Update: Cut for about an hour before it crashed and released tension. After a bad encoder read. Leading to a 7mm trench cut through the sheet which has rendered it unusable. I think static is a very real problem to solve here. I’m cutting bendy poplar ply which is light and fluffy and, so I assume, a potent generator of static. I am going to try earthing as much as possible and coating with home made anti static spray (Isopropyl Alcohol, Fabric softener and water).
I’m not convinced that dust is involved in the errors - my current theory is that the static builds up enough without the dust collector anyway. I will not be able to source an anti static hose till next week but given my timelines I will try without and with much earthing etc.
The dust (coupled with the sled sliding across the workpiece) is the source of the static.
Dust collection just rubs all that statically charged dust together along with a constant source of new air while pulling it into the tube, exacerbating things.
So far so good with the static spray and earthing of the sled. Just a guess but is it possible that static builds up in the increased dust in the air when there is no dust collector? Would be consistent with my experience of a shortening time between errors throughout 4-5 hours - as the dust in the space increased.
Just a question, where are y’all located, and what is the weather like? High humidity dissipates static electricity, I’m in very hot very humid Texas (see the above y’all reference) and have only seen this error once when I was running it in an air conditioned space.
I’m along the Mississipi in Illinois. Humidity is the norm, especially when the corn in the region is tall, just leeching water into the air.
I run my machine in my garage, which only has the main garage door, which is “open” even when closed because the top panel leans inward.
I’m thinking I only saw the error once, maybe twice, and I haven’t upgraded firmware since my first successful relief carving…2 months ago? So I shouldn’t have any of the fixes that have been put in to help mitigate the issue, yet, either.
My garage electrical doesn’t appear to be grounded, I haven’t done any soldering or anti-static/emf mods, etc., and introducing dust collection with a cheap cyclone, bucket, and tiny shop vac caused no new issues.
I’d estimate I have run the machine for about 30 hours at this point, post calibration.
The only issue I had (other than the 1 or 2 times I had a job stop for what I remember being the 15mm error super early on that I mentioned already) was the “Push To Extend” bug rearing it’s head twice, the second of which chewed up a belt and called for a reassembly of that arm.
I struggled with this error and others a lot prior to grounding things. Still not a ton of runtime for me, but grounding the sled and vacuum hose and under the PCB board using m8dn’s mount and copper tape on the bottom of it has helped me a lot. I run in my garage, and it gets up to the 90s inside there (100s on hot days). I am in Colorado and in general where I am the humidity is low (35-50% relative, less in summer than winter).
Checking in to report back my results. Yesterday I replaced the ethernet cables with longer and shielded cables and this initially seemed to have solved my problems. Was able to cut a small simple project (M4 Stand) no problem. I even removed the grounding setup as I would rather not have that attached if it is not needed.
Then started another project that requires a lot of pocket clearing. Was running great for ¾ of the cut and then the cutting stoped with no error message come up! Ran the cut again and got a Motor current on Bottom Right axis exceeded threshold of 4000 error this time. Both cuts appear to have been going off track from planned cut upwards. Forgot to take pictures but will today.
Not finding much information on this error on the forums right now but going to run some more cuts today to see how repeatable the issue is. If anyone has ideas of anything I should try let me know!
Today went a lot better! No issues and got through some larger cuts. Thanks everyone for the items to try to resolve this! Really excited to be cutting in earnest and work through my project backlog.
I’m in Wellington NZ. Not particularly dry but I have got the machine set up in a small space. With an anti static hose connected and the other grounding I already had in place I have yet to see another error after 12 more hours of cutting.
OK, I am still struggling with the 15mm and Encoder Read error failures even after implementing all the mitigation steps we’ve considered in the past (grounding DIR pin, copper shield between control board and router, grounding everything including the dust collection hose to earth ground, cleaning and sealing the motor ethernet ports).
I am still consistently getting errors on my Top Left motor when trying to run a test job. This is without a bit installed but router running (dust collector off too). However, it isn’t entirely broken - the machine will run through the job for 10-15min before finally failing (but has failed at least 3 times in a row today).
I’ve attached the logs. Does anyone have an idea what I could try next?
Is it possible my solder bridge may not have connected on this board and cause the problem to persist? I really don’t think that is the problem but at a loss at the moment.
Or maybe I need a new encoder board? Which means I’d need a new motor and base, too, since I couldn’t get the screws out of the motors after loctiting them originally. Maslow-serial (7).log (21.1 KB)
Yes, I did that too - but using the 3d printed TPU molded covers instead of hot glue. And replaced the ethernet patch cables with shielded versions too.
I think that the hot glue is going to be quite a bit better than the covers. The hot glue will hold the encoder wires and connectors together physically and prevent vibrations from causing the encoder to disconnect.
The super-tight fit of the tpu cover /should/ do the same.
The PLA one I made with just slightly larger tolerances prevents the clip from engaging, but fits so tightly I had to use a good bit of force to pull it off, so it would have ostensibly prevented separation, too.
Currently on deck for me is to make it easier to don/doff the cover while also making it cover the whole clip. I’m seeing some potential issues with needing to have openings, but I’m thinking a piece of felt or open cell foam would work perfectly to create a compliant seal, considering the size of the particles we’re trying to block would struggle to get around those.
Hey everyone, I am also dealing with the 15mm error and I have noticed an issue with one of my belts not spooling tightly anymore during the initial setup. The other three arms remain tight. There seems to be a mismatch between the how much the motor turns and what the encoder reads while extending the belts. I noticed the coil was way too loose and took this video. Sadly this caused my belt to bind on the retraction step. Does any one know of a method to monitor the current reading of the encoder as I “slip” the belt? My best guess is that some current is “leaking” into the encoder when the router is spinning and I’d like to compare the readings from that belt on a dry run. Also I glued the Ethernet port before ever running my M4 so the problem is not a bad connection.
I too am struggling with 15mm errors that seemed to get worse after I took all the steps to remediate all suspected causes (grounding DIR pin, shielding, hot gluing).
I think it would be very helpful for some verbose logging that we can toggle that could show detailed logging of what the controller is reading from the encoder over time. I would like to know if it is jumping around sporadically. Being able to turn it on for a single motor would be ideal.
