Flickering LEDs with 3d printer

[kuon]

I have a large 3d printer which has a 1600W bed heater, and that heater is connected to mains and controller by an SSR.

The issue is that when the printer is maintaining the bed temperature, LED lights in the house flickers. I suspect that is because the printer is doing PWM on the SSR with the heater.

Do you have any recommendation on what to do to remedy to this? I don't have a lot of experience with AC powers that high. I guess I could just put a double conversion UPS, but it seems overkill and is pricey.

I do not have a proper differential probe to measure main so I used two channels to measure neutral and phase, and I see 16V peak to peak variation when the heater is on which seems like a lot.

As a side note, isn't a machine supposed not do that?

[james_s]

That sort of thing is pretty common, loads like that can confuse the dimming circuitry in LED lights that are meant to work on regular phase control dimmers. The easiest solution may be to connect the printer to its own dedicated circuit, or at least a different circuit than the lights are on.

[kuon]

That sort of thing is pretty common, loads like that can confuse the dimming circuitry in LED lights that are meant to work on regular phase control dimmers. The easiest solution may be to connect the printer to its own dedicated circuit, or at least a different circuit than the lights are on.

My labs has only one circuit, so I'll have to add a new one from the main breaker room. So what you mean is that the voltage drop will be attenuated at the main breakers?

[Kim Christensen]

Sounds like cheapo LED lights... Do all the LED light flicker, or just those on the same circuit as the printer?
One fix might be to put an EMI filter on the feed to the 3D printer.

[james_s]

You could try non-dimmable LED lights, or smart bulbs that use something other than phase control for dimming. A low pass filter in the line to the printer may help too.

[thm_w]

What is the SSR model and what frequency is it set to switch at? 1600W is crazy for a bed heater.

Sounds like cheapo LED lights... Do all the LED light flicker, or just those on the same circuit as the printer?
One fix might be to put an EMI filter on the feed to the 3D printer.

Yeah, less than 5% variation in voltage and the lights flicker.
If the SSR is zero cross, then I would not expect to see any EMI issues. If its non-zero cross, and its randomly being fired, it could be?

[james_s]

That's certainly worth checking, a zero cross SSR might resolve the issue if it isn't already. I wouldn't trust what it says on it if it's a Chinese SSR.

[kuon]

The SSR is no brand one, I replaced it with a zero crossing panasonic one and the issue persists. The switching frequency seems to be 10Hz.

Yeah 1600W is a lot, because the bed is large.

For the LEDs, most of my LEDs do not flicker, only the one with E27 screw and those yellow LED string in it. All the flickering LEDs are OSRAM parathom, the rest is Philips.

But I am more interested in understanding the problem and fixing the machine than fixing the light

[Kim Christensen]

But I am more interested in understanding the problem and fixing the machine than fixing the light

What's the maximum continuous load permitted on that circuit? Because if you're already close to the limit, then wiring in a dedicated circuit is your best bet. It would suck if the breaker popped in the middle of a print because you decided to use the drillpress or something.

"fixing the machine" would probably take more effort that it's worth. You'd basically have to change the way it PWMs the resistive load. Right now, the "PWM" is at a really slow rate. It would be better to use a much higher PWM rate (100Khz), (have it well filtered so it appears to the mains like a slowly changing resistive load) but that would take much more sophisticated circuitry than what you have now. Maybe even a firmware change on the printer, though you could probably avoid that.

[james_s]

But I am more interested in understanding the problem and fixing the machine than fixing the light

LEDs that are made to work with phase control dimming are sensitive to the peak voltage, a large load will cause the voltage to sag and this can confuse the dimming circuitry. I suspect if you installed ordinary incandescent bulbs those would flicker too, though the thermal inertia of the filament would likely make it less noticeable.

[kuon]

But I am more interested in understanding the problem and fixing the machine than fixing the light

What's the maximum continuous load permitted on that circuit? Because if you're already close to the limit, then wiring in a dedicated circuit is your best bet. It would suck if the breaker popped in the middle of a print because you decided to use the drillpress or something.

"fixing the machine" would probably take more effort that it's worth. You'd basically have to change the way it PWMs the resistive load. Right now, the "PWM" is at a really slow rate. It would be better to use a much higher PWM rate (100Khz), (have it well filtered so it appears to the mains like a slowly changing resistive load) but that would take much more sophisticated circuitry than what you have now. Maybe even a firmware change on the printer, though you could probably avoid that.

I have 13A C breaker for my lab (lights and plugs). As for the machine, maybe I can add an EMI filter? But I guess I would have to properly measure the AC line to select a filter and I do not have the tools for that.

But wouldn't a zero crossing SSR be unable to switch at 100Khz? (unless the PWM is in sync with the AC waveform)


I realize now that the LEDs that flickers are "dimmable" and the non-flickering ones are not (I don't have any dimmer in my house), I was confused because at first I thought "why would a LED driver flickers" but now I get it, it's because of the phase control support.

[Kim Christensen]

You could simply buy a premade filter for your printer that's properly rated for your country. It would be a bit of an guess/experiment and it may or may not solve the problem. (Hard to tell how fussy the flickering LEDs are)

But wouldn't a zero crossing SSR be unable to switch at 100Khz?

You're correct. This wouldn't work. You'd have to totally redesign the heater control circuitry which isn't worth the effort.

I realize now that the LEDs that flickers are "dimmable" and the non-flickering ones are not (I don't have any dimmer in my house)

With this knowledge, the easiest solution is just to replace the dimmable LEDs with non-dimmable ones.

[kuon]

I'm going to try a pre-made filter and change the flickering LEDs.

Thanks for all your input, it's been very educative.

[thm_w]

I'm thinking the filter wouldn't do so much, if its due to voltage sag alone. Which it seems you've determined is the cause.

Some other ideas:
- reduce the maximum duty cycle, and power, of the heater, until the point that flicker is gone. But you'd end up with a slower heating bed which you likely don't want.
- play with the switching frequency, if the heater only switches on and off every 10s would that mean less noticeable flicker? Or is it still enough to be annoying

[pbgossett@gmail.com]

Curious if this has gotten any solution?

I've been having this issue different circuits, including off a subpanel from one another.

Debating either A: running a 240V dedicated branch circuit for the 3D printer or;
B: running a 120V to 240V transformer on the current 20 amp branch circuit in both cases flipping the power supply voltage switch.  Just trying to get "off phase" with the LEDs.

[kuon]

Curious if this has gotten any solution?

I've been having this issue different circuits, including off a subpanel from one another.

Debating either A: running a 240V dedicated branch circuit for the 3D printer or;
B: running a 120V to 240V transformer on the current 20 amp branch circuit in both cases flipping the power supply voltage switch.  Just trying to get "off phase" with the LEDs.

I've been living with it for now, the filter didn't do anything, which I kinda expected, my understanding is that it would require large capacitors with active electronic that monitor the AC signal and compensate it, a bit like solar inverters that follow AC signal sync.

I am re-wiring my labs with separate circuit with larger cables from the breaker room with dedicated circuits, I'll let you know how this goes as I am not finished yet and after that the electrician needs to certify the new breakers and measure everything before it can be turned on.

[pbgossett@gmail.com]

Will be curious on outcome. 

Ive also considered a single phase to single phase vfd.  My VFDs dont flicker my lights. 

https://www.ato.com/3hp-vfd-single-phase-input-output

Mine doesnt draw this many watts but yiu get the idea.