Forum Replies Created
-
AuthorPosts
-
roberlinMember
well, I spoke a little too soon. It is not 100% yet with the ferrite.
Failed late into the third trial. Still much much better than it was.
I have a few more of those things I will snap on and then start thinking about more serious measures if necessary.roberlinMemberPutting it on a separate circuit did the trick!!
Since that isn’t a viable long term solution for me (only 1 circuit in my garage), I ran down to Radio shack and bought (per that link you sent’s suggestion) some of those snap on ferrite things. Put one on the power cable for the spindle and one on the USB cable, and it seems to be working properly on the same circuit.
It is really weird that the proxxon is causing this much trouble. Maybe I messed it up somehow.
Anyway, thanks for all your help, guys!!
roberlinMemberWhen I handheld the spindle I kept it within a few inches of where it should have been, but not touching anything, to try to minimize the distance difference. But it is possible my shapeoko is acting like an antenna, and the contact is necessary for that? 🙂
Anyway, I did some more testing and am leaning towards the emi as well. Like it was still stalling when I totally disconnected the tinyg from the shapeoko and moved it onto the concrete floor where it shouldn’t be getting any vibration. (was tracking progress via he graphics in tgfx).
I will get out an extension cord and try a different circuit like you suggest. And I’ll check out that link.
roberlinMemberhmm… ran it with the spindle on in my hands and following as close to where it “should be” as possible and it actually made it all the way through. This would seem to suggest vibration.
But I can’t seem to get it to stall by jiggling stepper motor solder joints or other electrical connections…roberlinMemberMost of this has been on a 2.5 year old macbook pro 13 inch.
I am pretty sure I was hitting the same driver issues as above also, further complicating the matter.
But the rest of the testing will be be booted into windows and sending the file via tgfx.On the shapeoko forum they suggested that the contribution from the spindle might be vibration rather than EMI.
I am going to run it with the spindle handheld instead of mounted on the machine to try to distinguish.roberlinMemberWell, I no longer think my current problem is due to software.
After running the file a dozen more times (at 12 minutes a pop) it seems to make it through the file if the spindle is off
and stall out if the spindle is on (although I don’t actually have it cutting anything to eliminate the variable of load on
the motors).
I moved the power cable for the spindle (a proxxon) as far away as I could and it didn’t make a difference.
Hmm….roberlinMemberSo, I had been experiencing similar issues as above but kind of intermittent. They’ve now worsened to the point where I’m basically unable to use the tinyg. This was with coolterm on a macbook pro, but also in other settings as I detail below.
I tried rolling back the FTDI driver to 2.2.16 and it didn’t do anything for me. Also tried tgfx and windows coolterm in bootcamp, and coolterm for linux on a netbook. All of them just sort of stall out in the middle of sending the file, and it happens with about everything I send at it now (everything is generated with cambam). But it doesn’t always stall at the same place and occasionally it will actually make it through the whole file.
I was wondering, does tgfx use the xon/xoff flow control also, or something else? (i.e. like the grbl g-code senders).
EDIT: to be more specific on what happens when it fails, with linux coolterm it pops up some dialog box about serial port error or “serial port disconnected” or something like that. With the mac coolterm, it usually just freezes up (although one time I sat there and watched as the percentage of file sent kept going up while it was indicating xoff in coolterm and tinyg was doing nothing).
- This reply was modified 12 years, 4 months ago by roberlin.
roberlinMemberWell, it was dropping characters when I sent maybe a 20 line file that consisted only of changes to settings and no gcode.
I sort of thought (may be off base) that such commands would never trigger an XOFF since they don’t fill the gcode buffer.Regardless the problem reappeared for me today after being “fixed” for a couple of weeks, so you are right the delay does not seem to be a cure.
I will give tgFX a shot.
roberlinMemberI had the same problem. It was also dropping some characters when I sent smaller files.
Fixed it by going to the connection options and adding a transmit delay of 100 after each line and 3 (I forget the exact number… somewhere between 3 and 6 I think) after each character. -
AuthorPosts