Part 2: Portable Lighting System for Beach Volleyball



This is part 2 in my ongoing DIY sports lighting project and details assembling a 24V DC power supply for the battery charger.  Part 1 of the project can be found here.

Most of the higher power balance chargers for Lithium Polymer (Lipo) batteries require a separate DC power supply.  For our project, we will be using the Revolectrix Power Powerlab 8x2 which can charge at 40 amps per channel.  Feeding this beast requires either an expensive (~$300+) 2000 watt power supply, or a bit of DIY modification to some widely available computer server power supplies to adapt them to the task.  Being as the overall lighting system cost was already running out of control, I decided to adapt powerful and reliable Hewlett Packard DPS-1200FB server power supplies which are readily available used from $20-$30 each.


**A bit of warning before continuing forward.  The descriptions contained in this post are for educational purposes only.  Be aware that the circuits inside of these power supplies operate at LETHAL voltages, and should only be handled by qualified technicians observing proper safety protocols.  Do not attempt this project if you do not have a thorough understanding of the circuits involved.  There are off-the-shelf products that can safely accomplish the task at hand.**

That being said, I selected the HP DPS-1200FB power supply based on information obtained from numerous RC hobby forums.  There are many computer computer server supplies suitable for modification and use with Lipo battery chargers, and I am sure over time, availability and serviceability of this specific model will change, but at the time of writing, the DPS-1200FB seems pretty optimal.  These HP power supplies run at 12V but the Revolectrix Powerlab 8 does not charge at full output unless it has a 24V supply, so we will configure two DPS-1200FB power supplies and wire them in series to achieve our desired 24V output.  Because the Powerlab 8x2 has 2 discrete channels, we need to build a power supply for each channel.  This is beneficial because at full charge, a single household outlet will likely blow a fuse, so splitting the task between 2 outlets on different circuits will allow this setup to operate without special 220V circuits in the house.


The first step is to acquire some DPS-1200FB power supplies.  I purchased mine on Ebay.





In order to safely wire 2 DPS-1200FB modules in series, we will need to isolate the circuit board of one unit from the outer case.  A quick test reveals that in standard configuration, the negative output terminal is directly connected to the outer case.


We will need to isolate the internal PCB of one of the power supplies from the outer chassis in order to connect the 2 in series to produce 24V.

First, remove the 4 screws that secure the top cover.



With the screws removed we can open the top cover and peel back the insulating membrane.



Next, remove the 4 self tapping screws that secure the cooling fan.



Next, disconnect the IEC safety ground wire from the chassis.



Wiggle and lift the IEC module from the chassis.


With the IEC out of the way, one of the 3 screws that attach the PCB to the chassis is revealed.



Remove the 3 PCB mounting screws.





Next, push the power indicator LED in to detach it from the metal housing.


At this point, all of the internal PCB unit and attached fan and IEC can be removed from the chassis.




Remove the insulating membrane from the metal chassis.


There are a variety of ways to isolate the PCB from the chassis.  The most direct way I could think of is to use isolating shoulder washers and retain the original M3 screws and standoffs.  In order to do this, we will need the following parts:

1x 12mm M3 screw
2x 8mm M3 screw
3x M3 flat washers


We will use the shoulder washers to electrically isolate the PCB from the chassis.  Install them on the top and bottom of the PCB.





Some of these will have a tendency to fall out when handling the PCB so be cautious when re-assembling.

The shoulder washers on the bottom of the PCB will raise the entire assembly and some adjustments will need to be made.  A quick check with a dial caliper shows the PCB will be raised .040 inches.


We will have to grind down the metal standoffs on the chassis by .040 inches to fit the board back in correctly.  The standoffs are .125 inches tall.



So. . . 

.125" - .040" =  .085"  

Which is the new standoff height we will aim for.  I decided to grind down the standoffs from .125 inches to .085 inches with a Dremel and a cutting stone attachment.




After cutting down the standoffs, I tested the new M3 screws that I purchased to make sure everything will thread in as before.  Sometimes grinding will create burrs that inhibit the screws from properly threading in.


With the standoffs set to the proper height, re-install the plastic insulating membrane.


Next, place the PCB assembly back into the chassis.  Because the bottom shoulder washers had a tendency to fall out when maneuvering the PCB, I decided to hold the entire PCB assembly upside down and then position the case on top of it so the shoulder washer would remain in place.



Put flat washers on the M3 screws.


And secure the PCB assembly back into position.





Next, re-install the power indicator LED.



Put the fan and IEC back in position.



Re-attach the chassis ground to the case.


Install the 4 self-tapping screws to the cooling fan.


At this point, we can check and see if the PCB has been isolated from the chassis.  If the shoulder washers have been installed properly, there should be no continuity between the negative output on the PCB and the chassis.


If that checks out, re-install the top cover and double check to make sure we still have no continuity between the negative output and the chassis.


With the PCB isolated, the next step is to "trick" the power supply into turning on.  I use a 1W 750 ohm resistor.


We will need to bridge pin 33 and pin 36 on the PCB with the 750R resistor.


I found it easy to use a clip to secure the resistor on one side while soldering the other end.



With one side soldered in, remove the clip and solder the remaining end.


Trim off the excess with a wire cutter.


With the jumper in place, I crossed my fingers and plugged in the power supply.  No smoke!  Humans win!  


And we have 12.34V output which is great.  The power supply does have an internal trim screw for fine tuning the output voltage, but 12.34V will work just fine.

With the power supply confirmed working, I install a banana plug/screw terminal  jack to the output lugs of the PCB.


First, chop the ends off of the banana plug unit.


Next, set the soldering iron to a super high setting since the banana plug unit and the PCB lugs are pretty large and will dissipate heat very quickly.  Then, secure the banana plug on one side with a clip and solder the other side.


Then, remove the clip and solder the other side.


For a little added security, I used a couple of the unused washers from the banana plug unit to secure the back side of the PCB.



We will be connecting 2x power supplies in series.  One of them will need to be isolated as described above.  The other one will only need the 750 ohm jumper applied to trick it into turning on and does not need to be opened up and isolated.


I used some velcro to quickly bundle the 2 power supplies together.  To make the series connection, we will need a short wire lead.  I used a piece of 10 gauge wire.


Connect the 2 power supplies in series.  Note, the top unit is our isolated supply and the bottom unit is not.



And our power supply is completed.

I power the unit up and confirm that we have 24.65V output which will work great for feeding one channel of our powerlab 8x2 charger.  I plan to build another identical supply for the 2nd channel of the powerlab 8x2, but it should be noted if we are mindful of our charge settings and the capacity of the household AC circuit we are using, we can use this single 24V power supply to run both channels of the powerlab 8v2.




10 comments:

  1. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
  2. Hi, very good work !!!

    I've a question, if I want two PSU in parallel, so 12V at 200A... Is the same connection/isolation as in series ?

    If I need 24v at 200A... do I use 3PSU isolated and 1 PSU original ?

    Thanks

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
  3. The resistor is bridging pins 33 and 36, your text mentions pin 35. Awesome procedure!

    ReplyDelete
    Replies
    1. Thanks for catching that error. . . it has been corrected :)

      Delete
  4. Hello silsil, it's a very nice idea to consider parallel DPS1200 SMPS units, BUT.....

    Unlike when SMPS's are configured in series, a parallel configuration needs to have a balancing circuit so that both supplies are equally sharing the load. If this isn't done, one supply may be working harder than the other. This isn't desirable.

    I have investigated this myself and there are circuits to do it, but you need to know where a feedback loop can be applied to the DPS1200 board to control it's output.

    If it can be done, putting two sets in series wouldn't be a problem.

    Some Meanwell SMPS's have this facility to enable a parallel configuration.

    These American made interface boards are a great idea, one of his customers used them for his charger set up.

    http://www.buttsimpleideas.com/

    Cheers.

    ReplyDelete
  5. Thanks for sharing.

    Plaese visit my blogs:

    http://nklpunya.blogspot.com

    http://edmet-nklpunya.blogspot.com

    Don't mind just spend a few seconds
    by clicking to the link below to give a
    vote for his determination ...

    Please vote by 31/08/2016

    TQVM.

    http://woobox.com/vikq2d/gallery/HyQpixBqKhU

    ReplyDelete
  6. Question #1) You stated the Jumper wire from positive to third bottom terminal is done on the non floating unit however to clarify are you suggesting it isn't necessary on the floating power supply to make it active? If so, why not?
    Question #2) If running 3 in series to achieve 36v it would be obvious to float 2 of the three PSU's. Would the Center and Last PSU be floated? Or does it not matter which order they are floated?
    *Results of my 24v I had jumpered both Units and had no issues. However I have attempted to make a 36v and while it started up and was working fantastic while under load for 15 minutes charging The capacitors inside the center unit started bursting one after another shutting down the center PSU and rendering it only operable as 24v after the failure. Any suggestion what may have caused this? Thank you.

    ReplyDelete
  7. Hola, para una PS750w, vale la misma resistencia entre pines 33 y 36? Cual seria mejor si esta no vale?

    ReplyDelete