Re: MK3 Power Supply Clicking
Odds are your lights are wired into the same breaker as your outlets and you are causing a serious voltage droop
A breaker is simply a switch. It does not isolate anything from anything else. The problems appears to be weak step up circuitry in the PSU, showcased well for 110v users. The clicking sound is likely an electrical arc due to massive short pulses of power, rather than high frequency pulse width modulation (super fast, tiny pulses).
Bobstro's experience of the sound passing in time suggest something actually got welded together. Extended, this suggests that something loose in this PSU. I can testify to the fact that quality control is weak, because when I inspected my PSU before install, I did note one screw loose that would just slip and not tighten.
Personally I am taking this issue *VERY* seriously. I will get this PSU replaced one way or the other, and perhaps retask it to run some LED strips.
Today I sent in my first warranty requests, five separate ones I think. I refrained from sending the 6th even though it was due to a defective printed part that broke during assembly.
Re: MK3 Power Supply Clicking
Odds are your lights are wired into the same breaker as your outlets and you are causing a serious voltage droop
A breaker is simply a switch. It does not isolate anything from anything else.
You may be right about the cause of the problem, but voltage droop is a real thing in a wire where items are placed in series. Often outlets and lights are wired in series. They should not be wired together, but in older houses (or when the homeowner does modifications to the house) they are wired together and in series...or there are simply too many loads on the circuit.
Voltage Drop - Definition
Wires carrying current always have inherent resistance, or impedance, to current flow. Voltage drop is defined as the amount of voltage loss that occurs through all or part of a circuit due to impedance.
A common analogy used to explain voltage, current and voltage drop is a garden hose. Voltage is analogous to the water pressure supplied to the hose. Current is analogous to the water flowing through the hose. And the inherent resistance of the hose is determined by the type and size of the hose - just like the type and size of an electrical wire determines its resistance.
Excessive voltage drop in a circuit can cause lights to flicker or burn dimly, heaters to heat poorly, and motors to run hotter than normal and burn out. This condition causes the load to work harder with less voltage pushing the current.
https://c03.apogee.net/contentplayer/?coursetype=foe&utilityid=wppi&id=4578
Re: MK3 Power Supply Clicking
Just live chat them, and you’ll have a replacement in 3 days.
Re: MK3 Power Supply Clicking
See my post here: https://shop.prusa3d.com/forum/hardware-firmware-and-software-help-f64/power-supply-failure-t13443-s550.html#p110589
The above was taken using a current probe at 1mV/100mA, so it's showing ~4.8A peak.
I'm going to take my Prusa supply apart and look at the components and ratings, but the short of it is that input side is dealing with bursts of >3A RMS. My expectation is the high current pulses, when running on 120VAC, is stressing components and folks that have a supply with parts that slightly weaker (but not out of spec) are seeing energetic failures.
Re: MK3 Power Supply Clicking
So I have the same problem, just assembled yesterday morning.
The click is defiantly there very sharp and noticeable.
I'm going to take my Prusa supply apart and look at the components and ratings, but the short of it is that input side is dealing with bursts of >6A RMS. My expectation is the high current pulses, when running on 120VAC, is stressing components and folks that have a supply with parts that slightly weaker (but not out of spec) are seeing energetic failures.
Good job on testing the PSU, I'v also seen your other post, maybe u could try to get a differential prob? (if u can afford that 😀 )
I guess someone has to tear the PSU apart and run tests to see where's the trouble than. (too bad it's around final time for me, other wise I'll do it)
Re: MK3 Power Supply Clicking
So I have the same problem, just assembled yesterday morning.
The click is defiantly there very sharp and noticeable.
I'm going to take my Prusa supply apart and look at the components and ratings, but the short of it is that input side is dealing with bursts of >6A RMS. My expectation is the high current pulses, when running on 120VAC, is stressing components and folks that have a supply with parts that slightly weaker (but not out of spec) are seeing energetic failures.
Good job on testing the PSU, I'v also seen your other post, maybe u could try to get a differential prob? (if u can afford that 😀 )
I guess someone has to tear the PSU apart and run tests to see where's the trouble than. (too bad it's around final time for me, other wise I'll do it)
I should note that I made a stupid math error in both posts, which have been corrected. Dealing with AC -- divide by two... 😳
Anyway, I plan to take apart and analyze the power supply. Might get to start this afternoon, interspersed with calibration cube prints. 😀
Re: MK3 Power Supply Clicking
I should note that I made a stupid math error in both posts, which have been corrected. Dealing with AC -- divide by two... 😳
Anyway, I plan to take apart and analyze the power supply. Might get to start this afternoon, interspersed with calibration cube prints. 😀
Any updates on this? The clicking was the first thing I noticed once I had mine up and running.
Re: MK3 Power Supply Clicking
I should note that I made a stupid math error in both posts, which have been corrected. Dealing with AC -- divide by two... 😳
Anyway, I plan to take apart and analyze the power supply. Might get to start this afternoon, interspersed with calibration cube prints. 😀
Any updates on this? The clicking was the first thing I noticed once I had mine up and running.
Unfortunately, no. I have the bare PCB sitting on my desk. However, work and family have greatly reduced my available time and manually tracing a PCB is quite time consuming.
Re: MK3 Power Supply Clicking
Just built my MK3 kit -- when the bed is on it appears to display the same PSU clicking in time with the heatbed indicator. (Though it seems to start and stop at certain PWM frequencies?) Firmware shows as 3.4.1-1356, but from a check of the Github release history nothing dealing with this specific problem has come out since then.
Running in Canada on 115V RMS, 60hz.
(Also, off-topic, the forum login form doesn't seem to accept my password, but the shop login page does and logs me in here as well. Probably the special characters...)
EDIT: Updated firmware to 3.5.0 (shows as 3.5.0-1749) and problem persists. If it's just a little noise, it's fine; but I'll be on the lookout for other power supply issues in the future -- reading through this thread and the linked material, it looks like it's (understandably) not been thoroughly tested in this application with North American power standards.
Re: MK3 Power Supply Clicking
I also had the clicking. It didn't bother me too much. But today the psu died completely mid-print. It was only 4 months old! They are sending a new one, but this seems like something is seriously wrong for it to die so soon.
Re: MK3 Power Supply Clicking
Hey guys,
This IS a power supply issue. There will be no fixing it with a firmware update.
This resolved it for me. https://www.thingiverse.com/thing:3082178
Print the new mount and purchase the PSU linked.
There is a theory floating around that you can silicone the hell out of the inductors in the PSU, but I have not tried it myself. I just jumped right to a verified fix.
Re: MK3 Power Supply Clicking
Well that much is obvious. But no matter what the problem is, you've got to perform the ritual equivalent of "try turning it off and on again, and unplug and replug the cable" so that a company accepts your problems as real.
I'm just waiting for Prusa to offer a more properly-specced and 120V-friendly PSU replacement either through support, in the webshop, or as a sanctioned modification tutorial [unlikely, because of mains-voltage wiring]; because using a non-standard PSU sounds like a great excuse to decline other support issues if they arise -- even though you probably didn't break anything, it's basically impossible to PROVE you didn't break anything.
I am quite happy with my printer in general though. The clicking is a little annoying, but the only thing that really worries me is how anecdotally people seem to think it correlates with reduced PSU lifespans. Maybe I should think about printing the plastic parts for this mod, even though I'm not planning on performing it unless my PSU dies, so that if it does happen I have the parts on hand...
Oh well. It can wait until I get some PETG.
Re: MK3 Power Supply Clicking
I did a detailed writeup about the MK3 PSU here.
The clicking is 100% normal and should cause you no concern, despite the alarmist claims throughout this forum.
Prusa Support will RMA the power supply if it happens to fail.
Re: MK3 Power Supply Clicking
I stumbled into this thread in my search for answers for some strange behavior with my new i3 mk3. Yes, clicking/ticking noises from the power supply, and LED lighting in the same room to very slightly but noticably flicker when the PWM for the bed heater is active. The voltage change isn't sufficient enough for UPS to care on even the most aggressive settings. The LED flicker is especially interesting because LED lamp designs have virtually no capacity to buffer against changes of the input and are by design more susceptible and indicators.
tl;dr, there are some interesting points to discuss and findings of my own. I found the article linked in the previous reply interesting. Seems plausible at first blush, but also omits some key aspects of power supplies. And the discussion of the DC voltages is interesting data points, but it ignores what happens in the switching power supply. And why uprating the power supply changes these symptoms immediately. Anyway, let me get to two important points:
a) efficiency
A power supply rated for x output doesn't have x to provide as a useful output. total output (240W rated) * efficiency = useful output. Does anyone know what the efficiency of the stock supply is? This matters even more because a switching power supply on a nominal 230VAC will be MORE efficient than nominal 110VAC. It's entirely possible power supply sizing is just fine on 230VAC line voltage and triggered specifically by 110VAC systems where you'd have to derate the capacity further.
If we add the observation of actual draw from the prior reply (which doesn't say how this was established), that means if you draw 215W from the useful output as mentioned in the prior reply you're drawing 89% of the total capacity. This would imply it's very much up there in terms of useful output and pushing the power supply pretty hard.
b) actual draw
The statement in the linked post in the prior reply says a 215W draw was measured. I can't confirm that from what I'm seeing. I first noticed this when I moved the printer to a dedicated UPS this morning. Interestingly, with 124VAC line voltage @ 60 HZ, it says the i3 mk3 draws 260W from the printer to the APC UPS when the preheating kicks in (not UPS to wall, so, this isn't introducing additional losses). The "power on" draw for the heater is at least 260W on the line side, and then slowly declines to 255W or so over the next couple of seconds. The actual draw at power on could be even higher when the PWM flank kicks in. But, I don't have any measured proof for that. Somebody who really understands how this type of switching supply works may be able to comment better on what PWM load spikes from an inductive load cause to happen because of to the unique way in which switching power supplies work (rather than, say, a linear power supply) and what this does to the line side in terms of voltage draw etc. PWM loads and constant loads are far from comparable.
I have since then also confirmed this draw this draw with two different clamp meters. What I haven't done yet is drag my scope into this and record the traces. Not sure that is worth it. I'm significantly above the 215W draw mentioned by the prior poster, and confirmed across 3 "instruments".
Bottom line.. If the power supply that comes with it is indeed rated for 240W, based on the above alone, we can say there is concern the power supply when used on 110VAC line side is underdimensioned. This explains why I heard clicking/ticking from the power supply and saw LED lighting on the same circuit flicker (spare me the outright dismissal that this is an electrical problem in the house line please; it's not, there is plenty of power monitoring in this house to refute that and it doesn't invalidate the points above but misunderstands how electrical systems work).
This all leads credence to the suggestion that increasing the capacity of the power supply does help with 110VAC supply side, and explains why this may not be observed in other parts of the world where line voltage is 230VAC. And why people are observing failures when a power supply is pushed harder due to reduced efficiency as a result of supply voltage. And it perfectly explains why uprating the power supply to, for example, 350W cures these ills.
The analysis referenced in the prior post misses all of these points. I think there's room for improvement here for what the power supply is rated for on 110VAC line voltage. YMMV.
Re: MK3 Power Supply Clicking
I stumbled into this thread in my search for answers for some strange behavior with my new i3 mk3. Yes, clicking/ticking noises from the power supply, and LED lighting in the same room to very slightly but noticably flicker when the PWM for the bed heater is active. The voltage change isn't sufficient enough for UPS to care on even the most aggressive settings. The LED flicker is especially interesting because LED lamp designs have virtually no capacity to buffer against changes of the input and are by design more susceptible and indicators.
...<snip>...
a) efficiency
A power supply rated for x output doesn't have x to provide as a useful output. total output (240W rated) * efficiency = useful output. Does anyone know what the efficiency of the stock supply is?
...<snip>...
b) actual draw
The statement in the linked post in the prior reply says a 215W draw was measured. I can't confirm that from what I'm seeing. I first noticed this when I moved the printer to a dedicated UPS this morning. Interestingly, with 124VAC line voltage @ 60 HZ, it says the i3 mk3 draws 260W from the printer to the APC UPS when the preheating kicks in (not UPS to wall, so, this isn't introducing additional losses). The "power on" draw for the heater is at least 260W on the line side, and then slowly declines to 255W or so over the next couple of seconds. The actual draw at power on could be even higher when the PWM flank kicks in. But, I don't have any measured proof for that. Somebody who really understands how this type of switching supply works may be able to comment better on what PWM load spikes from an inductive load cause to happen because of to the unique way in which switching power supplies work (rather than, say, a linear power supply) and what this does to the line side in terms of voltage draw etc. PWM loads and constant loads are far from comparable.
...<snip>...
Measured efficiency on my Prusa supply is about 75% when the bed is heating. Note that DC power supplies are generally rated by their DC output load. So the Prusa supply is rated to output 10A at 24V. It will draw more that 2A on the input side due to being less that 100% efficient, as you noted. Reduced efficiency doesn't reduce the output, it increases the input -- at least on any reasonable supply where the manufacturer isn't outright lying to you about the power supply ratings.
Measured draw under full heating load was about 10.7A. I would expect this could vary as much as 20% since the hot end and, especially, the bed resistance are not high precision. So if I were on the lower end of the resistance range, a 20% variance would put the lower end of the power draw around 205W and his 215W would be close to the middle. Of course if I were in the mid to upper part of the resistance range, the someone in the lower end could be pulling >280W from the supply. This is one of the reasons to over spec the power supply. The other big one being longevity. The power guys I have worked with all generally use 50% derating as a rule of thumb for power components, though a switching power supply is a bit different in that you also want to run it as close to maximum efficiency as possible. For lots of supplies, that could be somewhere between 50% and 70% load.
You might want to start reading this thread: Power Supply Failure
I have a full set of measurement at 120V starting here: https://shop.prusa3d.com/forum/hardware-firmware-and-software-help-f64/power-supply-failure-t13443-s540.html#p110243 . Others provided 230V measurements. I also touch on the interaction between the LED lights in my hobby room and the Prusa supply.
Re: MK3 Power Supply Clicking
Thank you. Agree and that makes perfect sense.
Lights no longer flicker with Mk3 firmware 3.5.2!
In the release notes for the latest Mk3 firmware, v3.5.2 I noticed this item:
- Heatbed regulation improved - reduced the PSU clicking noise
So, I tried it this morning. when printing with PLA, I didn't notice the flickering lamps in the house I have seen every time before! Also, I wonder if the flickering is worse for some of us because of bulb type, I notice it most on a GE Reveal Halogen bulb in a lamp.