Server Powersupply mod

BillC

Active member
If +vref is the same level for all regulators I wonder if you might have to provide the +12v reg . with its own seperate +vref in order to increase its output only, to the desired 13.6 v, just an idea..
 

VK3ZYZ

Moderator
Staff member
Applying 12V to pin 44, 0V to 46, gives 2.5V on 45.

EDIT: Changing R34 from 910K to 1M did not alter anything.
 
Last edited:

BillC

Active member
With LM324n, looks like if pin 5 goes low then output pin 7 also goes from high to low and pulls current through diode and 36k from pin 38 ? Wonder what is on pin 38 , maybe the main regulator.
 

VK3ZYZ

Moderator
Staff member
I was having a lot of trouble printing a new cover....
FailedPrints.jpg

These all stopped printing with jammed filament about half way through the print.
In the process of the black PLA prints, I had to pull the hot end apart to remove the filament and found it was a bit loose. That was the problem, I hoped. But no!
So, maybe the black is the problem?
Well, you can see that changing to the grey did not fix it.

To do these prints, I have changed the slicer program used. The slicer is the program that converts the drawing .stl file into the .gcode the printer understands. This new version has a setting to keep the holes the correct size. Using the older version of the slicer, the holes turned out smaller that set so I'd fiddled the hole sizes to fix that. The setting in the slicer does the "fiddling" for me now. As a result of using the new slicer I may go and redo the old files.
Anyway, it turns out the problem with the print stopping looks to the the retraction setting in the new slicer. The retraction is the filament being pulled back into the printer a bit so there will not be melted filament oozing out while the head is moving and not printing.
Mr Google showed one cause of the filament jamming could be the retraction set too high. If it is, more heat gets transfered up into the cold part of the hot end eventually causing the filament to melt too early so causing a jam.
The retraction was set at 6.5mm and a change to 2.5mm fixed it :)

This version starts out with black for 3 layers, then I change to orange for a bit, then back to black.

12V40A.jpg
 
Last edited:

VK3ZYZ

Moderator
Staff member
A worthwhile simple mod on the R2Z-6400P-R 28Amp version is to unplug the two 12V fans and hook them onto 5V. I kust taped the connectors together, bent a couple pieces of wire (resistor lead) into U shapes so they plug both "12V" connections together and same for the 0V. Then soldered a length of hookup wire to these links and ran them to the PS 5V and 0V connections.
Now, the fans run quietly, instead of as sirens!
 

VK3ZYZ

Moderator
Staff member
I'm thinking of using some PVC down pipe for the case as it will be a lot easier and faster to produce. There will have to be a few adapters made to fit as the pipe is bigger the the power supplies, but here is the first "working" effort.
The front is similar to the earlier ones...
FrontFront.jpg
FrontBack.jpg

but being removable will make wiring a lot easier.

FrontTube.jpg


It fits ok, and will have screws in from the sides, top and bottom.
 

VK3ZYZ

Moderator
Staff member
Here it is all printed out.
CaseParts.jpg

There is a front, spacer and rear feet.
Spacer.jpg
SpacerInTube.jpg

The spacer has air channels.
SideView.jpg

Feet can be fitted at the rear to level it.
RearFeet.jpg


FrontInstalled.jpg

I have not installed the 12V sockets as yet.

The fully 3D printed case may be neater but it takes a long time to print.
 
Last edited:

VK3ZYZ

Moderator
Staff member
Here are the steps to wire it up to a connector that has been cut off the power supply cradle chassis.
First, the power switch and LED. Note the 220R resistor in line with the LED.
Switch_LED.JPG

Next, the USB sockets. I have included a 5A Polyswitch, and no USB resistors.
USB_Power.JPG

Now to thea12V. I used Jaucar flat speaker wire as it is very flexible, and has a lot of copper.
First12V.JPG
All12V.JPG


The fans normally run on 12V and if you like sirens, you can keep it that way. I prefer a quieter supply so here are the steps to change to 5V operation.
First, take the PS top off.
PS_top.JPG
The fans need to be unplugged. There will be white gunk that needs to be cut to free the plugs.
A couple of wire links bent to fit..

Link.JPG
and the connectors wired together.
Fan5VMod.JPG
then secured with heat shrink or tape
.FanModSecured.JPG
Finally, with the top put back and the mounting sleeve fitted, the fan wires can be attached to the socket.
Fan_Wired.JPG

It works :)
ItWorks.jpg


( I used my soldering iron to "drill" screw holes in each side to hold the front on.)
 
Last edited:

VK3ZYZ

Moderator
Staff member
Here is another. Thanks to Josh for the connections to get is going.
This is for the TDPS-750AB A unit that is 750Watts.
The main connector has +12V at 54Amps, +5V at 33Amps and 0V pins. I just soldered heavy cables onto the pins as I've not got a connector.
Connections.jpg

Josh sent me the info to get it working. In his picture, there are a couple of black wires going to an on/off swicth. (I just linked direct to0V and used the power switch.)
EnableMods.jpeg

This is a nice flat power supply so your set could well sit on top of it. About 280x210x40.
One problem is the 2 fans installed are 12V at 6A!!!! They make great blower fans, or sirens if you prefer. And the power supply shuts down if the fans fail to get to speed.
There may be a hardware mod for this but my fix was a bit different.
Both fans were removed and the one with the longer leads replaced but turned 180 degrees so it now blows onto the main power supply heat sink.
First, I tried running the fan on 5V but it would not start. Then on 7V from en external power supply as 7V looks to be the minimum voltage to get it going to keep the supply alive. But the fan is still too loud. Next, running through various power resistors but that was not satisfactory. Another try was to use series diodes to drop the voltage, In this case, I have a bag of W04 bridge rectifiers so 3 of them in line dropped the voltage ok but they got hot and the fan is still too noisy.
The final fix I came up with is to cut the fan red wire and install a 12V 5W tail light lamp in series. This lets the fan speed up when the lamp is cold, then slows as the lamp heats. As the fan now slows down, the lamp cools so the fan speeds up again. This repeats so the power supply "breaths" but keeps working and has air to cool it.
PS_Fan_SeriesLamp.jpg


There was a web site than came up with maybe a part of the circuit...

TDPS-750AB A_cct_maybe.jpg

Another fan mod I hope to try is to add one of my PIC12F629 boards into the fan tacho line to double (or more) the pulse frequency to see if that can get the fan running slowly. And to increase the 12V to 13.8V too!
 
Last edited:

VK3ZYZ

Moderator
Staff member
Here is the latest version front for the server power supply mod making use of the 100x50mm PVC down pipe for the case.
The spacers are still to be printed, but enough for tonight!

I've added a spring terminal for those times when 12V is needed but no power pole connectors are fitted.
If I get a chance, it may get installed tomorrow
.
Photo on 23-7-23 at 9.28 am.jpg
 
Last edited:

BillC

Active member
Clever use of the varying resistance curve of a lamp on the fan speed control on your power supply. I have the same thing here on a system with a number of relay coils in a dc motor start and run setup, lamps in series reduce the hold in relay coil voltage during the motor run mode of the unit. Nice looking outlet panel on your power supply plenty of outlets too.
 

VK3ZYZ

Moderator
Staff member
Here is the "final" setup for the R2Z-6400P-R power supply. And it works :)
Photo on 24-7-23 at 8.15 pm.jpg


FrontConnections.JPG
Front panel connections

5V to Fan.JPG
5V to fans. It was 12V but then they are to loud.

PlugSide.JPG
Connections to the front.


PCB_Connections.JPG
Connections to the socket PCB.
I may add Polyswitches and programming resistors to the USB sockets, maybe ;)
 
Last edited:

VK3YNV

Administrator
Staff member
The adaptor PCB's arrived, so there should be a few finished by the time of the hamfest. I need to redo the USB connectors, and figure out why I couldn't plug into them after assembly. I suspect that the pins moved when I soldered the connections, so it might be necessary to have a male connector plugged in while soldering the connections.

IMG_5348.JPG

IMG_5350.JPG
 

VK3ZYZ

Moderator
Staff member
To make the USB connection easier, I've got some boards on order...
1692585213224.png

There is provision for installing SMT current setting resistors for those loads that have that option and an LED resistor too.
EDIT!: 20230825 the PCBs are in transit.
 
Last edited:

VK3ZYZ

Moderator
Staff member
1693297591143.png

TDPS-750AB A server power supply "front". Still needs the LED wired and the spring terminals fitted.
 
Top