co2 laser losing power

[assplas]

Hello.

I bought a 2nd hand laser cutter/engraver 18 months ago with blown board, laser and power supply. Got some guys to rectify the probs, which they did. They installed mach 3 board new tube and psu. Has been running fine but would only cut, not engrave. I bought off aliexpress or ebay an awc608 controller before I knew about LO and installed it and it works great. Installed 6/03/2015. My problem now is it seems to be generating heat much quicker than before and then starts losing power.

Machine is unbranded but is type 900LH and is 60w with a bed size of 900 x 580, has 100 litres of demineralised water but no chiller.
We run on 240v in Aus.

It was anologue controlled cutting on up to 27 mA and would cut any length of time without problems cutting at speeds 2mm acrylic at 16mm/s
Now it cuts set on 23 mA max for about 20 mins, 2mm acrylic at 18mm/s and then starts to lose power, water temp rises and wont cut 2mm acrylic with a speed of 10mm/s.
When this happened I changed the tube for a brand new unit and it made no difference. I went back to the old tube and left it for 60 mins and it was fine again. Happened next day, left it and was fine again.
At the mo we cut for 20 and leave it for 45 cycle.
Cleaned mirrors and lense, checked alignment of mirrors and bed, all fine and dandy.
Trust this is understandable.

Any ideas why this may be?

Thanks
Mal

[assplas]

I am sorry if I have posted in the wrong spot. New to this.

Does digital control generate more heat/power than analogue?

Can someone help me with this please?

Mal

[NickWL]

Sounds like it might be overheating because your cooling system isn't powerful enough.
The rising water temp is a no-no, the cooling system MUST be able to hold the water at the required temperature (around 20 deg C – though I run mine at 18 deg which I 'feel' is a better temperature).
Nick

[assplas]

The laser used to run on analogue with 20 litres of water and temp got very high but didn't play up. Had no temp guage but water was warm to the hand.
Would cut 10mm acrylic at 28mA with speed of 3mm/sec and cut a job 50 min duration twice over.
Beginning to think I have abused this baby and have been very fortunate.
Spoke to the former owners and they only had a 30 litre bucket under it with no chiller and didn't have probs.
Will install chiller and see what happens.

Mal

[Tech_Marco]

You are "cooking" the tube because a 60W laser tube should run at 22mA where 28mA should apply to a 80W tube

Keep your water temp at 22'C or around to make the tube happy


Marco

[assplas]

Thanks for the replies.
It worked fine on analogue but now with awc608 I have this prob.
I initially thought the tube was u/s when this prob first appeared so i installed another tube (brand new) and the thing still wouldn't cut.
I am not convinced it's water temp problem though I'm leaning that way.
Could it be the power supply playing up?
I didn't change the psu when switching to awc608.
Trying to get some pointers about it all.
Many thanks

Mal

[NickWL]

Water temperature...

[yaddatrance]

Cooling is the one of the most important factors (outside of gas mixture, which we have no control over in a sealed tube and current) in a CO2 laser.

Higher temperatures in the bore inhibit proper population inversion (via the He->Ne->CO2 loop), which affects efficiency. This is visible through reduced power output levels.
The worse problem is as efficiency drops, the more heat is generated which can cause thermal runaway.

Sealed tubes have to dissipate through the glass to the water (also not a very efficient mechanism) so cool (not cold) water temperatures are critical.

Long story short, the cheapest way to to prolong tube life and cutting times is to increase thermal mass (get a bigger bucket) and add a thermometer.
You can just stop cutting if the water temp hits 70 degrees F, you won't damage the laser by going slightly over 70, but efficiency drops off scarily fast past that temp.

Second cheapest is to add a radiator. But those stop being good enough if your laser is bigger than 60W...
Third is to chuck a bit of ice in the bucket when temps go past 70 (also don't want it too cold, keep it around 65).
Fourth, ice and salt in a bucket with the actual laser coolant running through copper coils (to keep the deionized water pure).
And of course the right solution is to buy a chiller... but even the $500 chillers don't really have the btu transfer capability to handle 80W continuously.

Basic rule of thumb is, regardless of mechanism, you need to remove or dissipate 10x the rated watts of the laser in heat, so 60W laser -> ~600W heat -> ~2050btu,

Also someone noted above, and you probably figured out by now, 28mA is way over the max current for a 60W tube, if you want anything close to nameplate tube life.

[assplas]

Thanks for the imput.
I can see I have run this machine too hard but can't see why it would do this to me after changing to awc608. Was cutting this am, room temp of 13C and water temp started at 11C measured in the 100l tank. Cut for about 20 mins on 22mA, water temp got to 14C but I lost 1mA on the meter. I know I am trying to cut as fast as possible, time = money. But it used to cut fine, now it don't. the longer it goes the more it drops (mA). Is this normal?
I have been cutting at speeds which the machine had no problem with before but now it starts well and then drops off.
A chiller has been installed but I have yet to test it in anger.
Will update this in the next couple of days.
Please keep feeding me.

Mal

[yaddatrance]

The current dropping, after a period of time, isn't something that happens normally. Laser power drops if heat goes up, but the current should be constant at any given setting.

My question is how are you setting your current levels? The 608 are normally pwm controlled (you could alter some early models to analog) while using the 5v reference from the PWM out.
Only way current would normally drop is if the reference 5V was drooping or if the PSU was acting wonky.

Do you set your max using the layer power percent settings or using the manufacturer max_power(%) settings?

Also what model power supply and tube do you have?

[assplas]

Max power is set by manufacturers max power setting. It is set at 85% which gives me 24mA.

PSU is manufactured by
Jinan Hongyuan Electric Co
Input: 220V
Output: 35KV 24mA max
Bar Code: 2 01301 11402 5
That is all the info available on it.

Tube is
Bill Laser Product
Model: blank
No: unreadable
Date: blank

Mal

[assplas]

Update
Water chiller installed. Temp set at 16C, water remains fairly constant 18-19C whilst cutting.
Makes no difference to problem. Current still drops.
PSU replaced, current still drops.
Why would this happen after upgrade from analogue?

Mal

[yaddatrance]

Can you send a screenshot of the 'manufacturer settings' in lasercad? And what wiring schema did you choose (especially between the DSP and laser PSU) ?

Most chinese PSU's implement an RC circuit off the 'IN' pin.

Can you hook up a pot for power control and a switch to test the laser power outside the DSP? Ideally immediately after the power has dropped to a noticeable level?

While unlikely, It *is* possible for it to be the fault of the DSP (specifically the cable between the screen and the distribution board).

There is one relatively easy way to test if its the DSP or not that I can recommend if you have a multimeter and can get a hold of a 47 ohm resistor and 1 uF (microfarad) capacitor. (I'm assuming you don't have a scope)

If you measure the voltage between PWM and ground when the laser is firing 'perfectly' and after the power level has dropped though the following circuit, then the voltage SHOULD stay constant.
If the voltage sags, the DSP is the problem. If the voltage does not sag, its on the PSU/laser end.



I would run this test twice, once with green phoenix connector UNPLUGGED from the laser, and once with it plugged in (so the laser is firing while cutting). This is to eliminate any weird current draw from the PSU affecting the result.

[assplas]

Thank you for the response

manufacturers params as per pic

layer params

Wiring is as per the downloadable file on this forum but with the wp connected to g as per your pic. Used cat 5 cabling from dsp to psu.
After installing the resistor and capacitor I got a reading of 4.06v connected and 4.07v disconnected. The machine cuts better with these mods.
A friend brought a scope round and we connected it up took some photos
At layer 100% manufac 85%

at layer 10% manufac 85%

Thanks again
Mal.

[yaddatrance]

Perfect...

Now with the RC circuit in (the setup with the resistor/cap and multimeter)... Run a job until the power 'sags' and then run the same test and measure the voltage...

If the voltage is significantly lower, then the problem is on the DSP side, if the voltage is roughly the same then the problem is on the PSU/laser side.

One thing I noted that was slightly odd was between the oscilloscope readings at 100% vs 10% (of 85%), the rail voltage seemed to drop almost 2 volts? The 100% reading impies a 6V rail and the 10% a bit over 4V? Normally both would be at ~5V ish, but it could be easily be due to the age of the scope.