Welcome to the SeeMeCNC Community

I notice when using the X and Y axis arrows on Repetier with a dial indicator mounted to the effector that I lose Z height on some moves where the effector is very far away from a given axis and I ask it to lift that axis quickly via the arrows on the manual control button. I usually lose between 10 and 20 thousanths of an inch each time it happens. It usually happens the first time I reverse movement going from one direction to another. I thought it might be the "Y" axis pulley slipping, because it consistently happened when I moved the X axis to the left when I had the effector way over on the right hand side. So I took it off and found that the blue loctite had turned to what appeared to be hard grease and the worm screws were no longer centered over the two divots I created on the motor shaft to hold the shaft in place. I replaced it with a new pulley and skipped the loctite given that this substance may have created a problem for me by effectively lubricating the motor shaft. This seemed to cure the issue with losing Z when I moved from the far right to the left, but it suddenly started cropping up when I moved from the far left to the right once I got the belt tightened back up etc. I don't hear any belts slipping and I don't hear any noises that I would attribute to anything abnormal, but I keep losing Z when I am at the extreme ends of the print bed's range.

I have been told that a printer radius of 299 is perfectly flat for our printers, but I found that I had to make mine 298.2 to make the effector ride flat from edge to edge on my machine. It's not a major difference, but it was worth noting IMHO. If not for the dial indicator, I never would have known why I was having trouble keeping the same Z across the surface.

Any help you folks can give I will gladly accept.

Thanks in advance,
Gr8Scott

When you're watching the perimeter movement like that with a dial indicator, there are a few factors that can cause that small of a deviance from differences in belt tightness, wiggle in your pulleys on the steppers, missed steps, friction in the arms, and the cheap-skates being too tight. If all that seems to be spot on, you can start eyeing an Azteeg X5 (or smoothieboard or beagle black, etc) and some .9° stepper motors

Cheap skates could be a bit too snug. It's a family trait to over-tighten things, so it's a definite possibility. I tend to go overkill on everything and try to eliminate as much extraneous movement as I can within reason. I might have erred on the side of "too tight" instead of "too loose". The arms seem to be moving pretty well overall and I have no wobble in my pulleys etc. Definitely no belt skipping or looseness in the belts. I'm fanatical about that sort of thing and check belt tightness regularly. Prints seemed to track really well with almost smooth sidewalls on the test cube. I had extruder issues with the original See Me CNC hot end and cold end, so I upgraded to the EzStruder and a V3d V5 Hot End. I haven't had a chance to put the new additions through their paces just yet because I haven't got the issue with the head crashing into the bed solved to my satisfaction. I'll try loosening the x and y axis carriages up a little and see if that solves the problem. I think for smaller prints this might be a non-issue as I can't recall it doing this when the effector was near the center. It would only drop z when near the edges. I also think running the machine at a low/slow feed rate or cutting the acceleration in half to 600 instead of 1200 might help as well.

I have a 24 volt power supply that I am currently using to power the heated bed via a SSR. The effector wouldn't be moving until the heated bed load decreases once it reaches temp, so I could theoretically use 24 volts to power the steppers also to give them more oomph. I have already put heat sinks on the stepper controller chips of the Rambo, so I don't think I'd have to worry too much about the board overheating, but the stress on the motors might be a bit much for them over time. I'd be willing to risk wearing out the steppers etc if it meant that I got solid prints out of the machine for a while. I can always buy new steppers if needed (plus I have a spare for the add-a-struder kit I bought and haven't used).

As much as I would love to run my machine on a Beaglebone Black and a Replicape, I don't have the extra cash laying around as I have already blown a boatload of money on this printer and have yet to achieve what I would call solid output from it. There have been tons of necessary upgrades to account for the warping of the Onyx bed, inadequacies of the hot end and original cold end, as well as the weak ATX power supply that would overload and shut itself off when the heated bed and hot end were running at the same time. I did get some help from customer service when my pulleys cracked, but that was about it really. I got no assistance regarding the weak ATX power supply. From reading other posts, it appears that I was not alone with my problems regarding the ATX power supply that came with my kit.

My birthday is coming up soon, but I have already committed to upgrading the LCD screen on my laptop to get 1080p Resolution finally (it has been driving me nuts with this low resolution for a number of reasons). Can't tell my wife to cancel the order and spend three to four times as much for a Replicape and a Beaglebone (though I would really enjoy knowing I had the whole processor horsepower thing put to bed though...) Maybe Christmas or something if I really butter up my wife...

Just wondering if it would be better to load the Gcode onto a SD card and then run prints that way instead of using the USB cable from my laptop? I get occasional stops and starts when running the machine via the USB control straight out of repetier. It's really strange to see the machine stutter after tapping one of those arrows on the manual movement selection when there is absolutely nothing else going on. How could the printer stop and stutter like that when there is no print and no extraneous data etc to cause problems?

Thanks for the suggestions LordBinky. You are a gentleman and a scholar.

When you have you're dial indicator installed, I would try play and play with lots of various parameters like acceleration, jerk, travel speed, and even segments per second or the homing speed for the endstop bounce, setting them both absurdly (but safe for the hardware) high and low just to get a feel for things, how it affects repeatability,etc since you have a obvious and concrete feedback.

Don't test this out if you think it might bother you, this path leads to madness. One of the things that bothered me and I wish I had recorded like I had planned (and I may still be able to and replicate with the smoothieboard) when I was playing with the dial indicator on the rostock, I had vastly different results on the perimeters if I went from Home (G28) to the perimeter point than if I put in a script to go to a perimeter point, lift up 20mm, travel to the next perimeter X/Y point, and go to Z0 there. It's even different from those two results, with accounting for the height of the 'layer', if you clean the glass and oil up the ball tip of the dial indicator (just to minimize possible factors) and run a print of a circle that runs through those points. I personally liked making my adjustments at that time to improve results on the actual print movements, as it was a apparent whether a change made things better or worse (such as tightness of various parts, and settings).

As for the Rambo stuttering, mine would occasionally stutter going from home to the first print point. Never figured out why, nothing about it was consistant granted I haven't checked it against the phase of the moon yet.

Long before I had even used a dial indicator to check things, I had a point when I was playing with printing lithopanes and I had successfully calibrated about 150mm square area to sub .1 levels by doing prints of single layer 150mm squares. That was before center point calibration and tower position adjustments were looked into on the forums, so it's possible to do it better and quicker than I did. This was a time consuming trial-and-error process using the actual print results to make my adjustments. By the point I got that area visably consistant, I would definitely have failed standard calibration procedure as well as not having an easily repeatable process to share.

I can only give you verbal gratitude Lord Binky. If you lived close by, I'd make you some slow smoked Mesquite BBQ chicken (my specialty). Your kind patience has rewarded me with my first good print. It started out pretty messy, but now it's repeating beautifully and it looks like a million bucks. Even with the stock arms, I'm getting really consistent output now. By printing something large and adjusting the feed rate etc, I was able to dial everything in to a reasonable rate. I got some seriously booger-ed bottom layers due to a tangled fan wire getting caught on my home-made J mount that I didn't notice for about 100 layers. The last thirty layers or so are nothing short of amazing to me. I never dreamed I could get this kind of quality print out of this machine and seriously thought about taking it outside and giving my DPMS Hunter a workout on it. I have had this printer for a solid year and I have not gotten a decent print out of it the entire time.

I found two things. I can't adequately tighten all the micro-switches to eliminate all movement on all of the home switches. One of them just won't allow me to tighten it any more as my screwdriver slips and yet it still jiggles around a bit. Can't do much about that at the moment as I have made a gazillion trips to the hardware store and I refuse to go back for a LONG time unless I absolutely have to. I also figured out that I can print a thick first layer and get by with a slightly unlevel Z in most cases. Sometimes the head skips around on the plate and skates so close to the aluminum plate that it prevents the extruder from extruding anything. It spurts out a thick bead of ABS after that and then resumes like nothing happened. I thought it was impossible for Mic 6 plate to warp, but if you heat and cool it enough times it will warp very slightly.

Maybe later I'll try to bring in a bed leveling feature etc, but for now I'm both shocked and happy.

Thank you sir.

Excellent! Thank you for the praise. I am glad you got this sorted out and thanks for posting your resolution of issue and results, I have searched through a few forums recently and it's begining to become bigger than a pet peeve when a thread ends without a final response (I get that when an issue is is cleared up so there isn't a need to post again, but c'mon just saying 'that fixed it' or 'gave up'/'moved on' is extremely helpful to others and makes for a good information resource).

One of the many things I need to document and post is I moved from mechanical endstops to the hall effect sensors. I bought the sensor itself and mounted (i.e. glued) 1/2" N50 magnets (already had on hand, any magnet type would work though) onto the cheapskate itself. I didn't find the onboard trim adjustments to be worth an extra $30+ dollars so I didn't get the full boards. Because I mounted the magnet to the cheapskate instead of a screw I don't have physical adjustments and have to do it with software, which I happen to prefer now. The nice part is you just have to use that 3rd wire for the endstop connections and it works fine going directly to the board. I zip tied and hot glued (overkill yes...but I don't have to worry and it's protected) to the bottom of the melamine where the switches went and directly over the magnets. I think this is a worthwhile upgrade since I don't have the cheapskate making physical contact with the switch and I know I'm not shifting them or somehow shifting the top plate around. It helped the consistency of my leveling and Z height and allowed me to have worry free high home speeds. I don't think there's anything wrong with the microswitches, but as you noted the little play the have is hard to get out of them and ramming them a screw can't be doing any favors and made me worry about them unless I slowed down my homeing speed. If you go that route, test the sensor and magnet before you mount them. Get the sensor wired up, then check the endstop status (M119) with the magnet close and far away. Which side of the magnet matters, so mark them. Mount the sensors, check them again with the magnets and gcode, and before you mount the magnet make sure one more time you have it the right way by testing it again. You can mount them on a screw as well, just make sure you position the sensor over where ever you choose to mount the magnet.

Ones I bought (5 for $5.46, there are others that are prime if you subscribe to that for the free shipping)

http://www.amazon.com/Amico-A3144E-Effe ... all+Effect

I have three hall effect sensors on boards that I thought about using on the Rostock Max, but I might just save them for a larger delta I plan to build later once I get the Rostock Max dialed in properly. I have an extra 5/16 aluminum plate from Sands Machine that needs a home plus I also have an Onyx to use. I want to use the full diameter of the plate without cutting it down and sit it on the floor instead of the desk to get extra height. Longer arms, belts etc. Prolly use a BBB and CRAMPS board with a few POLULU boards snapped in place if I can get a CRAMPS board or solder one together myself if I can etch one out and not screw it up. Maybe Sparkfun or some other enterprising group will start cranking out and selling complete CRAMPS boards one day. Even though the protection is a bit sparse, the modularity should help make up for it I'd say.