There's two wires to each Deans tab instead of one so from memory I just soldered one to each side. The only place you have to watch is the inside of the positive tab where the wire will come close to the negative one. It's fairly easy to push it into a flat shape when the solder is molten with the iron. If you have trouble soldering Deans normally you might find it hard though - you need to be quick so you don't overheat the plastic. Also it's easier if you support the first wire so it doesn't move when you add the second.

An alternative would be to just join the two wires on to a shorter single length, then add the plugs.

On my 24V 100 supply which has 4 wires per side, I use a mechanical connection with an electrical terminal block - much easier. You can then add whatever connectors you want and change them quickly. See picture.

There's no standard in this game, we all do it our own way. I recommend connecting all the blades together to share the current and minimise the losses but that's just me being conservative.

Originally Posted by Sit-N-Fly

Hey jj604,
I was looking at your original conversion of the HP supply and had a question. How did you tie those heavy gauge output wires from the 12V blades into your Deans connectors? Looks like you did something under the heatshrink but it sure is a neat job. And is it standard operating procedure for these conversions to tie all the +12V blades together and all the grounds together to get the most capacity? Or do you just figure out how many chargers you want to power and figure your outputs from there? Thanks for any replies.

I'm having some issues with my Dell NPS-700AB.

When I first tested it, everything worked fine and got 12.15V with no load.

So I opened it up and installed a 60mm fan grill internally (between the fan and the case itself). Mounts up fine, but now it no longer provides 12V on the rails.

If I short pins c1r3 and c1r4, the fan comes on. As soon as I bridge it to c6r3, I hear a switching noise, the amber light with an exclamation sign comes on, and the voltage doesn't ramp up.

Any suggestions before I go back in to remove the fan grill?

Originally Posted by liquoredonlife

I'm having some issues with my Dell NPS-700AB.

When I first tested it, everything worked fine and got 12.15V with no load.

So I opened it up and installed a 60mm fan grill internally (between the fan and the case itself). Mounts up fine, but now it no longer provides 12V on the rails.

If I short pins c1r3 and c1r4, the fan comes on. As soon as I bridge it to c6r3, I hear a switching noise, the amber light with an exclamation sign comes on, and the voltage doesn't ramp up.

Any suggestions before I go back in to remove the fan grill?

Check that you didn't unplug one of the cables that go to the secondary board.

Originally Posted by tjinguy

Check that you didn't unplug one of the cables that go to the secondary board.

Oh good call, I see it now. Nearer to the A/C plug cable side. As soon as you open the clam shell case it pulls apart. Now to figure out if I can put it back in without cutting the zip tie.

Thanks!

Edit: that fixed it!

I've been on vacation and had some time on my hands, so I've decided to improve upon a great server power supply. The DL580 G3 power supply is a great find as some in this forum can attest to. In it's stock condition, I've been able to squeeze 100A out of this thing even at 120v. I know the internal fans are loud. But when your getting something good, why does it have to be silent. Anyways, the two things that bothered me about this supply was the lack of voltage regulation and voltage control. So I've upped the voltage and tweaked the regulation. In its current incarnation, it will deliver 13.2v under no load. And 13.2v under full load! If for any reason the 13.2v setting doesn't work, then just back it off to 13.1v.
These voltage measurements are taken at the PSU. This modification can easily be done externally with the addition of two resistors. The process of finding these values and the right pins was not as easy however.

I've included a picture of what needs to be done, and a chart that shows voltage regulation at the specified voltages. The pins that need to be connected are identified as r1c3(row 1 column 3), r2c5, and any one of the 3.3v pins. The 3.3v output is not changed by this modification, which is why I chose to work with them. The resistance value between r1c3 and r2c5 needs to be 6.8k ohms. Any less will make the PSU output voltage increase as more current is drawn! That may be good for some type of battery charging system but it is not what we need. Also it may cause the output voltage to rise into the overvoltage Protection region as specified in the regulation chart. Any higher than 6.8k ohms and the PSU will drop out of Regulation and drop its voltage. I chose resistor values that were common or easy to obtain, and close enough to the needed values.



Thanks jj604 for starting this thread. Hope you don't mind me using and modifying your pics and pdf for this post.

You kidding? This is great work - thanks!

Now if only you could find the time to figure out which pin outputs the variable voltage with load and which one controls the fan speed.....

John

Originally Posted by xandrios

Thanks jj604 for starting this thread. Hope you don't mind me using and modifying your pics and pdf for this post.

Variable voltage with load pin and voltage regulation with output pin are the same. You just gotta be able to use them for what they were not intended for.

I thought as much.

Now the one we really need is the pin that controls the fan speed. It would be easy enough to set up a manual fan control if we knew that one.

John


PS I put a link to the Dell ASTN 7000245 Poweredge 6650 PS in the sticky thread.

Originally Posted by xandrios

Variable voltage with load pin and voltage regulation with output pin are the same. You just gotta be able to use them for what they were not intended for.

Cool. Still several more days on vacation. I'll be up to something...

With regards to the NPS-700AB, is bridging C2R3 to ground the lowest fan speed possible as shown in this picture-



Or will following the digram xandrios provided be able to reduce it further and allow it to ramp up to the medium speed indicated above?



Apologies, I'm not versed in reading circuit diagrams, but is red and green bridged with 1k ohm, with another 1k between green and ground?

Shorting to ground will achieve the lowest speed possible. The second 1k or R2 is between red and ground. You can adjust the value of the second 1k to meet your needs. Making it less than 1k will make it speed up less to somewhere between med and max. Just give it enough to cool the PSU down...

Quite a few people will be waiting on your efforts.

I still think the original 100A supply that started this thread is the best combination of power and form factor ...if only we could quieten it down without having to do internal mods.

John

Originally Posted by xandrios

Cool. Still several more days on vacation. I'll be up to something...

Originally Posted by KiloOne

I still have to open them up though since I want to isolate the -ve output for series use. The 7000814-0000 can be modified for this fairly easily.

Could you please tell us how?
I still think temp sensing is the better way of controlling the fan speed as there are more factors then just the current draw involved. Besides, I will pick up 10 sensors at the post office today .

Fred

Fred, I agree. Another option is to put a wattmeter in line and control speed based on the power drawn.

However. It would be a really nice "good enough for most of us" solution if someone can find a way to just control the fan speed using the external connections WITHOUT having to disassemble the PS at all. Very simple and easy to do for folks who aren't comfortable with internal mods.

John

Originally Posted by flarssen

Could you please tell us how?
I still think temp sensing is the better way of controlling the fan speed as there are more factors then just the current draw involved. Besides, I will pick up 10 sensors at the post office today .

Fred

Fred,

I will start a how to thread by taking some pics as I do the next 11 supplies.

Can't get to that for a few days and I may change this procedure a little when I do them.

Basically you open it up, remove the two screws by the output block that ground the output board to case, raise the board up 1/8 inch on that end, slide a thin metal finger under pcb board holes, drill pcb board out to diameter slightly less than screw head dia (metal finger keeps drill from drilling out threaded holes in case), slide a (about .010 inch thick) plastic isolator shim with a hole in it under each edge of board. I used some of the black plastic that was lining the case. Make a plastic washer to go under the head of the two screws and put them back into the enlarged pcb holes (better yet use some 'hat' shaped isolating washers).

Use at your own risk

Dale