Thanks.
Waiting for my Dps , i will probably use a 24V 10A psu i found for 15$
Go for it, that should work well and it is unlikely you will need more then 24v DC for 98% of things you would do with it.
Heck I mostly want higher voltage just to jack around with plating and such.
giorgoskok, by any chance is your PSU like this one discussed here for a voltage mod at Banggood? Maybe you may want to give it a try, just don't go overboard with the output capacitors. 
BTW, reading certain product Q&As on Banggood gives me mental skin rashes; people is so soo dumb, and Banggood's staff is so lame it's just unbelievable. :FACEPALM:
Cheers ^:)
A similar psu Barkuti , i found it on ebay for 15$.
I’ll probably also add a diode on the output of the DPS .
I was going to build it with a 220V AC to 30V DC 10Amp converter , but it’s quite expensive (around 40€)
The DPS3012 and DPS5015 do not need a diode to charge batteries or run motors ect. Only the smaller units need this.
Good to know TA .
Thanks
Yes. I am going to wire it in right along with the stock fan. The extra fan is 5v as well so it runs well. Not too loud.
I’m also going to put in a four port 9amp usb hub connected to a volt/amp/power meter. Then I will have a nice monitored usb supply for chargers and the likes.
This is the usb hub. I thought it was 9amps but it is 3amps by two layers for max 5amps. Should still be sufficient though.
I watched the DPS input voltage meter both under load and without load. Fluctuation seemed the same in both situations, but I only had a .4 amp load. All the other server PSU's I have measured in the past have been very tight (like within .02 to .03 volts) voltage range (not counting drop under various loads). I opened up the unit with the lower voltage and it clearly has had some rough usage. It had a lot of smoke residue that seemed to come from 2 different components. One was a shielded SMD inductor on the front board and the other was some component (I think a diode) on the back pcb towards the back. I haven't had a chance to open the other one up. I sure would like to hear what voltage ranges you all are getting on your EPS-470's.
I will have to measure again but mine was right around 47.7V IIRC and very steady with no noticeable fluctuation. Although it was not hooked up to anything and I was using the cheap DMM at the time.
Mine appears to be in very good condition and even still had some warning stickers on it and the cap over the outlet jack was still in place.
^
Tried like you said above. Turned on with just a DMM connected. Both read solid with no fluctuation. One at 48.3v. The other at 48.4v
Just put a cheap meter on mine. It reads 48.39-48.44 volts with no load. That is a steady 48.4 V
Might mention the DPS indicates voltage out to two decimal places also and it wonders around a bit.
My EPS looks new inside with no dust. The fans even look brand new.
Thanks for that report. Sounds like our EPS-470's are behaving the same. So maybe the fluctuation we are seeing on the DPS screen is just the DPS increasing and decreasing it's load on the ESP to get the power it needs. The fluctuations are quite rapid which seems consistent with that idea. It also seems consistent with how steady the output voltage from the DPS appears to be.
EDIT:
This is confusing me. Check out the bottom trace (second picture) for the EPS-470 output jack. I have removed the 3-pronged jack. In both pictures, the through hole on your right is DC Negative, Middle is DC Positive, and Left is for that mystery pin. I had a hard time removing the prong for the mistery pin so you see some damage to the through hole in the picture. The confusing part is that all 3 pins are connected to the same trace pad on the bottom of the PCB. You can see where I scratched the pad to test for continuity. Wouldn't this cause the output to just short out?
I see your pictures and think I follow what you are saying! But, with all due respect, you can’t be correct.
I have checked all three pins. None are connected to case and none are connected to each other.
For my build I plan to just use the existing pins and just plug into the 48v +&-.
Your doing great. Just keep it up.
Just guessing, the trace we see may be a con-conductive stress relief (something or other) for plugging and unplugging the connector.
^
I completely agree with you. The unit works fine. There is something I'm unaware of or something that I'm doing wrong. I'm just puzzled and want to know what is up. The pad is definitely conductive and connected to all three pins. I have a couple things I want to check tonight that I hope will make sense of this.
EDIT: I couldn't really check out what I wanted as I don't know if I altered the trace contacts to the thru-holes when I removed the stock connector. I will be more careful removing the connector on my other unit when I get around to it. I did check for resistance on a EPS-135 between DC Neg and DC Pos (Unplugged from A/C) and it has almost no resistance too.
EDIT2: Reassembled the EPS-470. It works fine still. It draws .25 amps at 122.6v A/C (.19 amps without the fans connected) with no load. It draws .29 amps A/C with a 12v, .42 amp load. I used a Kill A Watt meter to measure the A/C current usage.
Could the third pin be just a voltage sense wire? Purpose being to measure voltage after resistive drops in the power cable and connector for control feedback. I've seen setups with two but especially as this is a rack-mount supply maybe they can assume there are no significant drops on the ground? If I am right it should just be passed/connected through the case wall of the load device (DPS) and attached to hot inside the case. Of course you have a voltage regulator after the supply anyway. Don't blame me if that setup fries the supply. It's just a guess based on no experience with server supplies. If this theory is right it should not produce any current when connected up that way (or any way probably). I suppose one could try it first with high resistance and check for excess current.
Could I buy one of the cheapy power supplies such as this , install it into an enclosure such as this , add a small case fan, add some sort of standard output connector (anyone have suggestions?), and a 120v plug such as this ?
And I guess my next question, in your opinion is this an improvement beyond just buying one of these cheap power supplies and leaving it sitting on my bench?
Yes you can. I would be careful about having bare AC mains exposed on my desk without a case. You could always add enough hot glue I suppose and just cover the contacts that way. That could be reasonable way to go. It still seems a little unprotected to me though with all those holes, might work mounted under a desk where wires can't fall into it.
However I would also consider :
or
or
basically 48V brick/laptop-style supplies.
Those are nothing like 750W and at only 144W not quite as high as your example either, but they require no modification other than possibly changing the connector to suit your needs.
Do you have a decent way out for the air the fan draws in?
If not, there is no good airflow.
Yes. If you look closely in the top picture you will see light shining through the grated vent on the bottom of the enclosure. I think it is enough to make good airflow.