Losing Power randomly over Time

[Jonathan Bowen]

I have a Hurrican Laser Agnes. Their Tech support doesn't have any ideas other then throwing expensive parts at the machine. The problem is it's randomly losing power late in the cut files. It's not focus, lens, or anything like that because it's random on the table. Sometimes it doesn't have a problem at all. The only variable that seems consistent is time. I don't have problems early in the morning when I first start running the machine. If I run it for about an hour then cut a longer file (20 mins+) then I get perforated cuts and areas that don't cut through at all. This problem has plagued the machine since we got it over a year ago. About 4 months ago we replaced the tube and that seemed to fix it. Now the problem is creeping in again. Our laser cutter load has been increasing steadily for some time now. The laser can run anywhere from 1 hour to 8 hours a day. It seems to show up around 2-3 hours in to regular use. I have yet to get a consistent repeatable case. I've run the same files 2-4 times in a row and they all have problems but in different areas. I've cleaned, I've focused, I've leveled the table, and I'm at a loss. As a side note: There is a communication error that happens between the controller and Laser Cut 5.3. It seems to hang on the compile stage and then pop an error that says something about not being able to communicate with the controller. It's annoying and has persisted across 2 computer installs. I did a complete wipe and window reinstall between so it could be evidence of a controller malfunction.

I have several theories. Those are mostly based on educated guesses from my electrical engineering background. A power supply could be under volting the controller and it's heat based. From what the techs at Hurricane say, The controller outputs 0-5 volts to the power supply for the laser. Higher the voltage the more wattage. So an under volt condition from a failing MOSFET or something could cause power fluctuations. Besides, The first step to trouble shooting as always check the power :) I may test this by swapping around the power supplies and test this theory. I'd test the controller output vs power input if I had an O-scope but I don't have access to one right now.

Anyone have experience with problems like this and can shed some light?

Here are some pictures of the problem files.

https://dl.dropboxusercontent.com/u/61862/2014-01-10%2021.37.37.jpg

https://dl.dropboxusercontent.com/u/61862/2014-01-11%2014.39.27.jpg

https://dl.dropboxusercontent.com/u/61862/2014-01-11%2014.39.36.jpg

here is the error message from Laser Cut

https://dl.dropboxusercontent.com/u/61862/error.bmp

[Dan Hintz]

This problem has plagued the machine since we got it over a year ago.

The owner of Hurricane was just here a few days ago swearing how great his business was, tech support, etc. If this has been a problem since you got it, my response to them would be "If you can't fix it with parts, I need a complete machine replacement".

Random power like that is usually attributed to either the tube or the power supply failing. Since this is a DC tube, you have a better shot at determining if it's the tube by looking at it during lasing... do you see any abnormal arcing, plasma in one end of the tube alone, etc.? On a side note, I no longer see the Agnes listed on their website... could it be this is a problematic machine they no longer carry? I looked for it to find the power level... check the power supply to make sure it's rated for the size of tube you're running.

A failing power supply could also lead to communications problems... bad USB comms,etc. Something to look at.

[Ralph Y Thorne]

More details would help.

What is the power rating of the tube?
How are you cooling the tube?
What is the ambient temperature where you're using the laser?
Do you have a Ma meter on the machine?

The second picture shows the material not being cut on one side of the sheet. Is that side of the material on the near or far side of the laser source?

Then again, as Dan says why is your model no longer listed listed on their website?

[Jonathan Bowen]

In their defense, They have been working with us a lot but we just kinda got a lemon of a machine. We have had to do a lot off odd stuff to get this working right. As a new owner, some of it is my inexperience. The lens wasn't in focus for the first 3-4 months. We had to shim the second mirror in order to hit true focus. I of course didn't realize it because I didn't know that you could make the laser focus on one point that didn't move. I was just trying to get it into the center as best as possible. We also had the cheap air compressor go out just outside of warranty and after a small fire. The fan that cools the electronics died recently and we replaced the tube as I mentioned but the other one might be ok. We kept it. Edit: We also had a stepper motor "fail" early on. We replaced it but it turned out to be a bad setting on the stepper driver.

The Agnes was discontinued shortly after we bought it. The Lisa replaced it and is basically the same machine. We did recently add some relays and linked them to one of the power supplies. It shouldn't be the one that connects to the controller. I'll have to swap them and see if it helps.

BTW Hows that Trotec working for you? That is the model I want to upgrade to. I'd also like to pick up one of those UScutters. I have not justified it yet but thats about to change I think. If sales don't drop in the next 2-3 months then I might go ahead and look into financing for the Trotec. I don't really need it yet but I also can't live with this much down time fiddling with cheap chinese crap all day.

[Jonathan Bowen]

It's a 40 Watt tube. Cooled by the water chiller that came with it. The laser is inside and usually room temp (around 70F) That sheet in the picture is didn't cut on the side farthest from the source. However, that file does cut from left to right and the side that didn't cut was the last to be done. No other file has done that. It appears random but is always in areas cut towards the end of the file.

[Ralph Y Thorne]

You need to know if the water temperature is rising slowly as you cut. The tube will lose power as it gets hotter. This might explain why you are losing power late into a file. You also really need a meter to know where you're at power wise.

Additionally your beam alignment may out enough to cause a loss of cutting power farthest from the laser source. Ray at Rabbit has a good tutorial on this.

These things may sound small, but they are enough to cause the problems you're seeing when working with a small output tube.

[Dan Hintz]

You need to know if the water temperature is rising slowly as you cut. The tube will lose power as it gets hotter. This might explain why you are losing power late into a file. You also really need a meter to know where you're at power wise.

Temp tracking is useful, but its effects wouldn't be so random (cut in/out), he would just slowly lose power. At this point, if the tube has been swapped with the same issue, I'd lean towards the PS being flaky (especially if it's the typical 40W-capable supply, which is the low end of the feeding pool). A failing PS would also show the effect of failing to cut near the end of a file... as it heats up, it simply can't supply the HV necessary to keep the tube lasing. I'm not convinced it's a location-based problem as you have one part of a circle cut just fine right next to an area that is not even singed.

[Kev Williams]

The material in the pics looks to be some pretty thin stuff, could it be heat-warping to the point the lens is out of focus? Happens to me all the time-- also, the air movement from the fan will lift lightweight parts at times...

[Dan Hintz]

Looks like 1/8" acrylic to me...

[Jonathan Bowen]

Yes it's all 1/8" acrylic. No heat warping. Also of note: the pictures are off the back sides of the material. It cuts most of the way through the material. The power reduction is very small and just enough to stop it from cutting through the material. I have no clue why a direct current tube would pulsate like it does. It's basically perforating the material. Sometimes it drops enough to not cut through the paper only. The edges are rougher as well. It's usually rather smooth but the areas that loss power have more prominent ridges. I'm assuming this is due to the power reduction.

I cut a file late last night that was a puzzle for a friends kid. 2 hours cut file and it came out good. 1/8" MDF with a large etched area. Took over 2 hours of mostly etching The cuts were clean but it did leave some small pieces so it wasn't a true clean cut. They parts just fall apart but they are not cut through in small slivers randomly through the part. I did this with the machine cold.

I'll start recording start and stop temps for the water. I'm usually under 30C.

I started with focusing the machine and checking all the usual suspects. I just checked the head. The beam isn't perfectly centered but it's not hitting the sides. It needs to move in the Y direction a fraction of an inch so I might do that the next time I pull the machine out from the wall. If I remember correctly that was changed by adjusting the tube angle. I might not mess with it because it's not causing a power loss it's just not perfect.

I need to do a few production runs and I can probably stock up extra in a particular item so I'm going to run the same cut file with a cold machine over and over till it fails and see what fails and where but also record the water temp and times.

@Dan are you referring to the high power PS? The machine has 3 or 4 smaller ones and I had planned to do swap the one that feeds the controller with another one of the same type that feeds something else. I'd have to see when I opened it up again. If it is the high power one then is there support for those? Can I get schematics and stuff for them? I'd much rather try and fix it then just throw a new one in there. I suspect it's rather pricey.

[Jonathan Bowen]

So a quick update. I was testing a grid of 15mm squares and looking for something odd. Also testing to see if the problem accelerated as time went on. I took a short video that is uploading to youtube right now so I can post it here and send it to the manufacturer. It should be live in a bit when the processing is done. For a while now, I have seen small burn marks on the bottom of my parts. I attributed it to flashback from the table but I think they are something else. The video shows a burst and an odd noise about 3 times. Square 2, 7, and 8 all had one that I saw. The beam is also fluctuating a bit as it cuts but that could just be varying reflection from the material (which is 3mm mdf). These little square shouldn't be attached to the sheet after cutting but they are all held on by little slivers of material at random spots. I just triple checked the mirror alignment and table height. Hopefully this sheds some light on the subject. I might move it out and prepare to start doing more exploration in the back of the machine tomorrow. If it is the tube then I can probably get another one out of Hurricane. If not then I might have to pitch a fit about them selling me a lemon and see if that gets me somewhere. Every time I call it's well we don't really have that problem with the other machines... At this point I'm not only losing my time to work on it but also money with customers lining up demanding their orders and me with a lame laser.

EDIT: BTW the water temp was 23C while running the test with air assist on and exhaust running.


http://youtu.be/2C8qMBiIItA

[Ralph Y Thorne]

This may be for another topic. From Jonathan's post... "I have no clue why a direct current tube would pulsate like it does. It's basically perforating the material." Dan, could you answer this one? This can be seen in Jonathan's second picture.

I use an 80w Reci with a generic power supply. The power starts to fall off noticeably at about 25c to 26c. I've never seen mine at over 29c. It may not sound like much, but to keep cutting constantly I adjust the power up a couple of percentage points as the tube gets warmer.

"I just checked the head. The beam isn't perfectly centered but it's not hitting the sides." Have you checked the laser firing in at least the four corners? You need a material that will let you see the shape of the burn mark. White poster works good for this. You should have a nice round burn at each position.

In your video what power and speed are you using? The temperature at 23C is good.

[Jonathan Bowen]

I checked all 4 corners and it looks nice and round to me. I set it to 500ms pulse and all 4 look about the same size and shape. It's a 40 watt tube so my settings for MDF and Acrylic are 5mm/s speed and 100% power with a corner power of 90%.

I need to pack and ship what orders I can but then I'm going to swap the power supply feeding the controller and check it again. Still waiting to hear back from Hurricane. I sent them an email with the video I posted.

[Matt McCoy]

Hey Jonathan: I had an issue similar to yours on one of my lasers. For me, it was a power supply that was overheating which eventually failed. Are you able to monitor the Ma during vector cutting?

[Jonathan Bowen]

No I can't monitor it. I'm really hoping it's just a glitch in the controller power supply. That would mesh with the occasional comm issues. I don't want to replace the HVPS or the Controller.

[Jed Lawrie]

Sounds like a bad power supply. If you are to get another one, ask them to send you at 50 or 60 watt rated power supply instead. You'll see a small boost in power and most likely a lot more consistency.

[Jonathan Bowen]

So I may be onto the issue. I'm assuming the fan on this high voltage power supply should be... moving. I'm going to check with the manufacturer and see if it is triggered by temperature. I sespect that isn't the case. So I will have to figure out what size it is and work on getting a replacement.

[Jonathan Bowen]

So this fan should be running all the time. I replaced it with another of similar size after a connecter swap. Still didn't work. I already had to wire in a 12VDC power supply in order to run the quick replacement I did for the exterior cooling fan. So I pulled the cable out the corner of the case and wired it into that supply. I'm going to be running a bunch of files in order to get a good idea if that actually fixes the issue. Very annoying but thankfully cheap and hopefully it fixes it.

EDIT: So no dice. Still has the same issue. I reduced the speed a bit to prevent issues yesterday so I'll keep doing that for now and just write it off as a fluke in the HVPS due to an over heat and hope it doesn't drop anymore. I'm stuck with it for now but I'm going to have to start working a replacement into my plans for the future.

[Jonathan Bowen]

Just an update. I found the problem finally. Well it found me. A capacitor exploding and smacking to top of the case for the power supply sounds a lot like what I suspect would be a tube cracking open. I panicked and ran over to see the tube was dark. I can see it through a small opening in the back of the cutting area. So I shut the machine down and investigated to find this.

https://dl.dropboxusercontent.com/u/61862/tmp/2014-02-03%2011.59.35.jpg

So I overnighted a bunch of capacitors to recap the entire power supply. I also had a rough time finding a replacement for the power resistor that blew up. I suspect the small cap was leaking a bit and causing the power loss as it reduced the frequency of the switching transistors under it. I already suspected the caps but I figured it was the larger filter capacitors and those were already on order. I was going to replace those and then redo the rest of the smaller ones if that didn't fix it. Without the emergency overnight shipping of $30 I only ordered $10 worth of stuff and it was only that high because I order 2 or 3 times as many as I needed. At $0.50 each its better to over order and store the parts for later then it is to screw something up and need one more.

This sort of capacitor problem is common with chinese manufactured electronics. The use substandard parts and the caps are about all that can go. They are relatively cheap to replace and easy for someone with basic soldering skills. If your having similar issues with a power supply then it might be worth looking into or finding someone who does basic repair work. $10 in parts is a lot easier to swallow then $200 for a new power supply. (or $400 if I got one directly from Hurricane)