PID -Hardware needed for a software fix

New Home Forum Mostly Printed CNC – MPCNC Hardware Development – MPCNC PID -Hardware needed for a software fix

Tagged: 

This topic contains 466 replies, has 29 voices, and was last updated by  Ryan 2 months, 1 week ago.

Viewing 30 posts - 421 through 450 (of 467 total)
  • Author
    Posts
  • #67042

    Ryan
    Keymaster

    I just shattered the crap out of my Z axis…..I was really trying to push the PID and it started doing some weird stuff. I need to take it apart and double check everything but I was milling 3.5mm slots in aluminum. Once it catches the wrong way boy does it get violent. 6-8mm/s is too fast in mystery aluminum and when it hit a U-Turn (fully loaded bit) it really just gave up. The single flute caught and pulled the whole axis down into the material. I tried a 3rd run but the upper mounts are busted. Printing some new ones now. I think the RPM dipped just long enough for the bit to catch, game over after that.

    Not fun too watch but I had to let it go to see what would happen. Now I know it can 100% pull itself apart!

    #67043

    Jeffeb3
    Participant

    Pics? Cool? I guess?

    Do you think the PID loop should be better or were you choosing a speed that was too low?

    #67044

    Ryan
    Keymaster

    It hesitated (RPM dip) in free air, a few times. I have no idea what or why it happened. Never happened before. Someone came over to pick up there order and I was moving the PID stuff around so I am assuming something is loose now…Not sure. I printed the wrong dang part so I am printing the correct size now.

    I do have the fail vid, I will upload it now.

    #67045

    Ryan
    Keymaster

    #67046

    Jeffeb3
    Participant

    Whoa! Yeah. Hmmm.

    #67047

    Ryan
    Keymaster

    That video makes me pucker…and not from lemons. The previous run I put in 2300rpm instead of 23000rpm. I think that did it in. The one before that I was moving at maybe 10mm/s and loaded up the endmill and got stuck. The first run I swapped the plunge rate for plunge depth…never rush it… 4 of the hardest runs I have ever had and I just let it go for a while to see what would brake.

    #67053

    Ryan
    Keymaster

    Redemption!!! Pretty happy for no air assist, now I need to try Kevin’s air and see how fast I can go. This is three 1.1mm passes, two 1.65mm passes works but not this well. I would like to make a few more of these in different materials on the same machine to kinda show speed and depth differences of materials, (and I need to try steel because of Ben…).

    https://youtu.be/BRFeZESDo4k

    Slick
    slick-raw

    1 user thanked author for this post.
    #67062

    Jeffeb3
    Participant

    That is just mesmerising. Was that with the PID on? I didn’t hear the speed drops.

    #67065

    Ryan
    Keymaster

    No. I need to give it once over and do it again with the PID. I made a bunch of aluminum cuts yesterday, learned a ton. Now I have a solid baseline to do some experimenting.

    #67081

    Todd Hansen
    Participant

    That is way to cool!

    #67170

    Jason
    Participant

    That’s a nice cut. I tried alu for the first time today, and experienced chatter. I shut it down quick. I need a nother pass to tell if it was a cam issue . Looks like you don’t worry bout the rpms much with a router..

    #67171

    Jason
    Participant

    Wow. Just made a second pass using my idea about stabilizing the z plate, I ran the same code, no chatter. It cut great. You asked me to get my machine dirty. Done.

    1 user thanked author for this post.
    #67172

    Ryan
    Keymaster

    using my idea about stabilizing the z plate

    What is this, can we see? I have a slight design tweak for the gantry but I have to release the 25mm LR2 first….

    #67209

    Jason
    Participant

    I posted pictures previously. It got bad reviews. But it works for me. Made the z more stable. All it is is 6/32’s holding it down. There’s a nut in the inside. In one direction, it does very little but in the other it stops some flex in the assembly.

    Every little bit counts, right?

    I may be wrong. I just think it’s too simple of a mod to not at least try. Tell me what you think.

     

    Also, sorry if I come out wrong in writing. I am really good at talking, and get along with everyone, but on paper, I seem to find a way to get under people’s skin. So I’m sorry.

    BTW, read this entire thread. Wow. Congrats man. This couldn’t have been easy. I really like that you brute forced some things. That’s definitely how it goes when learning and developing new things. I can definitely relate to that.

    #67213

    Jason
    Participant

    Same suggestion as before, maybe a spacer can be printed? An opening for the nut on the inside?

    #67216

    Jason
    Participant

    Wow. Just realized something. When I added this, I cleaned everything. I removed it when I replaced the plates, but added them back today. There are some plasic shavings that showed up where the screw enters the x and y cases. So there is some movement I can’t really see, but if the screw is shaving out plastic in one direction it must exist.

    When I added this I only expected movement in one direction. I should have added a second but above the cases to really lock it in place. Unfortunately that means removing the gantry, and I’m kinda using it right now.

    #67329

    Ryan
    Keymaster

    It really should not be moving, is your Z screw bent? Sorry wrong thread, I will find your other one.

    I am messing with these parts in CAD right now. I have a lil idea, and looking into your mod as well.

    #68345

    Ryan
    Keymaster

    Sorry I haven’t updated this in a while. I have been using it the whole time.

    The issues I was having were just loose wires. It still works perfect.

    Yesterday I was making some new cuts and Mistyped 2,400 instead of 24,000 RPM. The 660 worked great. It did not seem to initially like it on the test bench but I let it cut at that speed until it failed. I thought 8000 was as low as it would go…I was wrong again. I am not sure if that is a bad rpm range or just didn’t work at the time for some reason.

    Estlcam seems to have been updated to fix the finishing pass RPM pass through in the gcode as well. Although I am not sure it works as intended yet. I will have to do a quick drastic RPM change to check, I got hungry and ran out of time.

    I really like using it now. On Off control, RPM is steady. I need to stick some sort of LCD control on to dial in RPM settings. I also need to check if feedrate changes on the LCD change the RPM..Seemed to me it might have.

    #68356

    Jeffeb3
    Participant

    You can also use your fancy silly scope on the marlin output to see what it’s trying to set the speed to. It won’t be in a convemient format, but things like finishing pass speeds and feedrate multipliers should be obvious.

    #68358

    Ryan
    Keymaster

    Ha, thanks. Not used to having such fancy tools!

    #68372

    Jeffeb3
    Participant

    I wonder if adding a bluetooth serial module would be a good choice for the interface? Banging on a terminal doesn’t sound like it’s much better than an lcd and knob, but it’s less hardware and if we wrote our own android program, we could make it as complex as we want without adding any additional hardware. I don’t know how the pod people would connect… Or people with dumb phones… Hmmm.

    #68376

    Ryan
    Keymaster

    I have one of those sitting right here. There must be some sort of generic iphone/android wysiwyg app software.

    That dude is killing it with the videos. Need to get him into CNC.

    #69654

    Guffy
    Participant

    it’s funny that on tindie KRIDA’s “PWM AC Light Dimmer Module 50Hz 60Hz” declared as 5A
    on their EBAY product page they declare it as 2A in header and 3A in details.
    their dumb module (uses same BT139) “1CH AC Dimmer Module Controller Board Heatsink” declared as 2A and peak 5A.
    Datasheet of TB139 declares 16A RMS

    similar dumb module by robotdyn uses ST BTA16 (same 16A RMS by datasheet) and is much cheaper

    #69655

    Ryan
    Keymaster

    I am not that versed in this stuff but I have always taken data sheets as absolute max with insane cooling.  I can tell you as it is implemented in this board with the current heat sink it does fine with 120v 5A, warm/hot but is stable and works for long periods with no extra cooling.

    When I build some sort of enclosure I will add a fan.

    As for other versions, in needs to be built for inductive loads. That one I see is $3.50 but no snubber. I am not a pro at this side of things. I am not sure what if any features I can sacrifice or should be added.

    #69660

    Guffy
    Participant

    Of course maximum current could be achieved with good radiator.
    The fun thing is that they declare different current values for same triac with same radiator.

    #69661

    Ryan
    Keymaster

    I screw stuff up like that all the time. Hard to keep track of it all. That is why I bought and tested one before I posted about it, I was unsure as well.

    #70979

    Racerflea
    Participant

    Curious if yall have looked into using a digital signal like the quadcopter quys use. They are starting to break grounds on 32bit esc’s. however the d-shot protocol is still using the standard processor and is insanely fast.

    #71055

    Ryan
    Keymaster

    I remember reading about that. I am not that knowledgeable about it but I will remember that for sure. Right now we are getting plenty of speed from the arduino, but are limited by how often the AC crosses zero.

    1 user thanked author for this post.
    #72793

    John
    Participant

    I was looking through the thread and there doesn’t seem to be a final schematic of how to put this together and how to connect it to Ramps or Rambo Board. Are these going to be released in the near future. I realized that this is one hurtle I didn’t expect.

    #72794

    Ryan
    Keymaster

    Well it is not really finalized yet, here is a link and schematic.

    pidgin

    It attaches to whatever pins you designate in the firmware on the Rambo I use, MX2 45 &31, shown here.

    I do have the dev boards / kits for sale in the shop

    Attachments:
Viewing 30 posts - 421 through 450 (of 467 total)

You must be logged in to reply to this topic.