Gravograph IS400 Head Positioning Issue

[Crystal Hurtle]

I have been using a Gravograph IS400 rotary engraver for a little over a year. I regularly use the matrix feature of Gravostyle '98 for part serialization (with multiple parts secured in a fixture to match the layout of the file). Recently, the head of the engraver has been causing the text to drift unpredictably in the y-direction. Even though the spacing in the file is uniform, the spacing of the text actually engraved increases as the head progresses through the job. When it comes time to start the next batch, the text in the top row of the next set begins nearly below the part. If I cancel the job and reset the z-reference between every batch, the text drift is minimized, but this is a huge annoyance when we're doing hundreds of parts.

The image below shows a sketch of what I'm generally finding:
316644

It's worth mentioning that this same file and fixture setup has been used before with consistent results. The text on the right-hand column of parts seems to be affected more than the text on the left-hand column.

Has anyone seen this problem before? Any insight would be greatly appreciated.

I am afraid it may be a mechanical issue with the y-axis -- we bought our machine used through a third party. Apart from my experience with this machine, I am relatively new to engraving.

Thanks everyone!

PS - another post in this forum was a HUGE help with getting GravoStyle '98 to run on a Windows 7 machine.

[Kev Williams]

I've run into this problem at various times with all my CNC machines. My opinion of WHY may sound like I'm out in right field, but it's the only thing that makes sense to me. Here goes:

Virtually every stepper motor on the planet (that I deal with anyway) has its movements based on METRIC dimensions. We in the USA have to deal with SAE dimensions. It's a fairly simple conversion process for a computer to do. HOWEVER- there comes a time when a stepper motor is told to go to a certain destination point, and the metric equivalent of the dimension is split. Meaning, the stepper motor moves to the next highest possible dimension. When this happens, the work is now out of position, and so is everything that follows it. On matrixed jobs with identical text, the movement just keeps adding up...

Example, if I engrave the word STOP on my XT machine at a certain letter height (I forget exactly, .21" or so), and run a matrix of this across, every next STOP will be about .010" below the one before it. It always happens when I hit that magic number. My BIL fights this issue with the old C2000 he runs.

The fix? Change the height of the text or graphic you're engraving, by .002". Try this with your job above. Whatever the letter height is, lets say .125" for example, change it to .127". This will slightly change every other dimension within the job, and you'll never notice the difference to look at it...

Run the job-- if everything lines up correctly, then you know that's the problem...

If NOT, change the height to .123", and run it again...

If you're still getting runout, then you may be looking at a controller issue, or bad data (try changing the USB cable)

(btw, I have an IS400 :) )

[Crystal Hurtle]

Thanks, Kev. I tried switching up the font sizes, to no avail. Additionally, the lines in the matrix are slanting downhill, not just offsetting. The letters of the bottom row are also distorted and look squashed compared to the top row. I wonder if the vise is level (easy enough to check). The attached photo has horizontal and vertical markings to contrast where the engraving actually ended up.

316918

The poor image quality is due in part to the scrap material and my phone camera, but this behavior happens on clean surfaces, too. I have the parallel port dongle instead of a USB interface; I wonder if that could have any kind of effect.

[Mark Sipes]

Kev has a valid theory. Does you system give you a Home point read out? y 0.0, x 0.0 ? and when you return to home at job completion it reads the same but in the wrong position. The stepper motor operates at 200 per/rev. there may be one of the counters that is not working in the y axis within the motor.... not sure what the internals look like on an IS400.

.

[Kev Williams]

Wow-- that's not 'math' issue the computer is overcompensating for, and I doubt it's a controller issue--

My money sez that there problem is purely mechanical---

I messed with your pic so I could see it a bit better, then drew the lines on it, starting at the baseline of the first numbers on the left...

http://www.engraver1.com/erase2/badoff2.jpg

--not only does the text start running downhill, each new line of text runs more downhill than the line before it. The last line is BAD!

The line spacing increases slightly from line to line, EXCEPT that it decreases from line 3 to line 4, then increases again...

Finally, I don't know if the text is a basic block, or if it's supposed to look like handwriting. Something tells me it's supposed to look like normal block!

I haven't had to dig into mine since I've owned it, but knowing it's design, with the head moving and resembling a fist on the end of your arm, leverage effect caused by the weight of the cutter head comes into play...

my thoughts are you might have a loose stepper motor pulley set screw(s), or if the lead screws are belt driven (I don't know if the 400 is belt driven, but I do know the screws on my IS7000 are belt driven), one of the belts may be extremely loose and is skipping. Or both...

Or none of the above? The issue in my head with a loose set screw or belt is that the text should run UPhill due to slippage, but yours is picking up distance on the right, not losing it like it should do if a set screw or belt was loose...

OR, is the slippage occuring on the left side as the machine returns to the left to run the next line- Since the machine zero's itself before it starts any new engraving, that might make sense...

Maybe the entire rail assembly is loose, it would explain all of what I see in your pic... and I've seen stranger things! The nice thing is, if it IS mechanical, then it's likely an easy fix, nothing more than tightening or replacing a few fasteners...

I've never seen an electrical issue do this. Electric issues are usually much simpler in behavior, like normal engraving ceasing and the machine proceeds to engrave a straight line towards infinity. That's always fun on someone elses parts...

Good luck-- I do believe your problem to be mechanical, just have to locate the source!

[Crystal Hurtle]

Yes, the text is supposed to be US BLOCK 1L from the Gravostyle library but starts looking like scrawled handwriting by the end.

I am still hunting around, trying to locate the source of the issue.

On the IS400, it looks like the y-axis lead screw is motor driven. I have removed the accordion-style guard along the bottom of the arm, the plastic curtain cover on the front, and a small access hatch on the top of the arm to see what I could see. Removing anything else seems like a major disassembly project.

New observations:

1. The set screw between the stepper motor and the lead screw was tight. I checked all the fasteners I could see/reach (not many), and didn't see any obvious issues.

2. Seemingly at random, the head "hunts" for longer than usual in the y-direction when it returns to the origin in the upper left. It will travel to the approximately correct position, then move small amounts in the y-direction 5-6 times before stopping.

3. The lead screw looks clean enough, but maybe lubricating it could do it some good.

4. Whenever I try to engrave in the air above the parts, the alignment looks much better. I can't tell if the passes in the air are completely straight, but the problem is exacerbated when the tip touches the surface to be engraved.

Based on what I've read on the internet, I don't expect GravoGraph to be too helpful (or affordable) with our second-hand machine, but I'm considering reaching out to local CNC machine repair shops to see if anyone can help.

Does anyone on here have any recommendations on where to look for help?

Thanks for the responses so far!

[Bill George]

Can you tell me how a Stepper motor can be Metric? Never heard of that before, its X number of steps per revolution. Motor drivers convert a signal into steps and tell the motor what direction and how many steps. Maybe I've missed something? Mounts, shafts and frames yes metric but 360 / 200 steps is steps.

[Bert Kemp]

The way I understand it is steps are measure in mm so many mm steps per 360 degrees of revolution more mm per step give more accuracy in head placement. At least I think thats the way I think it works.
this what I found
Let’s assume we have a standard
step motor with 200 steps per
rev. This motor is driven by a
driver set to 5 micro steps
per full step. A Gecko G210 from Geckodrive for example. The
motor is directly coupled to the lead screw
which has a pitch of 5mm per revolution. That
means the axis will move 5mm for each revolution of the screw.
So we’ll take the motors 200 steps multiply that by the drives 5 micro step. (200 X 5 = 1000)
The drive needs 1000 pulses (or steps) to turn
the screw one revolution thus making the axis
move 5mm. So now we take those 1000 steps and
divide by the pitch of the screw, which is 5.
(1000 / 5 = 200). In other words we need 200 steps to move one
unit
or mm.



Can you tell me how a Stepper motor can be Metric? Never heard of that before, its X number of steps per revolution. Motor drivers convert a signal into steps and tell the motor what direction and how many steps. Maybe I've missed something? Mounts, shafts and frames yes metric but 360 / 200 steps is steps.

[Bill George]

Lead screws can be metric or SAE inch based, motors are still 360 Degree rotation either 200 steps or the micro steps its the same, just a finer scale. I have built CNC machines, and did the calculations. Bought the Geckos.

Link> http://www.quora.com/Why-is-a-circle-still-360-degrees-in-a-metric-system-whereas-lengths-are-measured-in-multiples-of-10

[Kev Williams]

on 'metric stepper motors'... I wasn't clear on my metric explanation...

While it reads like I meant the steppers are based on metrics, what I meant was, our programs that drive the steppers are based on metric dimensions...

We enter SAE dims, and basic HPGL is written in SAE dims, but the converted metric dims drive the steppers.

I don't know this to be fact, but based on 34 years of putting up with these anomalies and their cures, it just seems the SAE to mm conversion causes stepper position errors.

How much resolution is actually factored into the conversions anyway? Guess it depends. One anomaly I've had to learn with my cylinder engraver is that the EP modules think that 3.2" = pi. If pi is that far out of whack, it doesn't surprise me that the stepper motors sometimes don't know where to go!
;)

[Bill George]

The odds are its not the metric to inch conversion as all the rest of those same machines would have the same issue. Something is going on. I am wondering if a non-standard part or controller has been installed? Software update that had gone terribly wrong....

But you said it was working fine until this point. Gets worse when the pen touches the work, sounds like a stepper motor or wiring issue, perhaps even loose connections on the motor driver module.

[Chris J Anderson]

Not sure of the is400 uses steppers with encoders or not, or servo's,
but could it be a case of a loose / faulty encoder or stepper.

Is it possible to swap x and y steppers in them ?
That might tell you something if the fault goes another direction when you swap steppers.

Hows Gravo been with their support ?

[Phil Vernon]

Update. See my above post for the cause

Hi Crystal,
Ex Gravograph Employee here and still getting contacted by old customers to fault find, seems Salesman with half a brain don't actually grow on trees after all.
Sounds like the machine may be binding up on the leadscrew and could be improved simply by spraying some silicone cleaner, or you need a new stepper motor. Why do i think this? You mentioned that the problem becomes worse when you are actually touching the surface, the machine will be under more load at that time. In the air there is no load.
Does it happen when the machine is cold or hot? Are the fans working in the control box? Maybe overheating and doing funky things.
Taking the top off the IS400 is honestly dead easy, just a couple of little tricks is all. If your still having an issue let me know and i will try and help. What country are you in?
Regards,
Phil.

[Phil Vernon]

Not sure of the is400 uses steppers with encoders or not, or servo's,
but could it be a case of a loose / faulty encoder or stepper.

Is it possible to swap x and y steppers in them ?
That might tell you something if the fault goes another direction when you swap steppers.

Hows Gravo been with their support ?

Steppers only. Not sure about encoders or whether it just sends out steps with no closed loop. But definitely only steppers.

[Gary Hair]

is that image actually what it looks like? If so, I'm betting you have an encoder and it has some dirt on it and you are losing steps. Even if it's not that precise that it shifts, I'm still betting on an encoder issue. It could be a loose gear or belt between the motor and leadscrew, but that wouldn't typically display such even shifting and it would probably shift differently depending on the direction.
Have you tried running the job from bottom to top? If so, does it shift about the same amount?

[Bill George]

Steppers only. Not sure about encoders or whether it just sends out steps with no closed loop. But definitely only steppers.
If a stepper motor gets overloaded in some way, it starts losing steps. Since it has no feed back like a servo with an encoder bad things can happen. The controller just sends out steps or pulses, does not have any feedback.

[Phil Vernon]

If a stepper motor gets overloaded in some way, it starts losing steps. Since it has no feed back like a servo with an encoder bad things can happen. The controller just sends out steps or pulses, does not have any feedback.
Exactly.
But I've just realised (yep that's how we spell it over here) what it is, sort of What Gary Hair referred to. The IS200 uses a rubber coupling between the stepper and the lead screw. This is where I have seen this error before. The coupling perishes over time and the connection becomes flexible hence the position changes. Imagine 2 small toothed gears, 1 on the motor and 1 on the leadscrew, they are inline with each other (they don't mesh) and these are then both encased in a rubber sleeve with the inside face moulded to suit the gear teeth (this handles the drive part). The sleeve splits allowing 1 of the pinions to sort of freewheel a bit and then grab again. Hence the position change. The IS400 must be the same, I've only ever replaced them on the IS200, The IS6-7-8000 all use toothed belts for drives.
Hence when it is under more load the situation is exacerbated as half of the coupling moves inside the sleeve more.
Ask for a replacement coupling.
Phil

[Crystal Hurtle]

Thanks for the responses, everyone. It may be a few more days until I can get back to the machine with the space and time to remove the cover and work on it.


Update. See my above post for the cause

Hi Crystal,
Ex Gravograph Employee here and still getting contacted by old customers to fault find, seems Salesman with half a brain don't actually grow on trees after all.
Sounds like the machine may be binding up on the leadscrew and could be improved simply by spraying some silicone cleaner, or you need a new stepper motor. Why do i think this? You mentioned that the problem becomes worse when you are actually touching the surface, the machine will be under more load at that time. In the air there is no load.
Does it happen when the machine is cold or hot? Are the fans working in the control box? Maybe overheating and doing funky things.
Taking the top off the IS400 is honestly dead easy, just a couple of little tricks is all. If your still having an issue let me know and i will try and help. What country are you in?
Regards,
Phil.

I am in the US. Sometimes I imagine the distortion gets worse if the machine has been through a few runs, but I may be making this up. The differences are small. The fans in the control box sound like they're running, at least.

Two questions:

1) What kind of silicone spray should I use? Is this preferred over other oils?
2) Anything I should know about ahead of time about taking the cover off? What are these "couple of little tricks"? :)


is that image actually what it looks like? If so, I'm betting you have an encoder and it has some dirt on it and you are losing steps. Even if it's not that precise that it shifts, I'm still betting on an encoder issue. It could be a loose gear or belt between the motor and leadscrew, but that wouldn't typically display such even shifting and it would probably shift differently depending on the direction.
Have you tried running the job from bottom to top? If so, does it shift about the same amount?

I haven't tried running a job bottom-to-top yet. It would be a good test. The machine also has a good amount of dirt and metal shavings inside the case from the previous owner. The whole thing could really use some good cleaning and lubrication.

I'll post an update once I get back to it.

(Edit for mixing up "bottom-to-top"...)

[Gary Hair]

The machine also has a good amount of dirt and metal shavings inside the case from the previous owner. The whole thing could really use some good cleaning and lubrication.

Take care of this first and you may just solve your problem. I had a cnc router I built that was loosing steps on the X axis and it turns out I completely forgot to lube the leadscrew. I cleaned it really well with denatured alcohol and put a very thin coat of lithium grease on it along with a few drops of 3-in-1 oil and it never lost a step again. If it's really bad you could use WD-40 to clean it then clean off the WD-40 with denatured alcohol, then apply the grease/oil combo to lubricate it. Don't leave WD-40 on it as it really attracts dust and dirt and will clog things up pretty badly over time.

[Phil Vernon]

Two questions:

1) What kind of silicone spray should I use? Is this preferred over other oils?
2) Anything I should know about ahead of time about taking the cover off? What are these "couple of little tricks"? :)



I haven't tried running a job bottom-to-top yet. It would be a good test. The machine also has a good amount of dirt and metal shavings inside the case from the previous owner. The whole thing could really use some good cleaning and lubrication.

I'll post an update once I get back to it.

(Edit for mixing up "bottom-to-top"...)

Hi Crystal,
So I assume that you have already removed the top plate on the Y-Travel, this lets you see the Y-Axis lead screw. This is on top of the beam (Y-Axis) that sticks out and holds the spindle. The Y-Axis is where your trouble will be as I am certain it is the coupling that connects the stepper to the leadscrew. I can't exactly recall where the position of the motor is but there are 2 ways of getting further access into the machine.
1. Either manually pull the spindle as far out as it can go, you need to do this so the large cover can come off. Or drive it ou with the power connected and then disconnect everything. The back of the machine cover has a bulge that the Y axis moves into when fully retracted, if you dont move it forward it stops the cover from coming off. Now undo and remove the 4 screws on top and the 5-8 (not sure how many) screws across the back at the bottom of the cover, below the aforementioned bulge. Then simply lift it off. Then you will see that there is not much to it. You now have access to the X-Axis leadscrew, but this is not where your problem is. This drives side to side motion, not front to back.

Not sure if you need to do this next one. The motor and coupling that you need to get at may be accessed by taking the top off the Y-Axis.
Maybe take some photos and either post them here or message me and I'll give you an email.
But if you want to get more access then continue on.......

2. Remove the bellows/dustcover at the front of the Y=Axis beam underneath behind the spindle. To do this you will need to hang the machine off the front of a table and get access underneath, or tip it on it's side or back (but I recommend the first way)


Regards,
Phil.

[Kev Williams]

Hey Phil, just wondering if you know my rep Bruce in Salt Lake?

And while I agree with your diagnosis, I'm having trouble understanding why her machine is GAINING Y-steps on the way right and down the plate rather than LOSING steps?

[Phil Vernon]

Hey Phil, just wondering if you know my rep Bruce in Salt Lake?

And while I agree with your diagnosis, I'm having trouble understanding why her machine is GAINING Y-steps on the way right and down the plate rather than LOSING steps?

Hi Kev,
No, I don't know Bruce mate. I used to be the Victorian Territory Manager looking after Melbourne but I no longer work at Gravograph.

Gaining Y-Steps?
If you go back to the image you will see that it is a matrix engrave. It does the first 2 lines in the top left corner and works its way across, when going from the first line to the 2nd line it is not running fast it is at engrave speed (not ramp speed or reposition speed) nor does it have far to travel. But when it finishes the 2nd line and re-positions to the top of the next grid to the right it hits reposition speed in the case of the IS400 I believe is 60 or 35mm/s. Now if you imagine the coupling as I explained above, at this speed the connection freewheels a bit as the rubber connection is damaged. It cant hold itself together on the loads applied at acceleration (Re-position)
So as it engraves across it does Line 1 > Line 2 fast Reposition (Coupling slips?), Line 1 > Line 2 fast Reposition (Coupling slips?), Line 1 > Line 2 fast Reposition (Coupling slips?), etc.
And when it slips if it slips it does not go all the way back to what it believes is the top, but of course when it goes home at the end it homes to a limit switch so you would be none the wiser.
I put the question mark in because looking at the sample it sometimes slips and sometimes doesn't, may just slip enough to move 1 tooth or might just stop short of 1 tooth and then drop back in the tooth.

So if I have explained it clearly enough you can see how it moves down, its not really gaining steps but more not re-positioning at high speed as the coupling is Kaput! This is a image similar to the rubber casing the 2 pinions one on the leadscrew and one on the motor
319027

With the coupling damaged it will also cause slanting text as the coupling is not rigid but sloppy.


Cheers,
Phil.

[Kev Williams]

Makes sense-- I was giving most thought to steps changing while engraving, not so much thought about the move UP to the next line from the bottom of the previous line.

[Crystal Hurtle]

Phil --

Sure enough! Take a look at this:

320802

The y-axis is on the left. X-axis on the right for comparison.

Does anyone know the best place to find a replacement coupling (or how they are typically sized so I know what I'm getting?)

[Kev Williams]

Not only does that coupler look like it's toast, that screw looks to be severely worn down too..?

Some couplers are in 2 pieces, but I don't think that one's supposed to be? And if that screw is really as worn down as it appears, its likely been putting up quite a fight just to move at all. Could explain why the coupler is in half, those stepper motors are pretty strong...

I'm guessing you're going to be needing a new lead screw assembly too, these 2 screws should look the same, and they definitely don't...

http://www.engraver1.com/erase2/LS.jpg

[Phil Vernon]

Does anyone know the best place to find a replacement coupling (or how they are typically sized so I know what I'm getting?)

Hey Crystal,
Sorry for delay in responding. Was Unemployed for a while but landed a new job and have been a tad busy.
I would just ring up Gravograph and send them the picture.
The picture on the left shows the grub screw on the shaft side and i'm guessing that the other grub screw is behind the motor flange. There are 2 size couplings (different shafts) and it doesn't really matter which one they send as the rubber sleeves comes off. So you keep the existing aluminium parts, look at my picture above and attached (the black bits are the Aluminium and the white bit is the black rubber), you can take the rubber sleeve off as its knackered. I have seen peeps do a temporary repair with some strong tape...but it is only temporary.

Sorry to disagree Kev, but I would be very surprised if you needed a new leadscrew, I have serviced a heap of these and never seen a worn lead-screw. One looks lubed one looks dry. You can turn it by hand to check for binding. They are great machines it's a shame that the new Management in Australia breaks Employment agreements, otherwise I may still be selling and servicing them.

Any questions let me know, maybe with a pm also :-) I will try and be more diligent!
Regards,
Phil.
321456

p.s. I am the father of 4 boys and 3 of those boys just landed in the states for a 3 month visit across the whole USA with a stay in Atlanta for 6 weeks.
So if they get into any bother I may need to reach out through here...hope not though. :-)

[Phil Vernon]

Hey Crystal,
Did you get this sorted in the end?
Cheers,
Phil.

[Crystal Hurtle]

I did get it working! I improvised by clamping the broken coupling together while I finished a job, but an entirely new part is installed now and working great so far. Engraving is just a small part of what I do, so a lot of time passed in between...

I owe everyone here a lot of thanks!