The graph on the interface could be used other functions. For instance, plotting spindle load on a second axis with spindle speed. The hardest part is likely getting the information into the TI board. After that, only a small amount of code is needed to display the graph.
This is the best project I've seen in months. I would like to see an Override Button with Plus and Minus Buttons, under the graph on the HMI that would be enabled/visible only in the feed mode. After touching the Override Button (change to red), touching the digit/decimal place on the HMI would set the increment or decrement of the feedrate by 1 of that placeholder. Touching the Override again (back to default color) would set the "increment/decrement by" back to zero. Really nice project. It's all way over my head, but I love it!
This is why open source is beautiful, someone put an idea out there, and someone else built upon that idea. I'm going to order parts and hopefully contribute as well. CNC lathe doesn't do anything for my work flow, but CNC aiding to avoid gear changes is awesome.
I was bitten hard recently relying on a closed source program to collect results. A bug caused all the data to be lost. Lesson learned was the most important tasks need to be verifiable and simple.
@@kentvandervelden Sorry, can't wrap my head around this but would it be possible to thread away from the chuck with the programming of the ELS/touchscreen as it is now? Right hand turn, but away from the check for added peace of mind. Also, I actually liked the drop down items of your first version as it's easier to make a quick selection of a feed rate from that rather than typing. Is that function still there? Best would be to have both the drop down - potential with favorite settings storable - plus the "calculator" mode. Finally, I also liked when the enable/disable button was more prominent as in the 1st version. Maybe it would be possible to have skin where it took over the graph area? I know I am asking a lot, but just thinking out loud here.
@@AntiVaganza RH-thread + LH-feed: This may be possible with the tool mounted properly. Maybe upside down or mounted on the backside. Drop down menus: Here's a link to the version with the dropdown menus: github.com/kentavv/electronic-leadscrew/tree/feature-nextion and a the link to the arbitrary entry version: github.com/kentavv/electronic-leadscrew/tree/feature-nextion-pure I don't have the hardware to test with so I'm no longer making changes to the code, but others are welcome to submit improvements. Hope these suggestions help you. Thank you for asking.
Hi Kent. I have James' ELS set up on my lathe. This development very interesting and clearly a lot of work has already been put in. I'd like to suggest a hybrid arrangement. Your user i/f is much better for setting up parameters than compiling code and the display of the settings may also be preferred by some. Machinists tend to have oily fingers which is where problems with touch screens arise. But this work, linked to a 'in use' physical button interface would solve that problem. Set up available options with the touch screen and use it to display the option(s) selected but make the selection with physical cursor keys, keeping the display screen clean when the machine is in use. We can always wash our hands before changing options etc. BobUK
Thank you Bob, that's a great idea! Since the video, I added a display of rotary position, in degrees, dec., min-sec., counts, etc, which may be useful if lacking a dividing head. Implementing your idea, could be done either using a GPIO into the Launchpad or into the touchscreen (and avoid the C++ code.) Full disclosure: I'm not planning to do more work on this, but your suggestion will be added into new projects. Best wishes, Kent
As far as the oily fingers go. glass or plastic tablet covers can be used with these types of displays. This would protect the display from oils and impacts. I use this brand Supershieldz and been pretty happy with it on my phone and tablets, which I have used with my hands oily and dirty. For this display I would say a Playstation Vita PCH-1000 protector should work for it. I ordered one for this project. Once I have than delivered, I can then start designing and building a cover on my 3d printer.
Great video and great work on this touchscreen! I like your future ideas of a pot on the bottom left .. also, could you do "more" screens? I'm wondering if one could emulate the actual feeds and speeds plates of any given lathe and input it as an additional screen that would allow a user to "nostalgically" revert back to "their" feeds and speeds plaque on their lathe as an alternative to the highly custom manual input .. Does James known of your mods? I think he should ;)
Thank you! The Nextion can have many screens. The rough way to add a screen is to draw it up in a graphics program, highlight regions that are touch sensitive, and then add code to response to the messages delivered over serial. One could certainly do this to create a lathe make-model specific skin. A creative graphic designer or UI designer could a lot without touching much code.
Geat work! A touch display is nice but I would prefer to have critical controls on real buttons. I would appreciate if there would be a possibility to have a momentary start/stop switch so that you don't have to control the movement of the tool by engaging or disengaging of the half nuts. Moreover, a manual override of the feed speed would be very useful to dial in the feed on the fly. Therefore, a rotary encoder would be nice. For these functions it would be nice if you could access them on the touch screen and have the possibility to connect external switches and a rotary encoder.
All great suggestions! I put links to other ELS systems in my notes (in the video description). One uses a rotary encoder for most of the features and it's a very nice interface. Someone had interest in writing a feed override and I made a video going over one way to do it. That's probably all that I know. Thank you for commenting.
Great indeed fantastic project. You could add two more engines to make it all automatic, the board would be the same or should you do a new project? Thank you so much the videos are very intuitive.
You want to add two more motors? You may be able to start with the existing board and software, but heavy changes would be needed. I would prototype with LinuxCNC, make a custom display, attempt to reduce to a smaller size. The time cost of a custom ELS is higher vs. a CNC system that's more functional. Hardware costs are similar. Hope I understood your question correctly at my opinion was helpful.
just found this. Fantastic implementation! I particularly like the graphic of the part and cutting tool. What is the need for the graph? I see previous comments echoing my thoughts, but you must have had a reason for it. Kudos. I am in awe.
Wow, thank you! The graph is an experiment, to see if a trend line is useful. Spindle speed may drop if the motor starts getting bogged down. Maybe temporarily or maybe its regular. For the visual sort, who can't do math in their heads. Count me in that group :)
This screen would make it possible to loop a position sensor (from a dro set) to the TI main board and control the spindle speed based on cutting speed (handy in facing operations) as well as some level of speed control under load, when used with a VFD controlled motor. Through this setup it would be easier to set the cutting diameter. And have the CPU compute the output for the VFD.
That's a very interesting idea! I'm not sure that the value of a ELS makes sense when so close to full CNC. But a tool like you're describing, as a standalone, could be made pretty affordable, and the results especially nice if used with power feed. Thank you
@@kentvandervelden the cross slide would still be operated manually. Just that the VFD gets added to the control to set, and maintain speed. With the speed regulated by the working diameter.
@@FlexDRG Sorry, no I don't. I've had to buy some stuff on eBay at inflated prices, because everyone is sold out. E.g., last I looked RPi's were 3-4x list.
I had to look up GTFO :) Open source can be pretty cool. Probably not technically identical, but ever heard Lawrence Lessig talk about remixing? Thank you for all the support!
@@kentvandervelden I hadn't, but a cursory read of the Wiki entry, and it indeed sounds about the same. Luckily, for most software, we've got a number of anti-licenses to chose from that make forking and pushing changes back upstream a common practice. I can't imagine a band ever including a fan's cover of a hit song on their next CD. Hahah. And that's really a shame. I'm happy to see you got a PR open on Clough's project. I've got enough projects of my own that I won't be digging into the code any time soon (eg. I'd have to get my lathe functioning first... hahah), but I'm very glad to know there are references for this work (and the opencv/HAL modules you've done) that I can draw from and improve on when the to-do list gets shorter.
@@swolebro If you run into something where a little coding might help, let me know, and I might be able to help. But I'm not testing pepper sprays, as curious as I'm :) NiN might have experimented with fan covers and they released the raw multi-tracks to at least one of their albums. Probably the only one big group to have done that. Here's a great (lo-fi) video of LL's style and the subject of remixing. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-7Q25-S7jzgs.html
hi I'm a little stuck I have the same setup as you did in the video but the screen doesn't seem to be getting signals from the board is there something i have missed on the board I uploaded the code to the board i have tried the code with Support for the seven-segment display but the screen still didn't work but the seven segment display did any help would be greatly appreciated
Completely guessing... could something be off with your serial setup? Do you have means to test that serial is being sent and from the display? An oscilloscope will let you know something is happening. One with serial decode will let you know messages are not corrupt.
@@kentvandervelden just wondered if I had missed a switch or jumper that wasn't set in the original clough els I don't have a means of testing if I can't figure it out ill have to go back to the original els for now thanks for the quick replay it is much appreciated
@@zerot001 No jumper changes that I remember. I don't see anything about it in my notes or in the SCI block diagram. I'm sure you've already tried the simple stuff: common ground between Nextion and Launchpad, triple checked the right pins are used on the Launchpad for serial, tried reversing them just in case, and checked that the right baud rate is used on the Nextion. Another possibility is if your board has a different main frequency than the one that I used (different revision?), serial might be off if using a divider of the main clock. Sorry I couldn't be more help.
Hi Kent, in your "Touchscreen for Clough42 ELS notes" you mentioned in "Suggestions for improvement" that the TI processor has a coprocessor, the Control Logic Accelerator (CLA). Do you know if this CLA is now used and neccesary for the ELS to work in James and/or your versions?
Hi. Going through all the comments as well as all the documentation. So I also ended up with the version where the 7 segment display is also working so I could ensure all is working correctly. The problem is that I cannot get the Nextion display to work. So I am now open for a couple of next steps. If you or anybody still have a compiled version that was done with the old CWARE and where the screen was working I would really appreciate if this can be shared as I do not see it in the GitHub files. I have now in the meantime ported the code and everything else to the latest libraries but the problem stays the same. 7segment and servo is working but Nextion does not work. Any help for a compiled working .out file would be appreciated for my troubleshooting process.
Kent, awesome video and useful expansion of Clough42's implementation! For those of us who are coding-deficient, could you explain more about how to add this to James' ELS? Is it an addition to or does it replace the Clough42 code? Thanks for your efforts!
Thank you! This is a modified version of the original. If you built the original using Code Composer, you'd follow similar steps with the modified code. You can find the modified code at the following link. As a fallback, save the original binary that you loaded to TI board, which can be reloaded if anything goes wrong and save you the time of rebuilding. Wishing you the best of success! github.com/kentavv/electronic-leadscrew/tree/feature-nextion-pure-dev
You could add the limit switch to the original project. You'll need to add a small bit of code. I recommend using a proximity switch or interrupter instead of a mechanical switch to simplify debouncing.
Hi Kent.. this is a little off topic, but... were you the one who made a "V" bracket and mounted it between the Z column and the wall to stabilize the column? ie: bolt the small part of the "V" against the column and the wide part against the concrete wall.... and it really made a diff in the column flex!... if you made this vid please let me know which one.... Thanks! ( if not... try it! ) Phil
Hi Phil, thank you for thinking I might have come up with that :) A couple other people have told me about it and I have a few links. I'll send them this evening.
Hi Phil, I had a couple video links. The best implementation is probably ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-X_6i4SN05HU.html But a close second, with very interesting discussion is ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-JwzA67_-cAM.html If you make either of these, I would enjoy hearing your thoughts. I was warned by someone who tried bolting to the wall, that the performance may be better but the noise might be worse. Best of success to you.
Hi Kent, my good working Nextion display glass broke and after I uploaded the same firmware into a new identical Nextion display it doesn't communicate with the TI-board. In the Nextion Editor debug mode I connected the display and could send/receive commands. The Nextion display seems to be good but no connection to the E.L.S. Any ideas? Thanks.
I have the same problem. Spent about a week on it trying to figure out why the board was ignoring commands from the screen. I eventually gave up. And unfortunately, it seems as though we're not the only people who have experienced this. Give us a heads up if you figure it out, and I might have another go at it, because right now, I've lumped it into the category of a failed project.
It's possible, but the cost would need to be written. Given the RPM, Alarm, etc. passed to the make-model specific touchscreen C++ class, return the current feed, pitch, etc. If you are interested, drop me an email and I'll work with you to make the interface to the touchscreen generic to your code alongside the Nextion.
Hello Kent, can you share your design in the nextion editor I would like to see how you did the switch between turning and tapping. I can't handle it Regards Jacek.
Hi Jacek, all the files should be on GitHub. Let me know if something is missing. Here is the direct link to the Nextion file: github.com/kentavv/electronic-leadscrew/blob/feature-nextion-pure/els-f280049c/nextion-els.HMI
Hi David, multi-start threads are not provided by the ELS. An ELS is to help manual turning by providing a programmable gear ratio between the spindle and lead or feed screw. The stock ratios are otherwise limited and may be inconvenient to change otherwise. To cut a multi-start thread on a manual lathe, one way is to index the part. The rough steps are: cut the the first thread, stop the spindle, disengage the leadscrew, rotate the spindle and part to the next index (e.g.: 0, 120, 240 degree for a three-start), re-engage the leadscrew, start the spindle, and cut the next thread. A lathe might have a graduated ring on the spindle to help make this more accurate, but small lathes are unlikely to have that. Your question is a good one because the answer shows how an ELS could help produce mulit-start threads: show the current rotary position of the spindle and fill the role of the missing graduated ring. Pretty interesting... maybe I'll make one more change...
This video shows a few different ways to cut a multi-start thread on a manual lathe including a graduated ring, slipping the gears, and a dividing plate. Pretty cool. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-4OmD0bqxpXg.html
Just added spindle rotary position. With a 4096-count encoder, degree-minute is the best that make sense. I may make a YT Short this week about it. Thank you for the idea :)
