
That's why it worked with osciloskop connected. Bet if he disconnects osciloskop it wont work anymore.
@@aljosamlinaric8705 yep
Morten, I believe the Eaton UPSes we use at work have a sensitivity button that changes what the UPS tolerates. Does your UPS have a setting like this?
Humm I do not know,, would be nice to be able to lower the sensitivity,,, this is a cheap UPS :-/
100W of solar-power. Nice.
Yesterday we got a little more at the company i work for: 380kW solar-power from big panel-arrays at the roofs. And additional 100kW from our 5MWh battery-container with 1MW inverter. While this test-run the whole company took about 700kW power 😎
That is like 100x of what I have :-)
@@MyPlayHouse It's a food-company that makes fruit-juice. And the big battery and inverters are also an UPS for the whole production.
Does the UPS log give no indication as to the problem ,maybe the voltage output is unstable?
Interesting,, I did not check..!! I will have a look
Had a look,, No usable information :-/
you should do more testing and perhaps read up and test other models to get a system that will work optimally - almost there, i think you want more watts coming in but the power is almost less of an issue than ventilation and humidity control - you may want redundant and slightly overlapping systems #failover
Well,, we did get a steep closer :-)
i caught the monitor on the first reset, and it appears that the sine wave, right prior to the ups cutting off, appears to be faster before the power cut out, maybe even double the freq, then after resetting the sine wave appears to return to the normal hz,
Hi @netsspam
Thank You very much! glad you liked the video :-)
Thank you for watching! :-)
Hi @netsspam
Thank You very much! glad you liked the video :-)
Thank you for watching! :-)
Is the wave form monitor youre using not considered a load by the allpower? That could be keeping it's outputs "live" when the ups complains and shuts off the input,then tries to accept it again?
That would be my thought?
The oscilloscope only uses a few watts,, not enough to be a real load,, and I was powering it from the grid,, as I would not be able to see stuff if the R2500 was powering on and off the AC..
its german "Reine sinuswelle" = "pure sinewave"
Hi @dbmaster46446
Thank You very much! My German just was not good enough,, the oscilloscope did show the same.
Thank you for watching! :-)
@@MyPlayHouse now im curious how GOOD your German is :D
Maybe the UPS is too picky about the input power. Can you change the sensitivity in the UPS Software? On my Rack-UPS I can set how "clean" the incoming power has to be to be accepted.
Just a suggestion, I hope it works!
My thought too. According to the scope the frequency was moving between 49.8 and 50.2. it might go further out sometimes and that triggers the error.
Yes very likely,,,
What if you put a 1 or more watts 24/7 load on the allpower box?
So it keeps supplying power?
Maybe,, I think a bit more is needed,, but yes..
thank you very much for this video
Hi @skynetcybersystem3tech
Thank You very much! glad you liked the video :-)
Thank you for watching! :-)
UPS do not like that there is no ground i think. Many ev cars wont charge when there is no ground
Hi @ImbaCore
Thank You very much! glad you liked the video :-)
Thank you for watching! :-)
Proof you can work in a data center all your life and still guess muhahaha
I do not get that joke,, to analog for me..
👍cool
Thank you.
"reine Sinuswelle" is german for "pure sine wave".
Yes,, annoyingly that was one of the first comments,, and my german just was not good enough. :-)
@@MyPlayHouse Your german is much better than my danish 😉