Welcome! Log In Create A New Profile

Advanced

How Hot the Heat Sinks Inside a PSU?

Posted by Garry Bartsch 
How Hot the Heat Sinks Inside a PSU?
October 02, 2014 02:37AM
I am modifying a Corsair CX500 PSU to power my Mendel90 variant. I have clipped the un-needed wires short inside the PSU case (twisted, soldered and insulated the same colored wires together) and now want to tuck and place them neatly so they don't touch the PSU fan and then close the case.

What I want to know is how hot do the aluminum heat sinks inside the PSU get when the PSU is running? Can the bundled wires touch them? I am assuming they won't get hot enough to bother the wires but I don't know and the last thing I want to do is melt plastic inside there and maybe cause a short or fire.

Thanks for any advice or comments.
Re: How Hot the Heat Sinks Inside a PSU?
October 02, 2014 08:49AM
Not that hot. You'd want to leave room for air flow around them so if you can route the wires away from them that would be best, but you won't melt the wires or any if they are near. If they got hot enough to melt the wires, then that would be an extremely inefficiently designed power supply as it's wasting energy in the form of excess heat.
Re: How Hot the Heat Sinks Inside a PSU?
October 02, 2014 08:57AM
For the benefit of anyone else looking at this route, it's easier to get a modular PSU in the first place, e.g. the CX430M which I used to power my 3D printer for a while.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: How Hot the Heat Sinks Inside a PSU?
October 02, 2014 10:26AM
Quote
dc42
For the benefit of anyone else looking at this route, it's easier to get a modular PSU in the first place, e.g. the CX430M which I used to power my 3D printer for a while.
It took about 5 minutes to open my existing supply, clip the unnecessary wires, tape the ends, and put the cover back on. No need to spend the extra on a modular supply that you won't be using almost all of the modularity. Plus many that are going to build their own 3D printer likely have a old power supply laying around anyways...or can get an old used one fairly easily from somewhere.
Re: How Hot the Heat Sinks Inside a PSU?
October 02, 2014 11:49AM
Thank you for those answers. One less thing to worry about ..
Re: How Hot the Heat Sinks Inside a PSU?
October 03, 2014 07:12AM
Instead of stuffing these wires back into the housing I'd unsolder and remove them entirely. Even easier are RepRap controllers which require no PC-PSU modification at all and accept their connectors as-is.


Generation 7 Electronics Teacup Firmware RepRap DIY
     
Re: How Hot the Heat Sinks Inside a PSU?
October 03, 2014 08:27AM
Quote
Traumflug
Instead of stuffing these wires back into the housing I'd unsolder and remove them entirely.
I did that for my first power supply. Groups of 3 or 4 wires were crimped together with a crimped lug, then soldered into the board by the manufacturer. This made unsoldering them quite difficult due to the mass as well as the heat wicking away into the wires being removed. It was just as easy to clip them close to the board and then tape over the ends to prevent them from accidentally making contact with something in case they moved. IOW, desoldering them would be cleaner, but it may be more difficult and there isn't really an advantage gained one way over another.
Re: How Hot the Heat Sinks Inside a PSU?
October 03, 2014 12:45PM
Thank you both. I had the case back together before reading about de-soldering. All went well just tucking them but de-soldering would surely be cleaner. The reason I used this ATX PSU was I had it sitting around; plus the first Mendel90 I made works well with the same mod (and fits onto the machine's frame nicely) so I thought I would stick with something I knew.

One thing I like about this ATX PSU is it is totally silent. Are the OEM type PSUs silent as well?
Re: How Hot the Heat Sinks Inside a PSU?
October 03, 2014 04:49PM
Quote
Garry Bartsch
One thing I like about this ATX PSU is it is totally silent. Are the OEM type PSUs silent as well?
There are all sorts of designs. Some are thermostatically controlled so the fan only kicks on or ramps up if the temperature rises. Other's have better or larger fans that spin quieter or slower. Some are cheap or have smaller fans that run noisier and faster. Based on the noise your printer's mechanics will make when printing, you'll probably never hear the PSU fan when your printer is doing something even if you had a moderately noisy fan.
Re: How Hot the Heat Sinks Inside a PSU?
October 03, 2014 05:37PM
Thanks! Mostly I am concerned that the machine is quiet when it's not working because I don't turn it off between prints. You are right; as soon as it is working any PSU fan would likely not be heard.
Re: How Hot the Heat Sinks Inside a PSU?
October 04, 2014 04:50AM
The 12V 25A oem supply I now use for my printer is fanless and so silent. I got it via ebay uk for less than GBP20. To use these supplies, you need to be comfortable with doing mains wiring and you need to make an insulating cover for the terminals. I have published a design for a printed on one Thingyverse.

These PSUs have a voltage adjustment. This is useful when printing ABS if the heated bed struggles to reach operating temperature using 12V.

There is a thread in the Ormerod forum on modifying an ATX PSU to improve regulation on the 12V rail.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: How Hot the Heat Sinks Inside a PSU?
October 04, 2014 02:30PM
Quote
dc42
There is a thread in the Ormerod forum on modifying an ATX PSU to improve regulation on the 12V rail.

Is this the thread? I read through it. Most was above my head without making a lot of effort to do the calculations but the thread made me curious and I decided to test the heat up times on my Mendel90 which always has seemed to get the bed hot very quickly.

These are the results:
--
Starting from 24C first thing in the morning having done no prints ...
to 60 in 1:15 minutes
to 70 in 1:45
to 80 in 2:25
to 90 in 3:10
to 100 in 4:05
to 110 in 5:15
to 120 in 6:30
to 130 in 8:30

When Pronterface read 130C I measured an average 110C on top of the glass with an IR thermometer. Now I know these thermometers are not known for accuracy - and even worse on glass - but my glass is coated with PVA so less reflective (a nice matte finish) and hopefully this reading isn’t horribly off. I can say the heat coming off the bed is significant. And I put a PLA part onto the bed and it started sticking within 5 seconds. So whatever the exact temperature might be this bed is hot and gets hot quickly.

So I feel like for me the right thing to do is the exact same PSU mod as the first Mendel90, using the same cables and wiring since I know it works. My only question is the controller; the original Mendel90 has a Melzi and I am planning to use RAMPS this time. Can RAMPS transfer power to the heated bed as well as the Melzi?

[It's funny; now as the bed has cooled to ~30C that PLA part is starting to crack/pop as it loosens just like it always does after a print.]
Re: How Hot the Heat Sinks Inside a PSU?
October 04, 2014 06:44PM
Quote
Garry Bartsch
Quote
dc42
There is a thread in the Ormerod forum on modifying an ATX PSU to improve regulation on the 12V rail.

Is this the thread?

The thread I had in mind was this one [forums.reprap.org].



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: How Hot the Heat Sinks Inside a PSU?
October 04, 2014 07:43PM
Thank you. Most of that was over my head; obviously there are some very advanced people on this forum. Hopefully I'll eventually understand more of those types of discussions.
Re: How Hot the Heat Sinks Inside a PSU?
October 06, 2014 09:39AM
Quote
Garry Bartsch
Mostly I am concerned that the machine is quiet when it's not working because I don't turn it off between prints.
What electronics are you using? RAMPS for instance can turn on and off the power supply, running the micro controller just off of standby power until full power is required. Many power supplies aren't at their rated efficiency when they are in essence on buy under an idle load. If you're always printing, then that's not really an issue. But if you leave it on for long periods and not printing, then you're wasting energy (granted, not a huge amount).
Re: How Hot the Heat Sinks Inside a PSU?
October 07, 2014 02:57PM
Quote
cdru
What electronics are you using? RAMPS for instance can turn on and off the power supply, running the micro controller just off of standby power until full power is required.

I am using a Melzi right now but plan to use RAMPS on the machine I am designing. How does RAMPS turn on/off the PSU? Is it automatic? Or do we use some software to control it?
Re: How Hot the Heat Sinks Inside a PSU?
October 07, 2014 11:59PM
Most, if not all electronics solutions use an EN (enable) wire to toogle the stepper driver output to the motors. And most, if not all firmwares do disable the stepper driver outputs after a certain timer of inactivity, like 30 secs or so. The gcode file should contain last commands to turn off motors, extruder and bed (if used). With all those turned off, the only remaining consumer is the uC itself, which is quite neglectable. If it uses like 100-200mA at 12v, in comparison that is like an 10mA led running at mains power, like the led of the tv or such things which are always on. So at this point we are counting two marbles. Coz if you want to minimize energy use, instead of picking on a uC consumption, you can add extra insulation to the fridge, insulate the house, change all lights to leds, use foltovoltaic cells and that sort of things, and that mcu consumption is among last things on the real list.
Re: How Hot the Heat Sinks Inside a PSU?
October 08, 2014 09:52AM
RAMPS has a PS_ON header. The (usually) green wire in the ATX connector is grounded with that header with the M80 command is called, and then ungrounded when M81 is sent. You can put M80 and M81 and the beginning and end of your print job to control the power supply. If you go this route, you'll need to power the MEGA board either via USB, a separate dedicated power supply, or just the +5VSB from the power supply that still is available when the supply is otherwise off.
Re: How Hot the Heat Sinks Inside a PSU?
October 11, 2014 01:46PM
Quote
NoobMan
So at this point we are counting two marbles.

So I guess the waste isn’t too severe smiling smiley

Quote
cdru
You can put M80 and M81 and the beginning and end of your print job to control the power supply.

That is very cool. I'll try this when the build is complete. Or I guess I can test it when I test the electronics before setting them onto the machine.

Quote
cdru
If you go this route, you'll need to power the MEGA board either via USB, ...

So if the USB is connected to the MEGA does that mean the USB is powering it? I'm guessing yes.
Re: How Hot the Heat Sinks Inside a PSU?
October 13, 2014 01:52PM
Quote
Garry Bartsch
So if the USB is connected to the MEGA does that mean the USB is powering it? I'm guessing yes.

I think the mega will choose which supply to use. If there is a supply from jack or ramps it will use that, otherwise it will choose to fed from the usb supply. So yes it will use usb power if only that is available.
Re: How Hot the Heat Sinks Inside a PSU?
October 13, 2014 02:26PM
Quote
NoobMan
Quote
Garry Bartsch
So if the USB is connected to the MEGA does that mean the USB is powering it? I'm guessing yes.

I think the mega will choose which supply to use. If there is a supply from jack or ramps it will use that, otherwise it will choose to fed from the usb supply. So yes it will use usb power if only that is available.

Thank you. I wonder if it matters if the USB was direct from the PC or if it would need to be through a powered USB hub? Does one method give more power than the other?
Re: How Hot the Heat Sinks Inside a PSU?
October 13, 2014 03:37PM
A motherboard usb has some sort of short circuit protection, a current limit (i think in range 1A or less), and if the load wants to draw more than that it gets disconnected instead. A usb hub provides 5v power from its own source, and probably would let a consummer draw a bit more than that, but otherwise it should also have its own kind of protection, probably lets say 2.5A or such. The mega will draw way less than 1A, so its quite ok for it to use the motherboard usb, it would not need a hub.

Power supply wise, regardin +5v and gnd, i dont think there is a difference. Communication wise, maybe, as the logic levels from computer usb are 5v computer and gnd computer, the hub needs to translate that logic to its on 5v hub and gnd hub, and i dont know how that goes in a normal hub. I dont use a usb hub and i have no experience with them. So you can test and see how it goes, maybe check if you get higher error rates one way or the other. But i doubt it should be any difference.
Sorry, only registered users may post in this forum.

Click here to login