Safely powering 3D printer?

[MigMic]

Hi,

Sometime lurker, first time poster here. I am a mech eng by training, so know just enough about electricity to be a danger to myself and others . I've had a look at other posts but haven't found the exact answer to my specific situation so will start a new thread.

I recently bought a "Migbot" 3D printer kit (https://www.3dprintersonlinestore.com/migbot-prusa-i3-kit), and am in the process of putting it together. I am generally happy with the quality of the build, but have read of a few problems on other forums so am proceeding cautiously. So far the biggest problems I have heard of are:
1) Someone getting "zapped" when touching the power supply case.
2) Someone frying their controller board when using the supplied PSU.
3) The PSU being under-rated for the kit and not being able to supply enough power.

I would like to avoid burning down my house or blowing up the printer (not to mention avoiding electrocution), so I thought I would run some checks before I hook up the power.

Problem 1:
The PSU I have is a no-name unit made in China, maybe a switchmode power supply (see photos)? The details on the other person's problem were sketchy, but from my limited understanding someone should only get a zap if there is a fault AND the case is inadequately earthed. Is that correct? I have checked the resistance and continuity between each of the power terminals and the case with a DMM, and found only the earth terminal is connected (0 ohm resistance/continuity buzzer went off), so if I wire up to an Australian power lead (live, neutral and earth) I presume it should be OK (even if it develops a fault later)? I plan to run it from an overload-protected power board as well.

Problem 2:
The controller board is a "MKS Base" v1.3 similar to the one shown here: http://reprap.org/wiki/MKS_BASE. Unfortunately it is not open-source so the information on it is a bit sketchy. Similarly the details of how the person fried their board aren't exactly clear. Aside from reverse polarity issues, and assuming there are no manufacturing faults with the board, do I need to be concerned that the PSU could damage the board? If people are finding the psu under-rated, then I presume over-current shouldn't be an issue. Is an over-voltage something to be concerned about?

Problem 3:
I'll check this once I have the printer hooked up.

Thanks for any advice you can offer,

Michael

[Rerouter]

Its a generic switch-mode module, the zapping issue would come from the EMI filter caps between Active - Earth, and Neutral - Earth, they leak some current on most devices, enough to give you a good zap when there is no ground connected given (forms a capacitive divider at half mains)


As for frying the controller, you could do that just by poking around the thing while moonwalking on carpet, if your suss, hook the thing up to your multimeter, measure the DC and see that its what you expect, then flick to AC volts, and see that its a small number, e.g. less than 0.3V, this is a very crude way to see that the output is stable, a scope would show far more, but seems like you may not own one,

These types of supplies are generally built pretty cheaply, look for a fuse inside the casing, or install your own, The fact your willing to wire up a mains lead to equiptment means i should already take 3 steps back from this question, but i'll play nice, 

[Mechatrommer]

1) if there is switch for 110V or 220V on the input, make sure you put the switch on the right position based on your mains rating before switch on.
2) if your home/socket has earth wiring (3 wires cable), connect earth wire to the case by any mean. if no earth, proceed to (3) and dont touch the case if you dont feel like to get zapped, its normal as rerouter explained.
3) switch it on in noncombustible area, make sure there is no big bang. if big bang, buy a new psu.
4) no big bang, check with multimeter the psu output for stable 12V. if not stable, buy a new psu.
5) output stable, make sure your 3dprinter board is rated to be feed with 12V supply, if no, buy new 3dprinter board or mod it.
6) everything is ok? switch all off, connect them per manual direction. incl steppers, sensors everything.
7) switch on the test the unit. dont touch or mess with psu sections. if big bang, buy new 3dprinter, or repair.
8] what else better to be suggested? buy a $5000 printer? ymmv.

[MigMic]

Thanks for the quick reply Rerouter, and for playing nice. Perhaps the fact that I am asking questions indicates a degree of caution when it comes to mains electricity  . As it happens I have an electrician friend who can do the wiring, but I want to do my homework first.

Even though there is a hinged plastic "shield" which goes over the terminals, they are not fully insulated, which makes me nervous. How do other people power their printers? When I search for power supplies the only ones I find are either the switch-mode style, or much more expensive "bench" supply units such as Jaycar sell: http://www.jaycar.com.au/Power-Products-Electrical/Power-Supply/Laboratory-Bench/13-8-Volt-20-Amp-DC-Power-Supply/p/MP3098

Thanks,

M.

Edit: added link.

[MigMic]

8] what else better to be suggested? buy a $5000 printer? ymmv.

Thanks for the tips. Yes, you get what you pay for...

[Mechatrommer]

Jaycar sell: http://www.jaycar.com.au/Power-Products-Electrical/Power-Supply/Laboratory-Bench/13-8-Volt-20-Amp-DC-Power-Supply/p/MP3098

the weight 7.8kg is about right for linear psu of this magnitude. today norm is smps, linear only lies in the region of low noise test bench stuffs. if your psu got big bang, your best bet is PC ATX SMPS PSU, easy to find and cheap and more powerfull. otoh if you can tolerate 7.8kg psu due to the elctrophobia then its your call, no harm no objection there, ymmv i also have lower version of your prusa (no lcd), i never experienced any bad things you've explained above. i guess if you are so lucky that your ticket is pulled out from the lucky draw box (1 in a million), you may become one of those who complained.

[PA0PBZ]

The specs for the printer say: Power Requirements 300 W, DC 12V / 30A  so I'm not sure your power supply will do a proper job.

[andtfoot]

Even though there is a hinged plastic "shield" which goes over the terminals, they are not fully insulated, which makes me nervous. How do other people power their printers?

I'm pretty sure those types of power supply modules are usually housed within another enclosure.
In my case, I built one into the control box I put together for my CNC machine. Here's a work-in-progress shot:

[MigMic]

Sorry, that was just the first bench supply example I found. You are both right that it is under specified and "oversized".

This one is much nicer and gets a good review from Dave. Unfortunately a bit out of my price bracket. http://www.jaycar.com.au/Power-Products-Electrical/Power-Supply/Laboratory-Bench/3V-to-15-Volt-DC-40-Amp-Regulated-Switchmode-Laboratory-Power-Supply/p/MP3090

Originally I did look at an ATX PSU, but when they mentioned on the RepRap wiki (http://reprap.org/wiki/Choosing_a_Power_Supply_for_your_RepRap) that it needed to be modified I thought I'd persevere with the PSU which came with the printer (opening up a PSU seemed more risky  ). I should have read further, because with an extension cable there's no need to do any mods.

Time to start researching ATX PSUs now! Thanks for your help!

[MigMic]

Even though there is a hinged plastic "shield" which goes over the terminals, they are not fully insulated, which makes me nervous. How do other people power their printers?

I'm pretty sure those types of power supply modules are usually housed within another enclosure.
In my case, I built one into the control box I put together for my CNC machine. Here's a work-in-progress shot:

Thanks, that makes sense. Nice looking control box andtfoot!

[MigMic]

This one looks like it will do the trick: http://www.msy.com.au/sa/adelaide/pc-components/14158-evga-500w-100-w1-0500-kr-500watt-80plus-power-supply-unit.html

It's in stock at my nearest supplier, is rated to 40A @ 12V, and has good feedback on Amazon. There's also an in-depth review here: http://www.jonnyguru.com/modules.php?name=NDReviews&file=print&reid=384

A few questions:
1) The review states the ripple on the 12V rail is 90mV. Is this likely to be a problem?
2) The review also mentions that there is no heat sink on the bridge rectifier. Is this likely to be an issue since I won't be running it at full load (est. 30A max)?
3) Should I consider putting some protection in between the PSU and the control board? If so, what?

Thanks again,

Michael

[Richard Crowley]

Did the kit come with a power supply?  Your description is confusing.
The specs on the web page says: "AC Input 100v - 240v"
And also: "Power Requirements 300 W, DC 12V / 30A"
That would appear to imply that a properly-rated DC power supply was included with the printer kit.

The power supplies you have shown in the photographs are INTERNAL modules. They are not intended to be used "standalone" outside a proper enclosure.

Using a PC power supply seems rather "dodgy" to my thinking.
A proper 12V 30A power supply would seem like the straightforward solution.

[Mechatrommer]

Time to start researching ATX PSUs now! Thanks for your help!

no need research, no need rocket science modification. you just take care of black and yellow wires (12V), parallel each color (black to black, yellow to yellow) as many as you can for greater amperage capability, each of it are connected to a single point inside the PSU anyway. and lastly you connect green wire to chassis, there is only one green wire there so dont get scary, the rest of colors? you can throw them away... but...

when i said about using atx psu as replacement, thats only when your existing psu goes big bang or you need higher amperage, ie not needing the existing 20A unit shown above. but what you got to lose? i say use that existing underrated psu until it dies. the 30A max is probably if all steppers, heaters are ON at the same time which is very unlikely during normal printing, the psu should survive for sometime i believe.

and why i did advice for it? in case the existing one breaks? because its readily available and cheaper than the specialized smps psu like that, depending on your location, its likely you need reorder from china, like my location.

[sleemanj]

Most 3d printers from china come with that style of psu, including my own.

It works fine (but of course 2 psus that look the same may be entirely different), the only things i did were

1. Throw away the 2 prong power cord and instead sacrifice a nice moulded earthed iec cord for it
2. Crimp ring terminals onto said cord to properly attach to the screw terminals instead of just jamming the wire under them
3. Double check the voltage selector switch which is just visible through the housing
4. Plug in and check output voltage was correct before connecting the control board

Yes, it's really not supposed to be used where people could touch the live terminals, so, don't touch the live terminals!  You could always print a better safety cover ;-)

[MigMic]

Did the kit come with a power supply?  Your description is confusing.
The specs on the web page says: "AC Input 100v - 240v"
And also: "Power Requirements 300 W, DC 12V / 30A"
That would appear to imply that a properly-rated DC power supply was included with the printer kit.

The power supplies you have shown in the photographs are INTERNAL modules. They are not intended to be used "standalone" outside a proper enclosure.

Using a PC power supply seems rather "dodgy" to my thinking.
A proper 12V 30A power supply would seem like the straightforward solution.

Sorry for the confusion. The photos in the first post are of the power supply actually provided with the 3D printer kit. The stated power requirements are 30A @ 12V, but the power supply provided is only rated to 21A @ 12V.

Yes, I now understand that the supplied PSU is meant to go inside another enclosure. Since safety is important to me I am considering finding a suitable enclosure or going with another power supply.

Is using a PC power supply any more "dodgy" that using the internal module without another enclosure? It seems to be common practice for 3D printers. I'm not saying that "common practice" is either safe or ideal, but at least a PC power supply has the mains power inside the enclosure. Drawing DC power out of standard ATX connectors for use in the 3D printer should be no different to powering a PC, or am I missing something?

Just trying to understand so I can make an informed decision. Thanks for your input!

[MigMic]

no need research, no need rocket science modification. you just take care of black and yellow wires (12V), parallel each color (black to black, yellow to yellow) as many as you can for greater amperage capability, each of it are connected to a single point inside the PSU anyway. and lastly you connect green wire to chassis, there is only one green wire there so dont get scary, the rest of colors? you can throw them away... but...

when i said about using atx psu as replacement, thats only when your existing psu goes big bang or you need higher amperage, ie not needing the existing 20A unit shown above. but what you got to lose? i say use that existing underrated psu until it dies. the 30A max is probably if all steppers, heaters are ON at the same time which is very unlikely during normal printing, the psu should survive for sometime i believe.

Your advice made perfect sense. I bought the printer mainly as a way to learn more about electronics, so I'm happy to research alternatives to improve my general understanding. I also bought a cheap printer expecting that I would have to replace or upgrade some bits, so don't mind replacing the power supply if I need to.

[MigMic]

Yes, it's really not supposed to be used where people could touch the live terminals, so, don't touch the live terminals!  You could always print a better safety cover ;-)

Very true. In fact I found some designs on Thingiverse especially for this application.

[Richard Crowley]

Sorry for the confusion. The photos in the first post are of the power supply actually provided with the 3D printer kit. The stated power requirements are 30A @ 12V, but the power supply provided is only rated to 21A @ 12V.

I would NOT a priori decide that the provided PSU is inadequate without at least measuring the ACTUAL current draw from the gadget. 30A seems like a remarkably humongous current for a little 3d Printer.  For that matter, 20 amps seems rather more than I would expect.

Is using a PC power supply any more "dodgy" that using the internal module without another enclosure? It seems to be common practice for 3D printers. I'm not saying that "common practice" is either safe or ideal, but at least a PC power supply has the mains power inside the enclosure. Drawing DC power out of standard ATX connectors for use in the 3D printer should be no different to powering a PC, or am I missing something?

IME, using a PC power supply is rather "dodgy" because PC supplies are designed for a very specific kind of load. They are expecting a substantial current draw at the primary (and secondary) logic voltages, while 12V is a secondary output for peripheral devices.  Drawing all the current from the 12V output is well outside the design parameters of PC power supplies. 

Now, some premium PC power supplies might be designed well enough to thrive on that kind of mis-application.  But it really seems presumptuous to just assume the the provided PSU is "inadequate" and run out and buy something that wasn't designed for the job.  YMMV, but it wouldn't be my first choice.

[MigMic]

But it really seems presumptuous to just assume the the provided PSU is "inadequate" and run out and buy something that wasn't designed for the job.  YMMV, but it wouldn't be my first choice.

I'm not presuming anything . If you have a look at my first post, I was actually asking how to safely hook up and test the supplied PSU (e.g. by adding additional over-current or over-voltage protection) while avoiding problems other people have mentioned. I was only planning to check the report that the PSU is inadequate after I got the whole printer running.

While I admit the thread has gone off on a few tangents (which I don't see as a problem, since I'm learning heaps in the process), I would be happy not to spend any extra money if I can use the supplied PSU safely.

If you could provide some suggestions on testing the PSU (if not yet covered in this thread) I would gratefully receive them.

Cheers,

Michael

[MigMic]

IME, using a PC power supply is rather "dodgy" because PC supplies are designed for a very specific kind of load. They are expecting a substantial current draw at the primary (and secondary) logic voltages, while 12V is a secondary output for peripheral devices.  Drawing all the current from the 12V output is well outside the design parameters of PC power supplies. 

OK, thanks. That makes sense.

[Richard Crowley]

If you could provide some suggestions on testing the PSU (if not yet covered in this thread) I would gratefully receive them.

I would grab an auto head-lamp (or two) and connect them to the output of your 20A PSU.  That should provide a good stiff load to test what it's got.

You should certainly have a DMM if you are doing things like building a 3D printer kit.

If your DMM won't measure 30A (or 20A),  you could rig up an appropriate current shunt to put in series between the PSU and your test load.
Then you could measure the voltage across your current shunt and use Ohm/s Law to calculate the current.

I would probably get one of those $5 digital panel meters from eBay that will display both the voltage AND the current. And build that into the enclosure that you make for the PSU.  Like this one....
http://www.ebay.com/itm/Dual-LED-DC-Digital-Display-Ammeter-Voltmeter-LCD-Panel-Amp-Volt-Meter-100V-100A-/271862702359
(just the first one I found of a great many)

[MigMic]

Thanks. Yes, I've got a lamp somewhere I can use to load up the PSU. I was although thinking of connecting the heat bed PCB directly to the PSU to test it's actual current draw.

Yes, I've got a DMM. Will check it's maximum current rating and get a shunt if necessary.
I like the idea of the panel meter - will investigate as well.

Cheers.

[Richard Crowley]

Honestly all those "problems on other forums" seem like free-floating anxiety expressed by people who don't really understand what they are doing. 

Most certainly, any power supply module like that is potentially dangerous because it was NEVER INTENDED to operate as a stand-alone box.  It was designed to be built INTO some product which provides proper grounding, shielding, mechanical stability, and ventilation. 

Complaining about the potential dangers of an internal PSU like that sitting out on its own is like complaining about a fan with the grille removed.  That is what makes me think the the people who are "complaining" don't know what they are talking about.

Millions of switch-mode power supplies (SMPS) are made in China every year and used in products all over the world. The one in your photo looks pretty average for build quality.  Dunno why one would just assume that it was any more dangerous to use with your gadget than any other SMPS you could buy.  Again, seems like ordinary "free-floating anxiety" to me.  Unless you encountered some statistically-significant string of customers who reported failures with that particular SMPS.  Sorry, not a very compelling argument, IMHO.

As for being "under-rated", If I had to wager one way or the other, I would bet that the "30A" specification is greatly padded. I would be quite astonished if it draws even HALF that.

[sleemanj]

As for being "under-rated", If I had to wager one way or the other, I would bet that the "30A" specification is greatly padded. I would be quite astonished if it draws even HALF that.

We can make some educated guesses.

A typical 3d Printer hot end runs 40W at 12v, so that's 3 Amp
A stepper is going to be 2 Amps at the absolute most, because that's what the typical stepper drivers can run.
A single extruder 3d printer has likely at most 5 steppers (X, Y, 2 for Z and one for the extruder).  2 * 5 = 10A for the motors.
The control board, maybe a display, we'll just be generous and say they did a really bad job and took an amp for that.
The heated bed however is trickier, looking around it seems that they have a resistance of about 1.5 Ohms for the 12v, so that gives us a surprising 8A for the bed if it's running flat out (~100W !)

Anyway, adding that all up into a very much worst-case scenario, we get about 22A at 12v.  But realistically, the bed isn't going to be running full tilt, neither is the hot end, the steppers probably drawing somewhat less than 2A, and the control etc is probably only a couple hundred mA.

[Richard Crowley]

I forgot about the heated bed!  Why on earth do they use "premium" regulated and filtered DC for such a "junk" load as a resistive heater?  That just seems remarkably stupid.