Heated bed temperature ALWAYS fluctuating (in idle too!)

klippy(4).log (3.0 MB)

Hi!

I have a very weird problem and at first of all: I found out, that it MUST be a software (settings?)-problem in klipper, because with marlin-stock-firmware from BTT on the exact same hardware-configuration the problem doesn’t exist!!

The heated bed temperature is always fluctuating about ±8°, even when it’s cold!

I have a very customized Creality CR-10s (ok, there is quite few left from it ^^):

Here are the modifications that maybe belongs to the problem:

750w-heatbed with NTC 100K thermistor Beta 3950K

Solid-State-Relay

the upper components are working for a year with original Creality-board and marlin.

APC UPS for emergency power-supply, because sometimes we have power-failures for a few seconds, but the power for the heated bed is NOT connected to it, because the UPS is to weak. And loss of bed-heating-power for a few seconds during printing is no problem.

Keep in mind that because of that I have a seperate power-cable for the heated bed!

And now I added a new Bigtreetech SKR E3 V3.0 with klipper 1st installed.

In printer.cfg I’ve set the heatbed sensor to “Generic 3950”, but it makes no change. Everything else is quite stock.

Now please look at the graph. The heated bed is always off:

You see that the graph is fluctuating in a weird way and in the right area the graph is stable. It is stable in this area, because I pluged off the power-cable from the heated bed!

Whaat? Why?

And remember: with stock-marlin-firmware on the Bigtreetech-board the problem isn’t existing.

So what can be the problem in this case? What can I try?
Please help!

thanks and kind regards!

Hello @BachManiac !

Is that a 2 wire cable that you use or a 3 wire cable with ground connected to the chassis/frame of the printer?

Thanks for your reply!

The 3-wire-cable goes only to the meanwell power-supply.
But now I’ve connected either the frame and also the solid state relay to the ground too, but nothing happened.

And the heated bed? What cable?

The heated bed is powered by the separate cable:

230V L------------Solid-state-relay-----------Heated-Bed---------------N

In this circuit there is no ground.

I assume you have a aluminium print bed.

You have to use a 3 wire cable for the heated bed and the ground attached to the bed.

If there is an issue with the heat mat and it gets connected to the bed, bad things will happen.

Also missing the fuse in your wiring.

1 Like

I found the problem!!!

While grounding the aluminium print plate I wondered why there were very small sparks beween plate and cable.

Then i measured with my multimeter and it shows 204V-AC between print-plate and ground, but I can grab it and grounding works too without triggering the fi-switch.

then I measured the solid-state-relay-switch and I realized that it was wired like this sheme:

N-----SSR-----Hotbed---------L

So I turned the power-cable around 180° so I have this scheme:

L-----SSR-----Hotbed---------N

And now I have no current between print-bed and ground and the temperature is also constant now!

graph1
It works! But Why??

And why is rotating the plug necessary on klipper and not on marlin? questions and questions…

What happens if you connect the bed thermistor to the port of the hotend thermistor and reconfigure the pin settings for testing purpose?
How is the thermistor cable routed inside the printer?
Is it along side other cables that might interfere here?
Or might other components affect this sensor signal?

This sounds like a serious issue. If you do not have the knowledge I would consult an electrician to check both your wiring and house electrical distribution.
230V AC is really not something to play with or take lightly.
I will not provide any further hints for what to look and I recommend that nobody else here does so as well. Playing with your own life and safety is your choice but not to any advice here.

1 Like