diy solar

diy solar

AC based charging using server power supplies

I’m also confused by this. As far as I’m aware when the battery cells reach 14.1v the power draw should taper off.
I never assume a power source has a current limit that is low enough that the heat created by that current is inside the tolerance of the insulation design. The supply very well might self limit the current, but I don't know that to be true. If you know the internal resistance of the battery and the voltage different, you can calculate the approx amp draw.

In real life this is a 1000 watt power supply for a computer, meant to supply a computer of less than 1000 watt load. As soon as you go to re-apply it to a new task, you take on the responsibility of making sure it will fit properly to the task. It surely has some type of self protection. Because I don't know what that is nor how it works, I will make sure I don't overload it.

About it getting hot without proper ventalation, I have not used it yet. Since mine has adjustable output voltage, I can bring it down close to the battery voltage and I should be able to manage the current by reducing the voltage different. I am about ready to start some testing.
 
If charging at 14.1v thru a psu, on a 280ah cell in my case a 560ah bank @12v. As long as I keep an eye on the capacity meter thingy and check the bms app, I can’t see how it can go wrong. Failing that if there is a runaway the bms should kick in and shut it down.
I won’t be leaving mine unattended and running it from a genny. At the 70a max I can get out of a psu &14.1v it’s only 0.25C. The cells are rated at 0.5C charging rate.
@Mark44, what’s a retired traveller in the USA? I’m guessing your not a New Age traveller like we have in the uk? Kinda like what I am. Or do you mean traveller as in gypsy? That’s my heritage, in descend from Scottish travellers(gypsys) from that I can not retire and have no wish to.
min not having a dig in anyway, just really curious of the meaning of traveller across the world
@Pyke,

No worries, I take no offense at all. While I may have Gypsy blood, I have in fact been a "gypsy" most of my life. I have lived in at least 20 different locations around the world. I have lived and worked on 5 continents. I have traveled more than 1 million miles easily in support of my career, and or my family. I have now retired somewhat. I teach intermittently around the world, and travel for that when not locked down.

Even after "retiring" I too have no wish or capability to stop traveling. My wife and I have a fifth wheel trailer, and while we won't be "full timers", we shall be spending likely more than 50% of our time in it traveling. That includes trips "over the pond", to "the great white north" and "OZ"(but without the trailer for the last 2! ). ;)

So, back to cases. I agree with your ideas, that if you are watchful, as to charging, and you have a good shunt in line to make sure you can measure and observe charging, then with some observation all should be well.

Additionally, like you I may use a genny at some times to charge, and there is a limit we can set, simply by not filling more fuel than the calculated run time. However for me, the genny is really a back-up system.

In addition, I am adding DC to DC charging from my truck, as I have dual alternators with smart technology, so that neither one will overheat. They are 150 Amps each, so I should be able to get 70 Amps or so.

So my only relevant concern I guess is using "shore power". This is where watchfulness will be of the essence in making sure that nothing untoward happens!

In actuality, I may never need to worry much about the charger, since I will have solar (eventually) and truck based for when hauling, so I may never see a need for charging from the PSU (but better safe than sorry). Your point about charging at .25C is a good one. I would like to know if there is any downside to that charge rate. also, wouldn't you be charging at 12.5% since you have 560 AH/70Amps =8, and 100%/8 =12.5%?

BTW, we too have new age travelers here in the US. Perhaps not as prolific as in the UK, but just come to Sedona, you'll find lots of them! :giggle:
 
I never assume a power source has a current limit that is low enough that the heat created by that current is inside the tolerance of the insulation design. The supply very well might self limit the current, but I don't know that to be true. If you know the internal resistance of the battery and the voltage different, you can calculate the approx amp draw.

In real life this is a 1000 watt power supply for a computer, meant to supply a computer of less than 1000 watt load. As soon as you go to re-apply it to a new task, you take on the responsibility of making sure it will fit properly to the task. It surely has some type of self protection. Because I don't know what that is nor how it works, I will make sure I don't overload it.

About it getting hot without proper ventalation, I have not used it yet. Since mine has adjustable output voltage, I can bring it down close to the battery voltage and I should be able to manage the current by reducing the voltage different. I am about ready to start some testing.
@DThames,

Great thoughts. I do have a question or two for you.

Do you know for a fact that the draw in the server cabinets where these PSUs are used are in fact less that 1000W, or less than ~65 Amps.? If you have info on this, please share if you would.:unsure:

I agree that you as the next user, are responsible for making sure that the product is "fit for duty". My understanding in what I have read of these units, is that they are capable of roughly 65Amps continuous duty at voltage.

Yes you can calculate the current draw, but would it not be easier to use the BMS to limit either voltage or current?

Lastly, if your load (the 280 or 560 AH of battery) is more than capable of ingesting the amperage provided, why not just let it run? I suppose the PSU could smoke, but I doubt it frankly, as they are well built and likely have integrated protection circuits (as you pointed out in your post). If I go and use a HACK for the over voltage Protection circuit, then I indeed need to be aware and careful.

So how will you add protection or limits so that you don't overload it? I too would like to make sure of that for my own units.❓

I think the ventilation thing is kind of a "red herring". They are made with cooling fans. The average temperature in one of the racks they live in is likely above most ambient temperatures we would subject them to (barring desert travel perhaps).

Test way! Please let us know your results as you go. Mine have not arrived yet, but I will do the same once I have them in hand.

Thanks for your posting and information, every post gets us closer to an answer (right or wrong)!:cool:
 
Just a quick update on my choice of server PSU.
Dell Poweredge 1100W Server power supply redundant L1100A-S0 PS-2112-2D1 1Y45R

Power output is just under 90 Amps at 12 V.

I am researching the tech sheets now to see if this is continuous or intermittent.
They have shipped (2 on order), now to anticipate their arrival.
 
@DThames,

Great thoughts. I do have a question or two for you.

Do you know for a fact that the draw in the server cabinets where these PSUs are used are in fact less that 1000W, or less than ~65 Amps.? If you have info on this, please share if you would.:unsure:

I agree that you as the next user, are responsible for making sure that the product is "fit for duty". My understanding in what I have read of these units, is that they are capable of roughly 65Amps continuous duty at voltage.

Yes you can calculate the current draw, but would it not be easier to use the BMS to limit either voltage or current?

Lastly, if your load (the 280 or 560 AH of battery) is more than capable of ingesting the amperage provided, why not just let it run? I suppose the PSU could smoke, but I doubt it frankly, as they are well built and likely have integrated protection circuits (as you pointed out in your post). If I go and use a HACK for the over voltage Protection circuit, then I indeed need to be aware and careful.

So how will you add protection or limits so that you don't overload it? I too would like to make sure of that for my own units.❓

I think the ventilation thing is kind of a "red herring". They are made with cooling fans. The average temperature in one of the racks they live in is likely above most ambient temperatures we would subject them to (barring desert travel perhaps).

Test way! Please let us know your results as you go. Mine have not arrived yet, but I will do the same once I have them in hand.

Thanks for your posting and information, every post gets us closer to an answer (right or wrong)!:cool:
I just did a quick test (more telling than my opinion or assumptions) I have two of those BYD used batteries in parallel (about 200 ahr I am guessing). The battery was at 27.6v. The series power supplies were at 29.1v (lowest setting with the pot that I have). I plugged them together with a shunt power meter plugged in, between. The amps shot up to 73amps, more than the power supply is rated for at 120v 60hz. The battery voltage shot up and the amps when down, which they would as the voltage was not closer to the supply voltage. I need to discharge the battery some before I could to any more testing.

Rating sticker attached. If I really wanted the amps out of it, I could connect it to 240v AC and get 80+ amps out of it. Since my plan is to use this for generator based bulk charging (emergency grid down), I will likely make a cord for a 220-240v generator plug and use a current limit resister. My PLA printed housing might not take too much heat, so that might be my limiting factor.
 

Attachments

  • HP-server-charger-rating.jpg
    HP-server-charger-rating.jpg
    116.8 KB · Views: 7
Just a quick update on my choice of server PSU.
Dell Poweredge 1100W Server power supply redundant L1100A-S0 PS-2112-2D1 1Y45R

Power output is just under 90 Amps at 12 V.

I am researching the tech sheets now to see if this is continuous or intermittent.
They have shipped (2 on order), now to anticipate their arrival.
Smoke means intermittent :)
 
I did another quick test. I took the battery down to 26.5v and connected the 29.1v server supplies again. The amps were about 40 amps....lower than when the battery was higher voltage. I am guessing the supply could not regulate the output voltage and went into some abnormal mode. I disconnected, waited a bit, and did it again with the same results. Then I let it sit for a few seconds. The battery voltage was climbing and now the amps as well.
 
I did another quick test. I took the battery down to 26.5v and connected the 29.1v server supplies again. The amps were about 40 amps....lower than when the battery was higher voltage. I am guessing the supply could not regulate the output voltage and went into some abnormal mode. I disconnected, waited a bit, and did it again with the same results. Then I let it sit for a few seconds. The battery voltage was climbing and now the amps as well.
The PSU is trying to supply a constant voltage and will only provide the amperage demanded by the load. Are you running this through a BMS or MPPT controller that might affect it?

As to the smoke, its just the Leprechaun lighting up a blunt! (heard that somewhere....) :cool:
 
The PSU is trying to supply a constant voltage and will only provide the amperage demanded by the load. Are you running this through a BMS or MPPT controller that might affect it?

As to the smoke, its just the Leprechaun lighting up a blunt! (heard that somewhere....) :cool:
No BMS nor controller. Supply was connected to the battery. It is a voltage regulated output supply. Not sure what happens to it if it is loaded beyond its ability to regulate the voltage but I am guessing it does something different, maybe a lower max current or something.
 
No BMS nor controller. Supply was connected to the battery. It is a voltage regulated output supply. Not sure what happens to it if it is loaded beyond its ability to regulate the voltage but I am guessing it does something different, maybe a lower max current or something.
From what I know, CV supplies such as yours will manage the amperage in an effort to keep the voltage constant. So I expect that it would lower current when needed.:confused:

I'm no expert, so YMMV.
 
If you use those HP supplies, drilling holes is not recommended. The PCB's are multi-layer. If you drill a hole you might risk of a short between the + or - and one of the internal traces within. Highly recommend soldering wires (multiple to share the current)

If you get some powersupplies from an old router you can get even higher currents. For example the Cisco PWR-C45-2800ACV ( 2800W power supply for the 4500 series) provides 12V / 113A (at 240V), along with 50V/28A.

A whole different approach might be using the solar controller. If you serieconnect 2 12V supplies you're getting 24V @ 40-100+A (Depending on the powersupply you're using). If you make the PV input of the charge controller switchable to either PV or 'DC from PSU' you don't have to worry with adjusting voltages, and, more important, don't have to worry with float/absorbtion and other charge profiles. Just let the MPPT do the actual charge parameters.

2x12V in series will work fine for most MPPTs with a 12V bank, a 48-50V supply works great for 24V banks. As long as your PSU can deliver more than the MPPT it won't go into hickup or whatever it will do by overload.

Only thing you have to take care of is cooling. Some PSU's have their internal fan -> fine, nothing needed. Others don't have a fan and cooling is provided by the server/chassis where they belong. But without these you have to add cooling / airflow yourself.
Easiest is just getting a PSU with internal fans.
 
If you use those HP supplies, drilling holes is not recommended. The PCB's are multi-layer. If you drill a hole you might risk of a short between the + or - and one of the internal traces within. Highly recommend soldering wires (multiple to share the current)

If you get some powersupplies from an old router you can get even higher currents. For example the Cisco PWR-C45-2800ACV ( 2800W power supply for the 4500 series) provides 12V / 113A (at 240V), along with 50V/28A.

A whole different approach might be using the solar controller. If you serieconnect 2 12V supplies you're getting 24V @ 40-100+A (Depending on the powersupply you're using). If you make the PV input of the charge controller switchable to either PV or 'DC from PSU' you don't have to worry with adjusting voltages, and, more important, don't have to worry with float/absorbtion and other charge profiles. Just let the MPPT do the actual charge parameters.

2x12V in series will work fine for most MPPTs with a 12V bank, a 48-50V supply works great for 24V banks. As long as your PSU can deliver more than the MPPT it won't go into hickup or whatever it will do by overload.

Only thing you have to take care of is cooling. Some PSU's have their internal fan -> fine, nothing needed. Others don't have a fan and cooling is provided by the server/chassis where they belong. But without these you have to add cooling / airflow yourself.
Easiest is just getting a PSU with internal fans.
I don't think anyone is drilling the PCB.
 
Half way through doing this and hit a problem. When I connect the resistor the colts only rise to 12.70. Without the resistor there sitting at 12.64.
any ideas where I’ve went wrong?
 
26A03559-D0A7-4B6B-B511-AAD233901D3C.jpegThe problem looks like I’ve got a different chip and the over volt protection must be a different pin. I don’t know what the pin is I disconnected or which one is the over volt protection. Any idea where I can find out?
ive put the chip number into google but no results.
 
For an extra layer of protection, consider adding a neon 120/240V AC rated indicator lamp wired between the two metal PSU cases. If either has a fault which causes the metal case to be live, the neon indicator will light.
What if they both fault?

edit: think I answered my own question. 120/240 means the lamp could handle 2 hot input should they both fault. Can anyone confirm this?
 
Last edited:
Don't assume the worst, just yet. Odds are the number difference is a build date/factory code or a code for temperature rating or similar. What value resistor are you using? Try a 470 ohm and then a 1K ohm and see if the output is different.
 
I tried a few different resistors and the voltage peaks at 12.7v. That was using a 820k and 1000k resistor. I joined 2 820k together and the volts still only went up to 12.7
 
Back
Top