Low laser power output from Lightburn via MG3

Quote from Mal Rixon on April 10, 2025, 3:51 pmHi All,
Just setting up my MG3 on my K40 and I notice that from Lightburn with 100% power selected I only get a reading of 10mA tube current.
I have a toggle switch installed so I can operate the Power intensity either from the MG3 or from the front panel. From the front panel with 80% power selected I read about 16mA and the laser is definitely brighter.
I changed $30 value to double at 2000 and half at 500 and from Lightburn I still only get 10mA.
I have included the G code which is just a straight line generated by Lightburn
Any help greatly appreciated
Cheers Mal
Hi All,
Just setting up my MG3 on my K40 and I notice that from Lightburn with 100% power selected I only get a reading of 10mA tube current.
I have a toggle switch installed so I can operate the Power intensity either from the MG3 or from the front panel. From the front panel with 80% power selected I read about 16mA and the laser is definitely brighter.
I changed $30 value to double at 2000 and half at 500 and from Lightburn I still only get 10mA.
I have included the G code which is just a straight line generated by Lightburn
Any help greatly appreciated
Cheers Mal

Quote from Paul on April 11, 2025, 12:36 pmHi Mal,
The PWM output pin on the MG3 provides a 5V PWM signal. Basically its a pulse width variable signal. That signal drives an opto coupler on the K40. If its wired with a pot, the signal level might be not 5V anymore. I speculate that the potmeter is interfering with the 5V level of the PWM and pulling it down via the potmeter resistance. You can check that with a multimeter, read the voltage on the IN terminal, by sending Gcode commands: M4 S1000 which says full power. The laser button does not have to be pressed on (test without the laser actually firing). When in stand still mode you have to set $32=0 (cnc mode) since laser mode does not produce a PWM output signal when the gantry is not moving. Or just send G1 X10 Y10 F100 S1000 in laser mode ($32=1)
Try to isolate the cause by just wiring the PWM to the IN terminal of the Laser power supply without the potmeter and see if that works properly. The $30 and the S value in Lightburn needs to match to get matching power and percentage values in Lightburn's application console but I think you understand perfectly that from you comments above.
If the toggle switch is a single throw and wired in the middle of the pot (wiper of the pot) then that should in theory work. Maybe install a triple throw toggle switch that switches the three connections : wiper and gnd, 5V connection of the pot.
pls let me know if you got further with that.
Cheers, Paul
Hi Mal,
The PWM output pin on the MG3 provides a 5V PWM signal. Basically its a pulse width variable signal. That signal drives an opto coupler on the K40. If its wired with a pot, the signal level might be not 5V anymore. I speculate that the potmeter is interfering with the 5V level of the PWM and pulling it down via the potmeter resistance. You can check that with a multimeter, read the voltage on the IN terminal, by sending Gcode commands: M4 S1000 which says full power. The laser button does not have to be pressed on (test without the laser actually firing). When in stand still mode you have to set $32=0 (cnc mode) since laser mode does not produce a PWM output signal when the gantry is not moving. Or just send G1 X10 Y10 F100 S1000 in laser mode ($32=1)
Try to isolate the cause by just wiring the PWM to the IN terminal of the Laser power supply without the potmeter and see if that works properly. The $30 and the S value in Lightburn needs to match to get matching power and percentage values in Lightburn's application console but I think you understand perfectly that from you comments above.
If the toggle switch is a single throw and wired in the middle of the pot (wiper of the pot) then that should in theory work. Maybe install a triple throw toggle switch that switches the three connections : wiper and gnd, 5V connection of the pot.
pls let me know if you got further with that.
Cheers, Paul