My take on the HP server power supply. - Page 21 - RC Groups
Shop our Airplanes Products Drone Products Sales
Thread Tools
Jul 30, 2011, 08:40 AM
Registered User
The top right pot in your last image adjust voltage.
I guess the bottom left one is current limit, but i did not tried to play with that.
If you want higher voltage, it is easier to look here:
Ribbon pin 8 and 470 Ohm resistor to +5VDC give me about 13.3V.
I found my iCharger seems to be best effective if input voltage is near battery voltage.
In my case, i charge 6S, so 25V PS seems to be tiny bit more effective then 26.5V one.
Sign up now
to remove ads between posts
Jul 30, 2011, 11:51 PM
Registered User
Nice thread…I have a question:

PS A has normal DC ground, and PS B has the DC ground removed. I can get PS A to power on without issues. In order for PS B to turn on, connect the same pins 4, 5, and 8. Now for pins 1 and 2 on PS B. Do they connect to the DC ground on PS B or PS A?

I suppose I could try things out…but I'd rather ask first!

Jul 31, 2011, 10:15 AM
Pitch - Roll - Y'all
Originally Posted by gxc11
Nice thread…I have a question:

. . . Now for pins 1 and 2 on PS B. Do they connect to the DC ground on PS B or PS A? . . .


PS B's pins 1 and 2 go to PS B's DC ground. The only difference is that PS B, with the isolated DC ground, has fewer DC ground "locations/options" for you to attach pins 1 and 2 because of the "isolation" mods.

I attached mine directly to the DC negative post via a ring terminal.

On PS A, you could attach pins 1 and 2 to the negative DC post, the screws holding the board to the case, or solder to the pair of intact jumpers on the board of PS A.
Aug 01, 2011, 03:45 PM
Registered User
papa_lazerous's Avatar
Going to chime in and try not to hijack here, I've just got my 2 DPS-600PB power supplies, and have succesfully removed the DC grounding on 1 so I can run in series. I know to power up I need to short 3 pins together on the outside of the hot swap connector.

But what I really want to do is remove teh hot swap board completely and take my 12V supply directly from the posts on the main board inside. The problem is that that I am not sure what to do with the 2 wires that goto the small hot swap board on mine one is purple and the other is blue... And secondly do I just need to short pins on the ribbon cable to get it powered up if so can someone post a photo? I have tried going through lots of posts but the threads have gotten so long its impossible to find the relevant info

Aug 02, 2011, 03:44 AM
Registered User
loudbikes's Avatar
Papa lazerous, dps-600pb version: connect wires 1 and 2 to ground and wires 4,5,8 together and cover with heat shrink. Purple and Blue wires are labeled on the main board. One is the +5 and the other should be the -12. I just cut them short as I didn't need them. - Tim
Aug 19, 2011, 02:48 AM
Registered User

12v x5 in series x2 in parallel for 60V, 5kw Battery Charger?

G'Day guys,

So I have now read all 21 pages of this thread so far, and it sounds like you dudes are on the ball.
I know this application is a little different from what you are using the chargers for, but the principles should be the same and I am hoping for a little direction.
Here's what I'm thinking...

I have 4 x 12v deep cycle batteries set up in series (=48v), and I have 3 sets of these in parallel for a total of 12 batteries at 48v = a total of around 18,000w storage, but intend to only use no more than around 9,000w per cycle (best not to discharge below 50% per cycle I believe).
I have a 5.5KVA generator to re-charge the bulk back into these before the sun comes up, then a few solar panels on an MPPT charge controller that should finish the last long, slow phase of the recharge during the day.

The problem is the cheap 48v DC chargers are usually well under 1,000 watts so I would be looking at running the generator for around 10 hours+ to recharge, and it doesn't use a proportionate amount of fuel depending on load. By this I mean if I ran the generator for 2 hours with a 5,000w load, it doesn't use 5 x as much fuel as if I had it running for 2 hours with only a 1,000w load. It is much much more economical (and faster and hence better for the life of the generator overall, and the environment) to get the job done quick instead of running it for half a day with very little load. A 48v DC charger to do big watts is up to $6,000 for a reasonable one.

So I’ve been thinking... I can probably get access to 12 x "Dell DPS-500CB A Rev 7" power supplies (12v~41A each) from 6 old servers here at work.
If I expanded your 24v (2 in series) approach with 5 in series for 60v (which is the ideal voltage for bulk charging AGM batteries), and I created 2 sets of these charging in parallel, I would have created a 10 PSU, 60v, 82A (4,920w) beast that could in theory do the business in around 2 hours thereabouts, and have 2 power supplies spare for backup replacements.

So what are the issues?
I see spikes in an oscilloscope pic in your thread, will these become just too noisy with 10 PSU's ?
What would happen if 1 or more failed in one array of 5, would the other more powerful array create a disturbance in the force?
Can they even handle 60v pushing through each of them?
Is 10 enough to reach critical mass and the whole Frankenstein (it’s Alive!!!!) charger just end in a fireworks show and melt a hole through the earth's crust, providing great amusement for the missus as we stand from afar switching the array on with a broomstick?

I will pre-disclaim my mad-scientist actions now, and declare that I will create such a monster purely of my own free will, accepting all responsibility for the consequences to myself and any ripples in the time-space continuum.

Legitimate concerns and advice please.
Last edited by webdude; Aug 19, 2011 at 02:55 AM. Reason: title now more correct.
Aug 19, 2011, 11:49 AM
Registered User
Well, there is one big problem, these server PS do not have Constant Current limiting, meaning they will shut down if you overload them.
Which you will, if you just connect them directly to batteries.

In theory, PS without CC could be used, if you are able to adjust the output voltage and measure current.
You need to raise the voltage during charge and keep the current under shutdown threshold.

Most industrial 1kW or stronger PS does not have CC limiting.

And if it would work, i would be very afraid of the voltage ripple and what would happen if one unit fails, especially at this number of PS.

One of mine 2x400W PS in series failed (burned), it was the grounded unit, the ungrounded unit survived.

Some options:

I know, not enough power, but Pb should be charged at 1/10C - what is the Ah value of the pack, 400Ah/48V?
If so, you should charge it at 40A, roughly 2kW.
Aug 19, 2011, 12:03 PM
Registered User
akschu's Avatar

What an interesting problem and solution you propose.

First of all, lead acid batteries are charged using constant current, then constant voltage, then a float.

What you are proposing is getting rid of the constant current part and just attaching a constant voltage. With 4 of these power supplies, it will probably average around 50.5v ish so how fast they charge the batteries will completely depend on bias between the power supplies and the batteries and how much current it induces.

Given that the constant voltage charge on these is around 14.7v per cell or 58.8v, I suspect that the current flow won't be that much. In order to calculate I would need to know the internal resistance of each battery.

All that said, if you connect these power supplies in series to get 48v then they are going to be pretty noisy. Two of them is pretty noisy, but our chargers are dc<->dc converters so they do a pretty good job filtering the noise, but in your case, you might see significant voltage jumps.

I think the best way to see if this will work is to get one power supply, find the voltage adjustment pot and crank it as high as it will go, then connect it to a single battery with a shunt or current sensing meter, and see what the current draw is. If the draw is enough to charge your batteries in the time frame you want then great, if not, your not out much.

Aug 20, 2011, 03:37 PM
Registered User
I am almost sure the PS will shut down.
These switching PS hold voltage too good.

I charge now 100Ah 12V LiFeYPO4 by using Maas regulated PS with CC limiting.
I set it at 14.5V/50A and if i connect it to the battery, the voltage drops to about 13.6V, holding 50A current.

Some industrial MeanWell power supplies does have current limiting instead of shutdown, but i was warned by user who tried it, that it is overload protection method, so should not be used this way, the PS does not "like it".

MeanWell does make nice Pb charger, but i guess 1kW is not enough:

It is no problem to get a cheap 12-15V adjustable PS with current limiting, with power around 350W.
It is difficult to get a unit with more power and much more difficult and expensive to find something for higher voltages, 24-48V.
The Mastech units were used in the past for direct charging LiFe cells, because there were not any powerful chargers on the market.
Aug 22, 2011, 12:00 PM
Mike Freas's Avatar
Looking to use one of your kits but I'm not sure what supply from feathermerchant I should buy? Would it be best just to get them from ebay?
Aug 23, 2011, 10:58 AM
Mike Freas's Avatar
Will there be any more full kits made? I'd like one.
Aug 23, 2011, 04:05 PM
Registered User
I'm interested in a full kit as well....
Aug 23, 2011, 07:23 PM
Registered User

RE: 12v x5 in series x2 in parallel for 60V, 5kw Battery Charger?

Thanks to Jointer & akschu for your replies.

I am not worried about the noise from 10 PSU's, the generator will be singing the main tune, the symphony of 10 fans will just add a chorus line.

Also AGM deep cycle batteries of all the Lead Acid types are apparently the best at taking a wallop charge real fast, I have heard before that you can pretty much recharge the bulk of them in under 2 hours without damage/complaint hence why I was aiming for using the full capacity of the generator.

I guess I am more concerned with the idea that I should be fixing current rather than voltage for the main bulk charging phase?
This could be where this experiment could come undone.

I am not fully over this (cheaper / more fun) option yet, but I will be looking more to see if I can snag a cheap second hand forklift charger perhaps?

Thanks again for your help and ideas, and congratulations on a very informative discussion.
Aug 24, 2011, 04:27 AM
Registered User
I am afraid the noise we mean is voltage "ripple" and not acoustical noise...
Not sure how fast AGM could be charged, but it should be written somewhere on them.
Charging Pb requires phase of constant current and then constant voltage.
Not possible with these server power supplies (unless someone modify them for constant current, which is probably not easy or possible).
They are good only for controlled load like chargers or... servers.
Aug 25, 2011, 10:58 AM
Mike Freas's Avatar
Well it's been a few days without any word on a new batch of kits. If anyone has a full kit they are willing to part with or even just the instructional DVD shoot me a PM. When I get home from Germany in a month I'm going to build one of these to take full advantage of my Duo. Thanks.

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion COnverting Server Power Supply - EMERSON pluto_26 Power Systems 17 Nov 02, 2017 11:34 PM
Question Hp DPS 800gb a power supply Nocept1 Batteries and Chargers 277 Oct 17, 2017 02:42 PM
Discussion Getting a Compaq HP DL580 Proliant Server Power Supply to Work earlwb Batteries and Chargers 556 Aug 09, 2017 10:04 AM
Help! HP Server Power Supply conversion memory Batteries and Chargers 4 Feb 14, 2016 02:47 AM
Question Can I Use This Compaq Server Power Supply? LowFlier07 Batteries and Chargers 2 Oct 28, 2010 05:23 PM