PDA

View Full Version : How much does water temp effect output power?



Robert Silvers
10-15-2013, 8:46 AM
My coolant temp ranges from 18 to 22 degrees. Is that enough to change my output power by a few percent such that if I etched a large solid area of glass it may be visible as banding?

Or are stripes only caused by power supply instability?

Dan Hintz
10-15-2013, 11:08 AM
Banding is typically seen as lighter and lighter followed by a sharp increase in power. To create such an effect due to temp variations you'd have to suddenly chill the tube from one line to the next.

I don't know how those Chinese chillers work, but a good design should chill a reservoir of coolant, not the outgoing stream, to avoid temperature fluctuations like that. If you're seeing the chiller temp vary back and forth over a very short period of time, they need to work on their control loop (or where they are getting their temp measurements from).

EDIT: If the banding is a smooth change from lighter to darker and back, that would suggest an oscillation between the chiller output temp and the tube heating it up. Bad, either way.

Dave Sheldrake
10-15-2013, 11:12 AM
Is that enough to change my output power by a few percent such that if I etched a large solid area of glass it may be visible as banding?


Nope


Or are stripes only caused by power supply instability?

Unstable output from the supply or heat build up on the final lens will do it (possibly heat build up on the mirrors too but at 100 watts unlikely)


that would suggest an oscillation between the chiller output temp and the tube heating it up. Bad, either way.

Or good for tube suppliers ;)

cheers

Dave

Dave Sheldrake
10-15-2013, 11:16 AM
If you're seeing the chiller temp vary back and forth over a very short period of time, they need to work on their control loop (or where they are getting their temp measurements from).

From memory they are only 6 - 9 litres of water Dan :) The pumps in the 5000 series are tiny little things, probably better off used in a fish tank.

cheers

Dave

Robert Silvers
10-16-2013, 12:16 AM
The default chiller settings are to bring the temp 2 degrees below current room temp, and then keep it there +- 0.8 degrees.

I changed the settings to be a fixed temp and +- 0.1 degrees and went back and forth 6 times at temps from 10 to 20, and in every case the tighter control, regardless of the temp, resulted in banding dropping a bunch. I engraved a solid grey halftone screen to make defects easy to see.

Even though I did this test, I still don't believe it. How can it matter that much to be off by 1 or 2 degrees?

Dave Sheldrake
10-16-2013, 2:12 PM
Question is will a 5200 chiller do the same if the machine is running for 3 hours?

Laser efficiency changes with temperature, last time I posted all the physics behind it my UK supplier got inundated with questions so I won't be making the same mistake again :)

cheers

Dave

Kev Williams
10-16-2013, 2:55 PM
Another reason for banding is the laser beam paths overlapping. My LS900 does this, the path actually wobbles. It's not a mechanical issue either, because I can repeat the same job 20 times and the wobbly path will line up perfectly each time. I have several samples burned into plex that shows this. If I can locate them I'll post pics. (if I can't I'll make new ones!) Different speeds and/or power settings will change the wobble, but the "new" wobbles will also line up perfectly. Been a thorn in my butt ever since I've owned it because it makes etching out large areas on plex or anodized material pretty much impossible because the result looks like crap. New Hermes only answer to this is "it's a function of the laser's ability to engrave photos". Running in photo mode does reduce the banding, but doesn't eliminate it. Vector path runs straight & true, it's only the raster path.

Dan Hintz
10-16-2013, 3:38 PM
It's not a mechanical issue either, because I can repeat the same job 20 times and the wobbly path will line up perfectly each time.

Actually, that points out that it is a mechanical issue. It's likely a worn or off-center gear/cog.