Update and bad news, after a complete flush of the cooling system, having removed all the HV wires (and replaced the cathode wire) and the cooling lines from the machine to have them isolated from each other, NOTHING has change, still get 9/10 an unstable beam like my video above… what’s next?!?!
Looking at what you’ve posted I’d have to remind you how one of these work.
Why it’s intermittently arcing, my little gray cells are still working on that, but I have suspicions.
A laser power supply or lps starts when the laser is enabled, at least on glass tube machines. This initial enable signal will allow the lps to raise it’s voltage, quickly to the trigger voltage, fast enough to do this in around 1ms reaching 90% of the placard voltage. The 150W tubes trigger voltage is about 19kv if I recall.
I keeps increasing until the tube conducts. Once it conducts the current is maintained as set by the IN signal (analog or digital.) If it arcs, then it will draw the maximum set current from the lps.
I know that because the maximum limit on a 150W Reci tube is about 28ma. If you hunt around on the Internet you’ll find the maximum goes from about 27ma to about 29ma. When you load it you can push it to 34ma, the current limiting of the lps should have blocked anything over it’s set maximum.
This lps is typical of most lps, it isn’t setup for it’s controller and tube properly or these current levels would be in much better control
It would be nice to determine why it’s appears to be arcing to the metal end cap. You’d think this would be insulated by the coolant and glass. It must be completing the path somehow or it would not arc or draw current.. or illuminate.
The other option is it’s going through the coolant. This should not occur with normal coolants as they don’t conduct.
I think if we had a fixed issue, such as a consistent arc, we could work on it from that angle. Since you’ve change the lps/tube combination, I can only guess you have a bad tube. The reason is that it’s intermittent and something intermittent is more likely to be somewhere is resonating, mechanical or other issue that can change dynamically.
Do you still have the original tube/lps combination? You could always swap out the tube or lps to see if you got some bad part.
Unfortunately, this area is rather void of information. We have an idea on how the lps works but virtually no schematics. I’ve seen one, it gives you an idea on what they are doing. Since this area is a hv area, many people just stay away from it. It’s also difficult to measure. I put a 600m ohm resistor in mine and a 50ua meter to watch the high voltage.
At this point all I can advise is to zero in on the arc itself. If I recall, you advised that even when it illuminates it may not be providing an emf output?
1 Like
Unfortunately, I now have my old RECI W8 and two brand new MWCLaser 150W, my old power supply, a brand new power supply, tried all the different combination and the problem remains. When using the W8, the result is different, the tube is less noisy, but still unstable, with visible beam leaks around the cathode.
I just played on the power supply, please tell me if it’s normal behavior…
When I press the test button and turn the knob on the ampmeter, it always indicates 0mA, but I notice a small (almost unvisible) beam close to the cathode only, and only when the know is mid range, low range and high range produce no beam at all:
When I disconnect the 6 pins connector, I can now produce a full visible beam, ranging from 0mA to 5mA when playing with the knob, but the power supply is really noisy:
Sometime, always when the 6 pins connector is plugged, the the ampmeter get stuck at 01mA even if I press nothing (see laser signal OFF and water protection OFF even if it is ON)
Sometime, the red light of the LASER stay ON (but dimmed) even if there’s no signal, and it turn OFF if I unplug the 6 pins connector:
Have you tried it with the meter unplugged?
Many people use an inline 30ma meter for that.
Do you have a scope?
I’ll try to unplug it tomorrow!
Yes I have a scope, what kind of test do you have in mind?
Pretty simply, if we eliminate the lps/tube as a probably cause, it’s wise to look at the input signals that drive the lps.
Dsp processors produce a pwm signal the entire time they execute a layer. So whatever you power settings are, it should be reflected in these signals.
The L input or L-On1 is the laser enable signal out of the Ruida and initiates the lase operation.
If you do a line, say 20mm and run it at 1mm/s, it will take 20 seconds for it to complete, giving you time to look at things.
The L input should always remain low if it’s lasing. The pwm should not change.
I’m not sure where elsle to look. It seems you did all the right things, so, while grasping for straws, maybe if we back up and see if the lps is following the commands of the controller.
Make sense?
As a suggestion, if you’re having issues stay away from high power.
hmm wait how did you have the controller set during that test?
see, i think it’s best to use the little potentiometer box to dial in 23ma @ 100% power. but at 50% power it will read 11ma.
did you tie the laser fire command high or low, or use the laser controller with a 100% power test fire? Is the controller’s configuration set to be able to do 100% max power?
if you set the pot box to go to 46ma@100% but in the ruida it limited it to 50% duty, that would read 23ma average. but what it’s actually doing is not well defined and could be different from model to model.
the meter reads average current only. it could be jumping from as much as 46ma to 0ma and read 23ma. that could cause arcing
I doubt that’s the case, I suspected the slow ~1ms response time of an HVDC excited laser was actually due to the response in the LPS, there’s a flyback capacitor, diode, and then there should be a capacitor. But I have never tried to get a high voltage divider probe and scope the actual high voltage response. And again what it does would be specific to the type of LPS.
You should never, ever, debug or test a machine at full power. This is asking for it.
What we’ve found so far is this pot just lowers the overall pwm (or analog) voltage signal. In simple terms, the wiper of the pot is connected to the IN input. With it’s one side is the input pwm, the other side is ground. This compresses your pwm to a lower voltage pwm signal. I believe it’s best, if you have one of these, to set it at 100% and adjust the lps to produce a current limit that you feel is good for your system.
Dealing with rms waveforms is not required for a glass tube co2 machine, as they use it as an analog device. The analog device/interface is not cycled on/off like the pwm on a diode or RF machine. The only way to limit the Ruida is to set the power level to 50% on the console before you pulse it.
No matter how you do it, they read the same, but for different reasons.
Anyway, I don’t suggest adjusting power this way. If you’re lps is setup correctly, this is unnecessary.
Most of the time, these machines are driven directly from the pwm to the IN of the lps. It’s worked this way for decades, without operators controlling laser power in the process.
Meters display the RMS value of it’s input waveform.
With a glass tube co2, it’s an analog device and handled that way, so there is no real waveform to deal with.
That’s why I suggested they just try it without the pot plugged in and the controller set to 50% power for testing.
Diodes and RF machines are digital and driven directly off the pwm, glass tubes are analog and may not be on or off.
Does any of this make sense?
I apologize for the intrusion, I’ve already realized this is a somewhat complicated and unusual problem, although I don’t have any experience with CO2 machines, I do have some experience using measuring devices.
This leads me to the following thought: What if one (or more) measuring devices are giving incorrect values and we are all relying on those values? 
What kind of resistance are you measuring from the earth pin on the mains plug to the FG terminal, and also from the earth pin to the cathode connection on the tube?
I cannot explain it but I read:
FG-ground: 9 to 12 ohms (doing back and forth)
Cathode-ground: 55 ohms
But the same reading if I unplug the 6 pins connector of the power supply:
FG-ground: 3 ohms
Cathode-ground: 55 ohms
With the controller at 50%, I get 27mA when the dial knob is at maximum
Unplugging the ampmeter change the sound of the laser tube but the problem remains.
Ok WOW, I might have found something, not the solution nor the root cause but it’s the first time I’m able to produce only GOOD and STABLE beam!!!
First, I bypassed completely the controller and jumped the IN and the 5V connector of the PSU:
Afterward, I noticed that at full dial on the ampmeter (40mA), my beam is perfect, I can even lower the amp and it remains stable. Same result as long I start the beam above about 30mA. If I try the same experience below 30mA, I get back my unstable beam…
It makes sense what you’re saying. but it’s at odds with what I’ve seen.
The older RDC6332 can set laser-out analog or PWM digital, before any secondary modulation for content. I’ve not used the analog laser-out functionality myself, I just see it exists in the manual.
The newer RDC644XG dropped the analog function entirely. Since it appears to be meant to supersede the 6332, leaving out that function implies to me there that wasn’t a good feature to be using in the first place. The bar is pretty high to cut an existing function out like that and alienating existing users, which should have been fairly easy to include as an option.
The control I saw is purely PWM to the LPS. The LPS pot is very necessary to establish what 100% power means. I saw a spec of “1ms response time”, and testing shows that it basically the case for getting a full turn-on and turn-off of the tube output. But 1ms to turn on completely/1ms to turn off completely, 500Hz, is not a cutoff frequency per se, it’s the freq where beam power stops being a trapezoidal shape, and transitions to more of a triangle or sine shape The beam power does smooth out entirely to a static analog level at about 2-4KHz. But operating from a PWM in this higher freq range creates beam instability on the low end near the “min power” point and pushes min power threshold higher.
OK, maybe I went off on a slight tangent than the original. I’ve not actually measured this but I assume the pot box works as you say- the LPS input isn’t TTL digital switching, but an analog amplifier that processes an analog voltage. which will be modulated on/off by the controller’s pwm power modulation for the “%power” fields and power adjustment when not at the commanded speed due to acceleration limits, and modulated on top of that by the controller’s content modulation (dithering).
Why I think this is a bad idea:
One of the reasons higher current degrades the tube is electrode sputtering, a highly nonlinear effect that happens with instantaneous peak current, not average. The mfg “long life” current of 23mA is going to be below the knee where sputtering rate drastically accelerates.
E.g., using a pwm that alternates between 33mA and 13mA gives an average current of 23mA, but half the time the electrode is above the knee and sputtering at a high rate. This will silently accelerate tube degradation the whole time.
It also relies on limiting the PWM Max Power in the controller and/or Lightburn, which produces all sorts of potential artifacts and ways to accidentally exceed the limits. Those methods seem like a bad hack
What do you mean by no analog outputs on the 644x series? As far as I know you can unplug a 644x type of controller and stick in a different type of controller like the 6445 type without an wiring changes..
I think the analog output will stay with us a while. I see a number of manufacturers using analog signals to drive the lps.
From the 644x manual, the analog outputs are still there or I’m not following you.
A lot of what you say just indicates the controller/lps combination isn’t set up properly. I you limit the current from the lps to 100% of your tubes current level, how can you overdrive it?
Do you actually own/use a glass tube co2? I know you have an RF machine.
Do you mean the dash, operator controlled lps pot or it’s internal limit pot?
I’ve run mine almost 6 years without a dashboard control.
Glass tube machines users are advised to use a 50us period (20kHz) for the pwm.
Of course, an RF machine can’t use the analog output as it’s on or off, you can’t run an 50% power all the time, like an glass tube can.
I wonder what you think of the original K40 on how they operate, via the pwm.
The laser tube and PSU appear capable of working correctly together. When I bypass the machine controller and locally force the PSU input by connecting 5V to IN, I can get a perfect, silent and stable beam, especially when starting above approximately 30 mA. However, when the machine controller is connected normally, the beam becomes noisy/unstable at partial power.
At 90% power in LightBurn or RDWorks, the beam is silent and stable. At 50%, the beam becomes noisy again. Also, if Max Power is set to 90% but Min Power remains at 50%, the beam sometimes starts correctly but sometimes still drops out. This suggests the issue is likely related to how the controller modulates the PSU power input at partial power, rather than a direct tube or PSU failure.
My next step is to scope IN-to-G and L-to-G at 50% and 90% to see whether the controller is sending analog voltage, PWM, or unstable firing signals.
You can look at the controllers wiring and see if it’s analog or digital. CN5-5 is the analog output and CN5-3 is the pwm. It makes little difference what you use, as either, will be converted to an analog voltage within the lps.
What you’re doing at these high current levels is a destructive operation. There’s a guy down under at MW laser that I’ve recommended to people with insane issues that we can’t resolve here, he’s commercial, but has a great knowledge of these. It may be worth hiring him to fix your machine. We seem to be going around in circles currently.
Since you did all I can think of swapping stuff around, it might be worth the money to have him take a look.
If you’re looking at the pwm, you can trigger off L of the lps as it lases when L goes low.