Flickering lighting LED's when heat bed heats

dc42

@JTBrown said in Flickering lighting LED's when heat bed heats:

Hi all,
I've hooked up my lighting LED's to the E1 on my duet 2 wifi and control them using a slider on the Duet control. They all work great with loads of help from this forum, unfortunately I've now just noticed that they are flickering when the build plate heats up. Can anyone help? Cheers Jim

It sounds that the power voltage drops when the bed heater turns on. If the dtop is happening at the PSU terminals, maybe you need a better PSU. But if it is happening in the wires between the Duet and the PSU, then you could use thicker wires, or use an external mosfet to control the bed heater. so that you can use completely separate wires from the PSU.

Increasing the PWM frequency as @Phaedrux suggested should make the flicker invisible. Try between 100 and 500Hz.

JTBrown

hi, I'm using version 2.05.01 firmware

JTBrown

Do you have any recommendations on a better PSU? I'm using the meanwell GST220A24-R7B power adapter that came with the Ultimaker 2+. Thanks

A Former User

GST220A24-R7B

thats a 220w PSU, MeanWell is generally a solid brand, if not cloned, it might just be under-dimensioned?

not sure how much the ultimatker2+ needs power wise, but if its below 220w then odds are dc42's suggestions will sort you out.

plan b could also be using an adittionalpsu dedicated for the bed instead of replacing it, but you might be constrained on space.

(hmm external power brick, not the easiest to rewire)

JTBrown

@Phaedrux Tried what you suggested for the PWM frequency and the PSU started buzzing, is this right?

JTBrown

Does the board use more power than the original Ultimaker board because that's all I have changed?

A Former User

@JTBrown said in Flickering lighting LED's when heat bed heats:

Does the board use more power than the original Ultimaker board because that's all I have changed?

could be, but very unlikely on a scale to cause any problem. more likely the original firmware ran at a even higher frequency. the sound is likely to increase with frequency, peak, and be reduced as the frequency increases even more.

JTBrown

I've just taken it to 600 and it still buzzes. The flickering stopped though so that's a start. Think I'll just keep the LED's turned off

A Former User

@JTBrown said in Flickering lighting LED's when heat bed heats:

I've just taken it to 600

i'll defer to higher powers for anything to do with black magic and inductors, but unless the psu is failing ultimaker would have had to run at a higher frequency or have had a larger capacitor on the board.

JTBrown

@dc42 could you suggest an external mosfet that’d work well with my board or they all pretty much the same? Thanks

dc42

An external mosfet won't help if the flickering issue is caused by the PSU. It will only help if the issue is the resistance of the wires between the Duet and the PSU.

I suggest you measure the voltage at the PSU terminals, and at the Duet VIN terminals. Do this both with the bed off and with it on continuously (i.e. heating up). If you tell us those 4 values then we can advise on the best way forward.

JTBrown

Unable to measure at the Power adapter terminals because it uses a R7B connector but at the VIN terminals it was 24.2 volts with the heater off and 23.83 volts with the heater on continuously. One thing I did notice was that the lighting LEDs flickered and the power adapter buzzed together when the temp reached 50°C. Also, because I had the printer on it's side, I could see that there was a red LED flickering next to the VIN to Bed fuse.

JTBrown

Has anyone got an idea as to what is going on please?

T3P3Tony

@JTBrown said in Flickering lighting LED's when heat bed heats:

red LED flickering next to the VIN to Bed fuse

That is the LED showing the bed heater is turning on and off (which is normal with it set to PWM)

The voltage drop you indicated (from 24.2 to 23.8) should not be enough to make the LEDs flicker.

Can you try setting the Bed heater to "Bang Bang" for now and see they flicker as the heater turns on and off ( set B1 in the M307 command for that heater)

dc42

@JTBrown said in Flickering lighting LED's when heat bed heats:

I've just taken it to 600 and it still buzzes. The flickering stopped though so that's a start. Think I'll just keep the LED's turned off

The buzzing is the PSU responding to the change in load at the PWM frequency. It's probably harmless.

JTBrown

@T3P3Tony Thanks that done the trick. Could you explain why? many thanks Jim

mendenmh

I suspect the improvement in behavior with increasing PWM frequency has to do with stored energy in the filter capacitors. You may be very close to the limits of your power supply, ad it cannot provide 100% to the bed. At low PWM, you are fully depleting the filter capacitors, and requiring the power supply to ante up a full-power pulse during the 'on' phase. At higher frequency, the filter capacitors are supplying enough energy, and the power supply only sees the average load. It is likely to work much better with a bigger power supply.

JTBrown

@mendenmh Hi, thanks for your input, it was the "setting the bed heater to Bang Bang" suggestion that worked. cheers

mendenmh

@JTBrown That is moving to the other extreme, in which the power supply's feedback loop probably has time to catch up and keep the voltage constant. If the supply has enough oomph, it may be ok. The observed behavior, though, was a sign you might be near the edge. The other possibility is that the supply is badly designed, so the low-frequency behavior of the regulation doesn't cross over nicely with the filtering of the capacitance. Incidentally, I had this problem, exhibited differently, and very annoyingly. For many months, all the LED lights in my house flickered annoyingly. It took a long time to associate it with the 3d printer. Somehow, the PWM frequency for the bed, which I had at one point set to 30 Hz, had gotten trampled back to the default 10 Hz. It takes only a tiny bit of line voltage variation at 10 Hz to crate enough flicker for the eye to see. It's pretty much at the top of the frequency response. So the whole house line voltage was being microscopically modulated, resulting in a drool-inducing flicker.