PID -Hardware needed for a software fix

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

This topic contains 458 replies, has 27 voices, and was last updated by  Guffy 3 weeks, 4 days ago.

Viewing 30 posts - 331 through 360 (of 459 total)
  • Author
    Posts
  • #61435

    Jeffeb3
    Participant

    You can just use wire for vias. You need to make the holes bigger, like regular through holes. You still need a pretty tiny bit though.

    Another fun thing to add or work on are smd LEDs, even if its just to indicate power.

    #61436

    Ryan
    Keymaster

    I also tested out the sketch without the second interrupt call for the sensor…works fine. No idea what happened the first time around. But that made me look at the PWM in interrupt loop.

    Do you remember or know why we are not using analogRead for the incoming marlin signal?

    #61437

    Jeffeb3
    Participant

    The incoming signal is a pwm, and not analog. Plus analog reads are slow. There is a pulse in function, but it’s also slow. What is the problem you’re seeing?

    #61438

    Ryan
    Keymaster

    Just trying to simplify the sketch. No problems. Just seems like there would be an easier way to read the PWM, no big deal.

    Trying to simply and kinda plan ahead, like leaving room for the direction pin in case we use it to trigger something, or even use it with a switch to signal a variable RPM.

    #61439

    Ryan
    Keymaster

    Moving around some pins in the sketch….so be careful.

    Marlin PWM in to Pin 3 swapped
    Sensor pin in to Pin 2 swapped
    Dimmer PWM out to Pin 9

     

    #61440

    Ryan
    Keymaster

    ehhh?

    1 user thanked author for this post.
    #61442

    Jason Dentler
    Participant

    This all sounds really awesome!

    #61443

    Ryan
    Keymaster

    I’m having a blast and learning so much.

    #61451

    Ryan
    Keymaster


    I totally get it now, prototype PCB’s are a huge time sink. I am sure this will be the last one I ever do this way. This would have taken maybe an hour with CAD and CNC. Instead of all the time I spent on this by hand. Add a few more wires and some surface mounts and this would be a huge waste of time. CNC 4 Life!

    #61455

    Jeffeb3
    Participant

    That is so organized.

    #61457

    Ryan
    Keymaster

    So maybe an LED for the enable pin, and the wobble pin if that becomes a thing. The Nano, rambo, and dimmer all have a few LEDs each. I think I can recycle the Nano pin 13 LED as the enable light.

    Kinda stoked to get into a real (basic) PCB on the CNC. Tighten up the dims, and match the dimmer footprint maybe, or at least figure out a mounting solution. All this power floating around I am bound to get shocked.

    #61458

    Ryan
    Keymaster

    Well this is great timing, first youtube suggestion. Thought I was going to take a break and watch a dumb video….nope.

    #61459

    Jeffeb3
    Participant

    It would be interesting to order a board and make the same one. Just to see how well you can do (and if it’s worth it).

    #61460

    Ryan
    Keymaster

    Shoot I am most interested in the file to cut. I mean I love CAD but….and then yup compare.

    #61461

    Ryan
    Keymaster

    So far so good, a few minutes in and I have the entire layout. Now onto the PCB part.

    #61462

    thesfreader
    Participant

    Well this is great timing, first youtube suggestion. Thought I was going to take a break and watch a dumb video….nope.

    Ah ! Was just going to suggest you check the video out :DDD

    #61503

    Ryan
    Keymaster


    Not bad for my first PCB, right?! 0.03″ traces.

    So I learned a lot here. First, Barry, thanks. The tape and super glue is freaking awesome! I think I need to surface my pcb work surface before I try and do it again. Pretty sure that was the only issue mechanically, you can see the top traces are a bit thin and dog boned. Some of the pads are a weird size, I didn’t pay enough attention during the design but it looks like different parts have different pad sizes like the two tiny ones for the cap, and the smaller headers are tiny pads as well. I forgot to put the tear drops on them, that would help for sure.

    I am happy with that as a first try, on PCB design, conversion, and milling! Tweaked my tool change commands, worked perfect. 0.8mm drill bits are great, the offset V-bits…I don’t understand them yet, ended up using the one from the shop.

    Those of you with a keen eye…The traces are on the wrong side, they need to be mirrored (or designed on the bottom layer), but I only had two sided PCB so I can’t use it anyway….yet. Two sided PCB milling is going to take a seriously accurate flip, like 3 pins at least.

    1 user thanked author for this post.
    #61505

    Ryan
    Keymaster

    The dimmers will be in the shop on the Wed the 13th, when the new insurance takes effect for those of you whom might want to play around with it… a dev kit if you will.

     

    1 user thanked author for this post.
    #61519

    Jeffeb3
    Participant

    Three pins because there is too much slop in two pins?

    What about milling a corner in your spoil board? You wouldn’t want to depend on opposing sides of the pcb, but if you flipped it, keeping the corner in the same spot… No, that’s crazy.

    #61528

    Ryan
    Keymaster

    I don’t think I could make a corner will work as reliably as two pins (three is just an exaggeration).  I don’t think I will ever make a bunch of PCB’s but a two side one would just be a cool trophy to bring to MRRF (or some other shows, I need to find another one)…. If I made more PCB’s, Surface mount and a few bottom side wires is okay with me.

    I would also like to try some of the other PCB’s suites, If I could get a direct DXF out that would save loads of time. The SVG output needs to have stokes converted to paths and then each trace combined/joined with the through holes.

    #61532

    Jeffeb3
    Participant

    Make sure you take some pics. In case you want to make a tutorial.

    #61533

    Jeffeb3
    Participant

    I don’t think pcb milling is advantageous for anything more than a few. It really is only useful when you’re doing iterative versions and you want to test each one without waiting weeks between versions.

    #61534

    Jeffeb3
    Participant

    Also, I found out my local train store (shout out to caboose hobbies!) has several tiny drill bits. If you ever need one quickly. I bet these would work great for the drilling pcbs. They also have some chucks that I assume fit in a normal drill, but can hold the smaller bits.

    #61537

    Ryan
    Keymaster

    don’t think pcb milling is advantageous for anything more than a few. It really is only useful when you’re doing iterative versions and you want to test each one without waiting weeks between versions.

    Through that site they are $2 each, with customs masks and printing, the only downside is it take a few days. For sure no one should attempt to do more than a proof. I am just a little worried about the noise so I wanted to make sure it that cap still works on a board.

    #61589

    Jeffeb3
    Participant

    WDYT? I don’t have any white out (the bottle I keep in my office drawer is from the 90s and it’s just for street cred). I used kilz primer instead and a sharpie. I’m the slowest ever. I get 15 minutes every other day or so…

    #61593

    Ryan
    Keymaster

    Much better than mine.

    #61691

    Ryan
    Keymaster

    Okay update!

    It works great so far. I just did a few test cuts. I can not begin to tell you how nice it is to have this much control. I ran my first cut at 10k, 10mm/s, 5mm doc, single flute in pine. PID worked great but way to slow for this cut. Next bumped to 12k…then 20k, ahhh much better. Then I found an estlcam bug, it does not pass off the finishing pass speed change (unless I deleted some setting).

    At 10k the router is really moving slow (I wouldn’t try to go under 8k I don’t think) I fully expected it to bog down, the CNC didn’t like it. I was beating it up, but it kept pushing the router through. The spindle held its speed great from what I can tell by ear. I will have to really look at the real data output from the Nano but I just wanted to use it already.

    So high five everyone, so far so good!

    negative-I think the router blips (~1sec) when the Rambo is booted but I have to triple check this out. Easily solved by not having it’s switch on, but it just shows having full control I forget to even use the dang thing!

     

    #61694

    Jeffeb3
    Participant

    negative-I think the router blips (~1sec) when the Rambo is booted but I have to triple check this out. Easily solved by not having it’s switch on, but it just shows having full control I forget to even use the dang thing!

    That’s a problem with the enable pin from the rambo not being pulled down when the microcontroller is starting, right? You could add another pull down resistor, so that it will get pulled down when it’s floating.

    #61695

    Ryan
    Keymaster

    I have to double check. I initially after I moved it all to the shop I plugged in the Rambo output ribbon upside down…..for testing purposes 😉 ….so power to the wrong port didn’t pop either board. After that I am not sure if it happened or not, I honestly can’t remember. Adding a resistor is easy enough, thanks, for that.

    #61857

    Ryan
    Keymaster

    I tried to stop it….but it just works too well. 😉 (p.s. This was the very first time the PID ever made a real cut.) Weird video artifacts from the crazy vibrations. If your cut ever sounds like this…stop it! This cut works perfect at 20k.

Viewing 30 posts - 331 through 360 (of 459 total)

You must be logged in to reply to this topic.