My K40 laser cutter was working perfectly well.
Then suddenly, about 4 months after buying it, and only few hundreds of hours of service, cutting power decreased a lot. It’s now impossible to cut 3mm plywood even at 10mA with 2 passes.
I got this image on a piece of plywood placed just in front of tube exit (in front of mirror #1
Do you confirm it’s a dead tube working in TEM01 mode ?
What could be the cause ?
I have to admit that I put tap water in my cooling system with some bleach to avoid algae.
And I turned also my tube 90° to facilitate bubble elimination and at the beginning some electrical arcs where present between anode and K40 as the distance was shorten a lot due to rotation.
If not the tube, could it be due to the K40 power supply ? I read on other streams that should not be linked to PSU.
I have analog ammeter installed on the machine.
Before tube death, I was able to cut 3mm plywood at 10mA (mesured in mA) in one pass
After, I’m just able to engrave the plywood, it’s almost impossible to go through it completely
So, in your opinion, it’s a dead tube ? Or could it be linked to the PSU/LPS ?
Unfortunately I don’t know what is weird sound
From the beginning, especially at low power, there was a high frequency rattling noise (like air ionization, I don’t know how to explain). This sound was vanishing at higher power.
No need to say that I never went above 10mA
I was wondering if electrical arcs when I turned the tube could have led to LPS breakdown …
New tube if on its way from China … Hope to receive it by end of May.
First, let me add some color to the answers previously provided:
I can be one or the other or both. The output mode you have shown makes the tube the primary suspect.
A bad LPS will not usually destroy a tube or cause it to jump modes. Mostly you just get low performance.
It’s very difficult to safely determine which is bad a tube or LPS. I usually evaluate the health of each component and then decide which part to buy first. Keep in mind whatever you replace, even if it turns out to be the wrong part, will not be wasted as these are consumable parts… i.e. they both eventually will fail.
Tap water over time will damage a LPS depending on how conductive it is.
Repeated arcing can damage a LPS. Usually arching from the anode means that the tube is dead. Rotating a tube to where the water jacket inlet is directed upward is not usually a problem unless it is very close to a ground plane or has poor anode insulation.
Note that your tube is not dead as much as it has shifted modes. Mode shift is usually caused by damage from overheating which changes the physical characteristics of the tube’s optics.
A hissing sound (from LPS) is common at low powers at or below 4ma. It will stop as the current is increased. At higher powers, you can tell if the LPS is arcing internally as the ma meter readings will bounce with the arcing.
Conductive water will also cause a hissing noise and/or a hot bucket.
From the output behavior of your tube, I would guess it is bad. Was it overheated?
The LPS also could have failed from arching or from conductive water.
Try this test while you are waiting for the tube to arrive.
Set the current at 20 ma [that may be about 70% if you have a digital panel] hold the fire button long enough to read the current on the meter and note if the meter is jumping (from 0 to its setting). LPS’s whose HVT’s are failing will hiss and the meter will not be steady.
Consider adding these important upgrades:
Safety interlocks on all covers… I know you will tell me you have laser glasses… not good enough
Add temperature and coolant flow sensors into the LPS interlock circuit.
If you need help adding safety for you and your components let us know and we will help.
First, thank you very much for this very detailed answer.
Never. i have added from the beginning a water flow indicator with temperature meter integrated. My water temperature is always between 16°C & 18°C
So in your opinion, tube won’t fail because of conductive water or arching. It’s more LPS that could fail.
I will ! But old tube is already on my desk …
I need some time to install it again with proper anode insulation before testing.
What I can say is that at 10mA, power on my analog ammeter was very stable.
I will try at 20mA
As far as I have well located the sound, hissing noise was coming from the tube, not from the LPS but I can be wrong on that.
I experienced this noise below 6mA in general
I experienced arching as soon as I have turned the tube. I didn’t managed immediately to understand from where was coming the sound because of insulation of anode that seemed quite good to me. When I unmounted the tube yesterday, I saw 1cm of Silicon above the top of the anode post. Compared to the video from the other thread on this forum, it seemed not so bad to me.
After adding a piece of 3mm plywood between anode and ground plane, no more arching
Other thing I didn’t mentioned earlier.
I have developed a shield for ESP32 dev board in order to use lightburn + GRBL.
This shield has a voltage shifter for the 3.3 PWM coming out of the ESP32 to 5V to go in the IN input of the LPS.
PCB has been manufacturing by an EMS and I’m using a Texas TXS0101 for voltage shift.
It seems working well, with ability to manage everything from my computer (movement, speed, power, air assist, safety sensors, …)
I have set PWM frequency to 15khz (I saw yesterday that recommendation is more between 20khz and 50khz) and I have calibrated my power on lightburn with my analog ammeter.
However, yesterday, I asked myself if this shield could have damaged the tube or the LPS.
So I have checked the PWM signal with a scope.
When my shield is NOT attached to the LPS, PWM signal is perfect.
When my shield IS connected to the LPS, I can see a peak of 5V signal then 3.3V for the remaining 99% of the high state.
Assuming that the LPS is just “interpretating” an average DC value out of the PWM signal, I assume it’s not a big deal.
Do you have other opinion ?
I just made the test.
Power on the ammeter set to 20mA for 3sec → No noticeable change or movement of the needle
However, very high frequency hissling noise was comming out the machine (probably from tube)
You are one of the most carefully prepared new forum members. Kudos on having done so many of these fixes that the K40 vendors leave off. Very few people have installed interlocks as well as flow and temp sensors before arriving at Maker Forums. This is why it has become a reflex to suggest adding them.
Your English is very clear. (My small amount of French is very much less so.) I don’t know what “EMS” means here and would like to learn.
If I’m reading this right, the PWM in has a current sink within the TL494 PWM driver (I don’t know how much current) which would make sense for your logic level shifter not not actually providing 5V at whatever current it is drawing.
I cannot imagine a way that your shield could damage the LPS.
That shows the loaded and unloaded state of the driver. Those drivers are not really intended to drive much load.
The simple answer is that that signal is filtered at the input to the control amp which creates a voltage (integrator) that controls the LPS internal PWM. If you have calibrated your ESP PWM for power then you should be ok.
That said, I think driving the L pin with an open drain driver is an easier approach and it leaves the opportunity to add a pot to IN for local control. There is an extensive explanation on my blog why I chose this approach. Adding this driver to a esp32 is not such a big deal.
That sounds like a good power supply to me … although it’s still possible that something else is weird with the LPS We will find out after you install the new tube.
One thing we never asked for on the failed tube, which would be interesting, is a video of the tube firing so we can see what the plasma looks like. No big deal now that you have a new one. Just still wondering what caused the tube failure.
Also, at night you can turn out the lights and fire the tube to discover where the hissing is coming from.
It may be coming from inside the LPS and not visible but sounds like it’s the tube.
Of course, it could be that the tube had a manufacturing defect from the start…