Power Supply failure  

Re: Power Supply failure

Finally built my printer last night. Being in Australia we are on 240v mains. So far the PSU is fairly quiet. Only very mild noises.

In fact unless you put your ear right up to the PSU you can’t hear anything.

Im starting to think that most of the failures people are seeing are relating to 110v countries.

As it’s a 240W supply, at 110v it has to draw two times the current from the ac side than it does on 240v. Maybe that’s the root cause.

I have no idea. Time will tell. I have only about 10 hrs print time on it so far. No issues as yet.

Jason.

Re: Power Supply failure

Quieter? Not necesarilly, or should i say probably not. More powerfull PSU's have tendency to include a fan for cooling, which was already confirmed in this thread.

Yes, I am one of the ones who confirmed it. I am running the Meanwell SP-320. It has fan noise, but personally I prefer to hear the white noise of a fan over the rattling, high-pitched vibrating/whining, and constant clicking of the PRUSA PSU. Guess it comes down to which noises are more annoying to you.

Now that I've ditched the stock PRUSA PSU, my printer is at least 50% quieter overall. Of course remote mounting definitely helps as well. It's shocking how much noise the original PSU was making, especially when attached to the frame.

Re: Power Supply failure

Finally built my printer last night. Being in Australia we are on 240v mains. So far the PSU is fairly quiet. Only very mild noises.

In fact unless you put your ear right up to the PSU you can’t hear anything.

Im starting to think that most of the failures people are seeing are relating to 110v countries.

As it’s a 240W supply, at 110v it has to draw two times the current from the ac side than it does on 240v. Maybe that’s the root cause.

I have no idea. Time will tell. I have only about 10 hrs print time on it so far. No issues as yet.

Jason.

Please post if you see issues with your PS. I'm thinking this is the issue and why I'm making space in my garage to test my theory. I wonder if I could test on a high amp 120 plug if it would work?
My idea was to keep the printer in my office, but if I have to keep it in the garage for the 240 power then I'll have to rethink things.

I guess I can hook up video from the garage to my office with hookup to my cell when I'm upstairs...

Re: Power Supply failure

As it’s a 240W supply, at 110v it has to draw two times the current from the ac side than it does on 240v. Maybe that’s the root cause.

This seems pretty likely. It would also explain why the ultra high failure rate of these units seemingly slipped past PRUSA. Being in Prague they are operating on 220v. They've likely done very little to no testing on 110v with these PSUs.

Re: Power Supply failure

As it’s a 240W supply, at 110v it has to draw two times the current from the ac side than it does on 240v. Maybe that’s the root cause.

This seems pretty likely. It would also explain why the ultra high failure rate of these units seemingly slipped past PRUSA. Being in Prague they are operating on 220v. They've likely done very little to no testing on 110v with these PSUs.

I'm also thinking that this is where the CLICKING SOUND comes from.

Re: Power Supply failure

Finally built my printer last night. Being in Australia we are on 240v mains. So far the PSU is fairly quiet. Only very mild noises.

In fact unless you put your ear right up to the PSU you can’t hear anything.

Im starting to think that most of the failures people are seeing are relating to 110v countries.

As it’s a 240W supply, at 110v it has to draw two times the current from the ac side than it does on 240v. Maybe that’s the root cause.

I have no idea. Time will tell. I have only about 10 hrs print time on it so far. No issues as yet.

Jason.

Jason, how many amps is the outlet that your MK3 is plugged into? I found an option to test called a "120v to 220v Circuit Re-joiner". The 240v circuit in the garage is 30amp and it's not wired or cabled. I would have to tear into the wall of a rental so this is my better option and I can stay in my office.... lol I just want to get something that mirrors what you have that's currently working.

Re: Power Supply failure

Jason, how many amps is the outlet that your MK3 is plugged into? I found an option to test called a "120v to 220v Circuit Re-joiner". The 240v circuit in the garage is 30amp and it's not wired or cabled. I would have to tear into the wall of a rental so this is my better option and I can stay in my office.... lol I just want to get something that mirrors what you have that's currently working.

The amp rating on your mains circuit is irrelevant here. Maxxed out at 240w and 240v, you're only pulling 1amp. 240w and 110v, you're only pulling 2.2amps. So less than 20% of the capacity of a typical 110v/15amp circuit.

Furthermore, the PSU is going to behave exactly the same regardless of whether it's on a 10amp, 15amp, or 30amp circuit of the same voltage. Just because a circuit has a higher max amperage capacity doesn't mean the printer will draw any more/less current. It's the voltage that matters here, not the amperage rating of the mains.

Lastly, the 220v circuit rejoiner is like $275 (and bulky as hell). Why wouldn't you just buy a new 3rd-party 110v PSU for $50? Doesn't seem to make a lot of sense.

Re: Power Supply failure

Just to add to this. The max Amps we have here in Australia in a std 240v socket is 10A. So it’s not related to current.

It likely is the voltage though.

So if it does turn out to be a 110v issue for the PSU, the cheapest option apart from a warranty replacement from prusa that works reliably on 110v would be the aftermarket Meanwell types.

I will of course report any changes I encounter. I now have a lot more print time on the machine and so far no noises or other issues to report.

Jason.

Re: Power Supply failure

Jason, how many amps is the outlet that your MK3 is plugged into? I found an option to test called a "120v to 220v Circuit Re-joiner". The 240v circuit in the garage is 30amp and it's not wired or cabled. I would have to tear into the wall of a rental so this is my better option and I can stay in my office.... lol I just want to get something that mirrors what you have that's currently working.

The amp rating on your mains circuit is irrelevant here. Maxxed out at 240w and 240v, you're only pulling 1amp. 240w and 110v, you're only pulling 2.2amps. So less than 20% of the capacity of a typical 110v/15amp circuit.

Furthermore, the PSU is going to behave exactly the same regardless of whether it's on a 10amp, 15amp, or 30amp circuit of the same voltage. Just because a circuit has a higher max amperage capacity doesn't mean the printer will draw any more/less current. It's the voltage that matters here, not the amperage rating of the mains.

Lastly, the 220v circuit rejoiner is like $275 (and bulky as hell). Why wouldn't you just buy a new 3rd-party 110v PSU for $50? Doesn't seem to make a lot of sense.

Josh didn't mean to stir things up I'm just trying to isolate some things before my MK3 arrives and maybe help a few ohters. NO I'm not an electrician and I'm not thinking I have to plug into a 30A circuit to get it to work.
Something is causing an "unknown percentage" of PRUSA power supplies to fail in the US and not in countries that run 240v circuits. I've ruled out two things.
1. I asked if people that experienced PS failures if they had other things on the circuit at the same time. You know how an old vacuum cleaner or space heater can draw power and make it dip?
2. I asked if any failures were caused when the MK3 was attached to a UPS. I received two YES answers. So this rules out dirty/clean power.

No I'm not an electrician, but I was in test and evaluation for over a decade in aviation and used those skills to troubleshoot issues for 8 years in a large football size server farm as a network engineer.
I don't know if the power supply that ships with my MK3 will display similar symptoms and failure or not. So my plan is to mirror the best I can the electrical circuit that is not causing PS's to fail.
After having a reasonable amount of printing time maybe 100 hours I'll move over to normal house circuits. If the power supply fails then we know the failure the lower voltage.

My test is to see if my power supply is going to fail anyways will it fail on 240v? If it doesn't, and after I move it to 120v will I get the clicking sounds and will it fail like so many others? Who knows, maybe I'll get a good PS and I'll be wasting my time.

Re: Power Supply failure

The one thing no one has mentioned so far is the AC Frequency. The majority of the world uses 50Hz while USA is 60Hz. It's been more years than I care to reveal since I worked power systems; so I admit my knowledge is a bit out of date. It just seems like that's the difference more likely to cause an issue.

Re: Power Supply failure

The one thing no one has mentioned so far is the AC Frequency. The majority of the world uses 50Hz while USA is 60Hz. It's been more years than I care to reveal since I worked power systems; so I admit my knowledge is a bit out of date. It just seems like that's the difference more likely to cause an issue.

Joe, with my questioning and comments that was assumed, but you are right. With my test on 240v if it fails on 240v, 60 cycles then that would be the one test that hasn't been look at.
If that is successful and I move to 120v, 60 cycles and it fails then we can rule that out. Like I said, I'm not an electrician, but I've played with power since I was a kid.

Re: Power Supply failure

Paul,

I've printed and mounted the PSU with your model. There is a slight problem though. The joint that connects to the aluminum extrusion is too high and the bed hits the joint. Can you lower it a bit?

I knew I was stupid to post before I'd mounted with it.

Yes, I can lower that angle brace, I know exactly what you mean. I 'dry mounted' it, but hadn't printed with it yet.
...

I've modified the brace. The other components are unmodified. I'm printing it now (this time I've learned my lesson, I'm going to mount it, wire it up, and print with it before I post it).

Changes I've made:
1) pulled the brace angle bracket at the bottom back so it won't interfere with the bed.
2) made the holes facing the frames vertical slots so tightening the brace can't bend the frame

Printing now, should have results tonight, likely post tomorrow.

Re: Power Supply failure

One interesting note:

My second replacement power supply came with a wire harness where the positive DC wires were solid red. My original and my first replacement had both black wires, with a red stripe down the side of the positive wire, but it sometimes made it difficult to tell which was positive depending on the viewing angle.

It could just be variability in the manufacturing process or a different supplier, but it seems likely they had enough people building the kit with reversed polarity on the PSU connection that they made it more obvious which was the red wire.

Re: Power Supply failure

Paul,

I've printed and mounted the PSU with your model. There is a slight problem though. The joint that connects to the aluminum extrusion is too high and the bed hits the joint. Can you lower it a bit?

I knew I was stupid to post before I'd mounted with it.

Yes, I can lower that angle brace, I know exactly what you mean. I 'dry mounted' it, but hadn't printed with it yet.
...

I've modified the brace. The other components are unmodified. I'm printing it now (this time I've learned my lesson, I'm going to mount it, wire it up, and print with it before I post it).

Changes I've made:
1) pulled the brace angle bracket at the bottom back so it won't interfere with the bed.
2) made the holes facing the frames vertical slots so tightening the brace can't bend the frame

Printing now, should have results tonight, likely post tomorrow.

Thanks very much. I'll keep an eye on thingiverse for the update.

Re: Power Supply failure

Josh didn't mean to stir things up I'm just trying to isolate some things before my MK3 arrives and maybe help a few ohters. NO I'm not an electrician and I'm not thinking I have to plug into a 30A circuit to get it to work.

No worries, more power to you if you want to spend money/time/effort testing out different PSU scenarios. I tend to think you're way over thinking it though. At the very most your tests are likely to support what we already anecdotally suspect is true: The PRUSA PSU is far more likely to fail on 110v than on 240v, and there isn't any more specific circumstance making it fail, nor a remedy for preventing it.

Either way, none of that changes the 2 options available to most 110v users: 1) Continue to hope you don't blow (another) power supply and continue to order warranty replacements when you do, or go ahead and spend $50 on a quality non-PRUSA PSU and move on.

Re: Power Supply failure

Yesterday my power supply failed quietly. It was in the middle of a print (PLA, normal temps), and it just stopped. No fireworks, no noises, just a dead printer. I pulled the fuse, and it was blown. Presumably the power supply itself failed and by attempting to draw too much current, blew the fuse. I can replace the fuse, but I'm pretty sure that whatever caused the first fuse to blow will reoccur.

I reached out to PR support via chat. Pavla tried to pretend that there were no known problems with PSUs, but when I pointed out this thread, the new answer was that it would have to be resolved with colleagues and that someone would get back to me via email. So far no emails.

[16:54:36] Craig Trader: The power supply on my Mk3 just died in the middle of a print.
[16:55:25] Craig Trader: This is from order # xxxxxxxxxx.
[16:55:26] Pavla: Pavla joined conversation
[16:55:45] Pavla: Ok, Craig, can you describe it a little more please?
[16:58:23] Craig Trader: I was running a fairly normal print (PLA normal temperatures); it ran for a couple of hours and then the entire printer powered off in the middle of the print, LCD screen black. Switched it off, switched it on, still no power. Tested the power cord, it's still getting power (110V). My Mk2, plugged into the same circuit is still printing fine.
[16:58:36] Craig Trader: The power supply is dead.
[17:01:12] Craig Trader: I have been running prints on the Mk3 for the last 24 hours, pretty much non-stop (same goes for the Mk2 beside it). Nothing seemed wrong.
[17:06:27] Craig Trader: Thoughts, comments?
[17:07:31] Pavla: sorry, Craig, I have 20+ people here, trying to do my best
[17:20:25] Craig Trader: The fuse appears to be blown, but that doesn't explain *why* the fuse is blown, given a normally functioning printer.
[17:21:18] Pavla: hmm, this is really weird
[17:21:42] Pavla: so everything was normal
[17:21:45] Pavla: nothing changed
[17:21:53] Pavla: not even the firmware, anything?
[17:21:58] Pavla: no new components?
[17:22:44] Craig Trader: Latest firmware (3.1.1 final). No new components.
[17:23:27] Craig Trader: I have a Pi3 connected, but that runs off its own power supply.
[17:26:39] Craig Trader: It would be helpful to know what size fuse to replace with, but I'm pretty certain that without knowing what's wrong, If I just replace the fuse, it will just happen again.
[17:27:02] Pavla: yes, I see
[17:28:24] Pavla: maybe the fuses could be the problem - faulty or something
[17:28:43] Craig Trader: From reading the support forum, I know that there have been problems with the power supplies.
[17:29:18] Craig Trader: Fuses only blow in an over-current situation (that's their job) -- the fuse doesn't *cause* the over-current problem.
[17:30:12] Pavla: yes, but only with clicking and we solved this issue - it was caused with the sheet which was vibrating on a certain frequency and is interfering - but this causes nothing
[17:31:02] Craig Trader: And yet I am not the first person to report a failed power supply:
https://shop.prusa3d.com/forum/original-prusa-i3-mk3-f30/power-supply-failure-t13443-s230.html

[17:31:23] #Craig Trader: Not 'clicking failed' but 'no power' failed.
[17:44:07] Pavla: Ok, Craig, we have to reslove this tomorrow with my colleagues. I will use the log of this conversation and we will get back to you via e.mail, alright?
[17:45:50] #Craig Trader: Sure.
[17:46:54] Pavla: Ok, thank you for the patience, will discuss it with developers and others and will let you know
[17:53:23] #Craig Trader: FWIW, I pulled the power supply out of the printer, and there's something rattling around inside the metal power supply itself.
[17:54:06] #Craig Trader: (Not just the screw wire connectors.)

Re: Power Supply failure

Thanks very much. I'll keep an eye on thingiverse for the update.

The latest version of my MeanWell 350 !24V! PSU holder is up on thingiverse:

https://www.thingiverse.com/thing:2783871

1) fixed conflict with bed (D'oh!)
2) made frame mount holes into slots to avoid any possibility of torquing the frame.

Re: Power Supply failure

I'm going to give an update on my 2nd PSU failure.

Talked to chat Saturday and there was one guy working. He asked me to send an email to Prusa describing my problem. Jennes responded at 9:05AM EST asking me to check resistance on the EINSY inputs to verify there is no short on the EINSY itself. I disconnected the two legs from the board and verified the following resistances. NOTE: I had ordered an EINSY board directly from Ultimachine for a Zaribo I'm saving up parts for and they matched these exactly. So it looks like the EINSY isn't the issue and it seems to be undamaged thankfully. I sent these to Prusa and I'm waiting for a response.

Resistances with PSU unattached
Input 1: 540 ohms
Input 2 (Heatbed input): 4.3 k-ohms

Heat Bed output : 113 k-ohms #This is normal obviously because it's not on asking for power.
Heat bed resistance: 2.8 ohms

PSU Resistances:
output 1: 320 ohms
output 2: 320 ohms

So this afternoon replacement fuses I ordered came in and I decided I'd try one out just for fuck's sake. Well..... Pretty much a mini explosion happened inside the PSU. Lord I was scared 😯 (get the reference?). Luckily I had disconnected mains from the EINSY and covered the connectors with insulation so I wouldn't accidentally short the PSU. Long story short, This IS a safety issue irregardless of what Prusa says. If just one spark was not contained inside the PSU, it could have started a fire. Should I trust a 3rd replacement? Over the years, I've had computer PSUs fail eventually and none of them failed this violently. I've never had a PSU fail since I've been living in this house (4 years)

Re: Power Supply failure

Here are pics from the failed power supply. When it failed there was a flash of fire from inside the case. The printer had been printing for almost 70 hours when this happened.

The first pic is the blown thermistor. there may be other parts that let out the magic smoke as well.

The second pic is the chunk of thermistor that fell out of the PSU case.

This is the exact same part that exploded in mine. This is supposed to be a resettable thermistor correct? It should interrupt power, trip, then reset as it cools.

Re: Power Supply failure

I would love for someone to send one of these failed power supplies to BigClive on YouTube.
https://www.youtube.com/channel/UCtM5z2gkrGRuWd0JQMx76qA

He's a dry witted Scottish electrical wizard who loves dissecting things like power supplies, tracing out the circuitry and working out why the "magic smoke" came out of something.

Has anyone had any feedback from Prusa when their old power supplies were returned to them for inspection?