GCC Mercury....Stubborn Problem...HELP ..

[William Wilkie]

Below is a copy of an email that I sent to our service guy tonight..If anyone has ANY ideas please help....Thanks in advance.


Regarding the status of the Mercury as of 10:33PM on Friday night....:
We did receive the Mainboard and Motor today, thanks. I did not have time to replace it at first and was going to do the replacement on Monday......until.....we received a pile of orders LATE this afternoon. I changed my mind and went back to the office at 6:00PM tonight to do the replacement.
First I changed the Mainboard, it did include the EPROM chips as we requested (thanks). When I powered up the first time the new board seemed to be more responsive than the other one so I said "OK now we should be good". After sending a few files we determined that the product was the same as before. NO NET RESULT
Second I decided to changed the motor, since it did show up there was no reason not to try it.
Upon inspection and taking pictures before making the replacement I determined that (1) of the motor mounting screws (the one in the round or "pivot" hole) was almost completely loose. At this point I said "OK case closed this is the problem, maybe I should just tighten everything and run a job". I then decided that as many times as we have been into this machine and since the new motor was laying right there, I would just replace it.
I replaced the motor including checking the pulley set screws in the adjacent wheel. Tensioned the drive belt just a little, put the cover back on and was SURE that this would fix the issue.
We ran the first job.....SAME PROBLEM as before. I ran several jobs from different software.....SAME PROBLEM. I adjusted the "tickle" or "trickle".....SAME PROBLEM. Moved the "tickle" back to the factory default....SAME PROBLEM.
I ran a few plates that I could send to someone....for review to see if you have any other ideas.
If you run something maybe 10 inches or so across the text on each end for about 2 inches looks perfect, the 6 inches or so in the middle almost looks "double printed" more so than the normal "ghosting" that you may see due to the "tickle" being a little off.
Reggie and I have both checked all of the mechanical things that we think could be even a little off.....rails, wheels, pulleys and belt tensions.
It almost looks as though the laser is just simply out of sync with the movement, but only in the middle of long X axis strokes.
We have tried running RAMP ON and RAMP OFF from the software side....SAME PROBLEM.
We CAN run jobs in a very small "cluster" mode with near perfect results, we just do not have the time to run LARGE jobs in "CLUSTER" mode.
I really need this machine running at FULL speed to meet my customer's demands....any ideas that you may have would be VERY much appreciated.

Thanks again for your service.

Best Regards,
William

[Richard Rumancik]

William - you say it looks "double printed". Do you have some magnification so you can look at the image closely? In the ideal case, the edges of an image should "line up" regardless of whether the travel is to the right or back to the left. If you are seeing that every second line matches up then this might help determine the problem/solution. In effect, excessive backlash or wind-up would create the illusion of a double image, because positive-going passes would create one consistent image and negative-going passes would create another. It would look like a double image. It doesn't show up on short strokes - but maybe it never reaches maximum acceleration on those strokes and so the backlash is negligible.

A loose pulley COULD create backlash/windup and cause a double image - I don't know what else might - but you said you checked the pulley. But try to determine if you have two "interlaced" offset images.

[Rodne Gold]

You could have an issue with the ribbon cable to the head or the little circuit board it goes to - the ribbon cables often crack and with flex make and break a circuit
You could also have an issue with the roller carriage and rails - perhaps worn rollers - clean and lube the rails/rollers etc that the head travels on , dont overlook the underside!!
The X/Y motors also have an issue in that they arent dust sealed adequately and the circular shaft encoder gets dusty - tho if you replaced with a new motor , this shouldnt be the case.
Tension on the belts should be "tight"..not just a little ....
I think the mercury has an adjustment if alternate lines are offset - called tuning..it should be in the driver , tho maybe it only applys to the explorer and spirits...

[Richard Rumancik]

A loose encoder wheel could create backlash - but since you changed the motor and locked the pulley it seems that this can't be it.

As Rodne noted - the belt should not have any slack, as it will cause "wind up" or backlash because the motor has to take up the slack before it starts moving the carriage again. Maybe you are being too gentle on the tension? The position of the carriage is determined by the encoder in the motor; ideally the slightest rotation of the motor will cause corresponding motion of the carriage. But if too much slack they will be "out of sync" and there will be a delay. Worth a try to tighten up the belt and see if it helps. Make sure the belt clamps at the attachment are snug. Inspect the belt carefully to make sure there is no damaged spot or any point it could stretch.

Is your service guy the GCC rep or a technician? You might need to get the GCC engineers involved. They could probably pinpoint the possibilities quickly.

I have an old Mercury but I don't think I have the tuning feature Rodne referred to. At least I can't remember seeing it in the driver.

[Kevin Huffman]

Hey William,

Send me an email to Kevin.Huffman@signwarehouse.com. I have a document that has another option for your double engraving.
There is a set of pulleys with some set screws. They sit on a shaft with a flat side. If these set screws are a little loose, when you run a smaller job the screw and lodge it's self against the side of the flat spot. When you run the larger job, it can create enough momentum to pull it loose and allow it to shit to the other side of the flat spot.
The .pdf is to large to attach to this post though.

[Dan Hintz]

Kevin,

Might I suggest fixing that spelling error in the 2nd to last line? I'll be a nice guy and not quote it... ;)

[Richard Rumancik]

William, how are you making out with this? Even if it has been resolved, please post what you ultimately determined to be the problem. That way we all learn something and the thread isn't left hanging . . . it may help someone else some time.

Hope you have made some progress with GCC.

[Robert Ray]

I am not really sure what Tickle is, but the problem you describe kinda sounds like the beam alignment is slightly off, so the further you go away from the top left of the system, the beam wanders a bit, and clips something, like the edge of the nozzle, causing a dual image to be printed at reduced power.

A quick test is to pull the nozzle out and test again?

-Robert

[Robert Ray]

I just thought of something else too...
I have Venus and C180's but in the new drivers, there are some new things called Advanced Vector, and True Image for raster. Might be worth a test to see if any of those things change anything.

-Robert