0
\$\begingroup\$

This question is regarding a 3D printer, specifically its heat bed. The head bed is a PCB heater with a resistance of 0.9 Ohms, so it's drawing around 13 Amps when supplied with 12 Volts.

I recently switched from Bang-Bang to PID controlled bed heating. The stock firmware pulsed On/Off with about 5 seconds each cycle. Now with PID, my bed keeps its temperature much more stable, the frequency is around 7Hz.

Although the total amount of energy used for heating should be less now since the heating is more efficient and doesn't waste energy by cooling down, I am a bit concerned of the fast pulsing of the rather high amperage.

So I was wondering, do the faster cycles cause more stress on the components (I am thinking MOSFET on the board and components inside the PSU) or is this negligible? I am particularly curious since this is a rather low budget printer (Anycubic i3 Mega, $350 printer with a generic unbranded PSU) and I don't want to push its limits too much. I suppose the power supply is a switching mode one. Are those safe to pulse like this at fairly high amps? The PSU is rated for 25A.

The only thing I can observe is a very faint noise with the same frequency as the pulses. Might have been the same on 5 second cycles, I was just not paying attention to it back then.

I'd be very glad if somebody could chime in on this.

Thanks in advance.

\$\endgroup\$
3
\$\begingroup\$

So I was wondering, do the faster cycles cause more stress on the components

Probably not. In fact, switching faster may stress component less.

Rationale: when switching slow, components have time to heat up and cool down on each cycle, which causes thermal expansion and contraction cycles and mechanical stress. Switching much faster keeps components at a constant (average) temperature and avoids thermal cycling.

7Hz is still quite slow, you'll still have some thermal cycling, but less than with the bang-bang regulation.

The noise you hear could either be due to the power supply and transformer's noise being modulated by the 7Hz PWM, or a ceramic capacitor acting like a piezo speaker, or simply from the wires. Variable current in wires causes a magnetic field which can make move very slightly. Since your current is pretty high this could be audible. Nothing to worry about though.

\$\endgroup\$
0
\$\begingroup\$

Although the total amount of energy used for heating should be less now since the heating is more efficient and doesn't waste energy by cooling down, I am a bit concerned of the fast pulsing of the rather high amperage.

It's probably about the same. The heater cools down and heats up for about the same time as before - it's just lots of short bursts instead of less long bursts.

So I was wondering, do the faster cycles cause more stress on the components (I am thinking MOSFET on the board and components inside the PSU) or is this negligible?

It is possible, yes. Switching a MOSFET wastes a little bit of energy while it switches, which heats up the MOSFET. Since you're turning it on and off more often, it will get hotter. Possibly it wasn't very hot to begin with and it's still fine now, or possibly it's undersized and gets quite hot (you say this was a cheap machine). You could check the temperature.

I think that having a more constant temperature puts slightly less stress on the heater itself.

The PSU may create a voltage spike when the load turns off and a voltage dip when it turns on. I suppose the spikes occurring more frequently could shorten the life of components. You could watch the PSU's output voltage on an oscilloscope - use a fast timebase and triggering to see how the voltage behaves right after the heater turns on or off.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the in-depth reply! I don't have an oscilloscope at hand but I think the voltage drop is not too big since that would cause a visible decline in print quality from lost steps on the motors and on the print head itself. I will definitely keep an eye on the MOSFETs temperature. The one used on the board is a WSK220N04, I couldn't really figure out if it's proportioned large enough, maybe you can have a quick glance at its data sheet? Cheers. \$\endgroup\$ – David Ramiro Dec 18 '18 at 11:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.