Graphtec 8600 Not Accurate

[Chicken Quasar]

Our FC8600 has started to become less accurate. It drifts in accuracy after about 2-3 feet.

See my photos below.
The W is dead on but 2-3 feet down the roll the C has drifted by about 1/8 of an inch.

Does anyone know what we can do to solve this?

Our grit rollers seem to be clean and grippy and the pinchers are on the tightest setting. This has been getting worse over the weeks,

 

[JBurton]

Did this also cut a weed border? Is the weed border hitting all of the reg marks dead on? (honestly the W is already off by like 20%, not surprising it's continued to get worse and worse.)
New wheels for the pinch rollers, replace the cutting strip, new blade, all are relatively cheap and the simplest things to address before tinkering with offset and other bologna.

 

[BigNate]

Are you using segmented area correction? you can set the segments pretty small to correct for slippage more frequently.... Without seeing you setting, try plotting a little slower with a little less down-force (just make the tests to make sure you have enough down force to cut - and a sharp blade always helps)

As a rule, less acceleration and less drag will mean less slippage - you get drag from the knife, from downforce, from media dragging on the platen, from pulling against the roll, etc. On media I have found is prone to slipping, I will form-feed the whole image through the Graphtec and then back to the start - letting the media accordion up behind the plotter, never re-roll it, the rotation inertia of the roll can be bad.

 

[Chicken Quasar]

Are you using segmented area correction? you can set the segments pretty small to correct for slippage more frequently.... Without seeing you setting, try plotting a little slower with a little less down-force (just make the tests to make sure you have enough down force to cut - and a sharp blade always helps)

As a rule, less acceleration and less drag will mean less slippage - you get drag from the knife, from downforce, from media dragging on the platen, from pulling against the roll, etc. On media I have found is prone to slipping, I will form-feed the whole image through the Graphtec and then back to the start - letting the media accordion up behind the plotter, never re-roll it, the rotation inertia of the roll can be bad.

I am not sure what "segmented area correction" is but I will look it up in my manual. We typically plot slowoly. Is there a setting for acceleration? Can we just turn acceleration off? Where in the menu are those things?

 

[Chicken Quasar]

Did this also cut a weed border? Is the weed border hitting all of the reg marks dead on? (honestly the W is already off by like 20%, not surprising it's continued to get worse and worse.)
New wheels for the pinch rollers, replace the cutting strip, new blade, all are relatively cheap and the simplest things to address before tinkering with offset and other bologna.
View attachment 172128

I remember once a few months ago tinkering with offset trying to set it up for perf cuts. Perhpas do you think that may be what is helping this? is there a way to reset the offset to factory?

 

[BigNate]

I am not sure what "segmented area correction" is but I will look it up in my manual. We typically plot slowoly. Is there a setting for acceleration? Can we just turn acceleration off? Where in the menu are those things?

Acceleration: first, speed is distance traveled per time, you media and your plotter head both have speed likely in mm/s. Acceleration is speed traveled per time, if it is displayed, mm/s/s. If you do not want any acceleration in either your plotter head or media, then you cannot plot - no you cannot turn off physics and still plot.

well, if you do not know what the segmented area correction is, then this is likely going to be your fix... it is where the press prints registration marks at intervals along the ends and sides of your image, then the plotter can scan and plot only a small area at a time, and between sections it can scan the next set of marks and make any correction to stay registered to the print.

there is a setting for limiting acceleration, but generally I look at the speeds things move, if it looks or sounds jerky, likely you have a speed too high for the condition and the resulting acceleration to achieve the speed is too high. Most plotters limit the acceleration for curves... for practical purposes you can think there a little inertia in the head and no inertia in the media for side-to-side plotting... However, there is no inertia in the head and lots of inertia in the media for in-out plotting (and there is rotational inertia when unrolling from the roll, but only for initial unroll of that part of media - why I form-feed the image first).... you need to try to keep all of these variables as constant as possible to achieve consistent results. (I have had to tape a heavy leader to the front of an image because the force of gravity puling down on the front was different than the force on the back when plotting the lead edge - this made the heavy banner media slip a little then the Graphtec pulled forward, but did not slip pushing back - adding the weight to the front made the forward/reverse more consistent for that job....)

your plotter will always drift compared to the print, the goal is to limit the slippage by minimizing all of the factors, but to also set a segmented area that is small enough that the drift is not out of spec before the plotter scans and adjusts for the next section.

 

[JBurton]

Acceleration: first, speed is distance traveled per time, you media and your plotter head both have speed likely in mm/s. Acceleration is speed traveled per time, if it is displayed, mm/s/s. If you do not want any acceleration in either your plotter head or media, then you cannot plot - no you cannot turn off physics and still plot.

I think you're getting too technical (don't get me wrong, I love it), if we don't want to confuse op. On the graphtec, acceleration comes in 3 values, 1 2 and 3. It could be a reference to something like acceleration due to gravity, or some other mundane SI value, but assuming he hasn't tweaked those settings ever, he ought to take a look and try setting it to 1.
What's getting me, he must be unrolling the whole thing to complete a scan operation, so it must be somewhere in the act of cutting that the plotter is experiencing too much drag, and the cutting strip developing a groove is so often overlooked, I'd put my money there. While segmented marks helps in accuracy, since he is losing length in one direction consistently, I don't think segmented marks would help as much as checking the blade, strip, wheels.
OP, what software are you using to scan registration marks?

 

[BigNate]

I think you're getting too technical (don't get me wrong, I love it), if we don't want to confuse op. On the graphtec, acceleration comes in 3 values, 1 2 and 3. It could be a reference to something like acceleration due to gravity, or some other mundane SI value, but assuming he hasn't tweaked those settings ever, he ought to take a look and try setting it to 1.
What's getting me, he must be unrolling the whole thing to complete a scan operation, so it must be somewhere in the act of cutting that the plotter is experiencing too much drag, and the cutting strip developing a groove is so often overlooked, I'd put my money there. While segmented marks helps in accuracy, since he is losing length in one direction consistently, I don't think segmented marks would help as much as checking the blade, strip, wheels.
OP, what software are you using to scan registration marks?

good thinking - to carry one point a little further, if he is pulling the whole image through the plotter to read the 4 corner marks initially, then I believe he is making a the 'measuring pull' under different parameters that the "plotting pull" --- or rather when the plotter is form-feeding off of the roll to get the initial corner scans, it has the back force from the linear inertia of the media, but also the rotational inertia of the roll.... Then when plotting the only back force on the pull is from the media (gravity and inertia)....

The point: if he is not form feeding the whole roll before the plotter makes the corner measurements, then the parameters for the measuring and for plotting are different - and you would expect it to be off by the plotter cutting slightly further down the roll than expected.... (from the pic at the top I would guess that the roll was to the right and the lead edge was to the left ---- since if this happened the plotter would pull slightly more media per distance when plotting than it did when measuring (more backpressure pulling media toward roll during measuring when compared to plotting)

I think checking the cutting strip is a solid suggestion. (quick recap: if the area plotted is slightly longer on the roll than the print, then not form-feeding would give results that match the data.)


most likely there are a lot of little factors adding up to OP's headache - work through some of the easier to do things and see. (if it is off only 1/8" per 3', set a segment area to 8" and it should self correct when it is about 1/72" off.)

 

[dragonmecprint]

Hi OP. Im having the same problem. I’m using a graphtec CE7000-130. Just wanna ask if you have found a solution coz it might work in mine. IMG_0644 is the correct cut but my cutter is cutting a little bit off going to the right thus the next photo IMG_0645 then it’s okay again around in the middle section. It’s a hit and miss for me that’s why I’ve been reading here.

Im looking at BigNate’s suggestion.

 

[coltonwanderson]

I know you said you have the pressure settings all the way up, but when you lower the pinch rollers can you verify that there is no play in any of the roller if you wiggle them.

Only ask because as a Graphtec technician, a common issue with the 8600s is there is a little piece of plastic on the pinch rollers that breaks and then they are loose even if you have the tension setting at its highest.

 

[dragonmecprint]

I know you said you have the pressure settings all the way up, but when you lower the pinch rollers can you verify that there is no play in any of the roller if you wiggle them.

Only ask because as a Graphtec technician, a common issue with the 8600s is there is a little piece of plastic on the pinch rollers that breaks and then they are loose even if you have the tension setting at its highest.

Hi. I’m sorry to ask, What do you think is the cause of my problem? My problem is almost exactly the same as OP. Already checked my pinch rollers. They’re all great. Any advise would be of great help from graphtec tech. Thanks in advance

 

[coltonwanderson]

Hi. I’m sorry to ask, What do you think is the cause of my problem? My problem is almost exactly the same as OP. Already checked my pinch rollers. They’re all great. Any advise would be of great help from graphtec tech. Thanks in advance

I would a run the following test to narrow down the issue to one of to major categories, either the tracking, or the cutting. Load a roll or lon gsheet, and put a piece of tape on the cutter where the edge of the paper is. Then feed it forward something like 10-15 feet. Then feed it back. If it lands right back where it was then you're tracking is fine. If it has shifted, then you know the tracking is the problem.

If the issue is not tracking, I'd feel the carriage to see if there is any wiggle. A loose carriage can cause cut quality problems. Sometimes old firmware or even software can cause odd issues. In rare cases the scan motor going bac can cause issues. As someone else mentioned I believe, sometimes you just need more registration marks for longer jobs. But I would start with that test to start narrowing down the issue.