Don't let people convince you that a touchscreen is needed. The best interface if the most simple one that will do the job. Fancy touchscreens sell products but aren't necessary a lot of the time.
I don't see why the fascination with Touch Screens. You have designed a beautifully simple and robust User Interface so stick with it. Remember - KISS!
I good touchscreen will have an equal or better life span then a mechanical button. But as a machinist the tactile feeling of buttons. Is very very important it lets you find and work the controls without taking your eyes off the work piece i like the other guys ideas about rotary switches. I also like the idea of a hand encoder for detail work.
Yeah touchscreen is nonsense here. LED displays are much more readable from a glance and real clicking buttons make for a much more satisfying and reliable interface.
This is actually better than a school lesson because you can actually see what's happening and how to solve the problem or at least a way to solve it. I'm definitely going to watch all the series and I don't know yet but I'll give it a thought about making an ELS for my lathe.
Excellent progress! Please consider the following: Do make kit considerations. I am too old with eyesight gone and memory shot to re-develop the work you are doing so well. I am sure there are few others who, for their own reasons, will join me here. Once you agree, I will swamp you with several "requirements" 😊 Thanx!
Here's how good the videos in this series are; I don't have the slimmest clue what you're talking about other than speeds, feeds, TPI and it's for a lathe but, I watched through the WHOLE thing and found it thoroughly enjoyable!
It´s just perfect, all this tinkering is what a good process is all about. I´m really blown away how good you are able to explain quite complexe stuff. Cheers
All the “”wall wort” power supply give you noise are topically half wave rectified, with no filtering. Battery’s under charging don’t need capacitance. “Because the battery essentially is a big capacitor” So take that nice scope and look at the out put of the wall charge it’s a lot of bumps all positives but humps. Every rising or falling part of the wave is a count. Put a 500-100 uF. Across the output of supply, it will get better. Keep up the good work J Hicks
Since you have innovated this yourself and most of the commenters are sitting on their hands I would not worry too much about their criticism. You obviously know exactly what you are doing and as you finish it and use it you will figure out what works well and what doesn't. Ignore the comments. Oh wait, then that means you should ignore this too. Ok, so ignore all comments EXCEPT mine cause I'm always rite.
Very nice job! I just bought an old mini lathe, and the leadscrew wheels are missing. Your project is exactly what I need to bring it back to life. Thank you very much for sharing your job.
"Absolutely briliant". You are an awesome genius dear person. Cant wait for each new episode. If you will indulge me with a bit of trivia: When "transistors" first came into "electronics", I learned and then later taught the theory behind it. This was decades ago of course; when ALL we had was germanium PNPs. Then we all went crazy when silicon NPN's came in, long before chips. And of course, as technology got on a "faster than sound" jet plane and has not landed since; we have gone into a totally new era in mankind. Amazing for sure. But of all the education and training I got; it was NEVER as good as what you do naturally when you teach. A tip of the hat to you kind sir; for you are truly gifted. Keep it coming. If a person is in electronics, and really wants to know 'HOW", these are the videos you should watch and listen to. I would give ANY thing just to ask you just a "few" billion questions LOL. I have NO doubt you would KNOW every answer. and teach it in a way that "even" I could understand Oh yes Siree "bobwizer"! WOW!
I'm very glad there are EE people like you, because I was about to take a very long walk on a short pier. Long story short, software engineer downloads Eagle, makes a ESP32 based irrigation system, order PCB from china, solder stuff on, connects full rect bridge supply and I2C bus spits software dude in the face. There is hope now. I just have to call an EE guy from uni. Thanks :-)
I get the feeling you don’t need to listen to the armchair experts - great work, really keen to see your project up and running. I’m sure this will become a commercial option very soon. Good on you for pioneering it.
There should be min. CS hold time called in the data sheet of the display driver. I would reduce the bit rate because the signals do not get to full HIGH. Great project!
Clough42 => Rotary switches for the interface would be nice. Then you have an interface that looks and feels like a gearbox. ;-) Also you can easy switch the buttons with gloves full of dirt. One for inch / mm; One for thread / plain cut; One for direction; An encoder (with heavy wheel for velocity) 2 select parameters / jogging; The LED display is perfect.
Negative on the encoders. While I agree with your comments that they are nice, it would require even more components to interface with them and I think the indication from most comments I’ve been reading are that he should try to keep this as simple and cheap as possible, and I agree with that as well!
@@EdAgers110 I understand that the second encoder sounds complicated, but it is not. It only must / should work when the lathe runs in plain cut mode (and not in thread mode) or when it stands still (for settings). You can simply switch between the endoder that is connected to the spindle and the encoder on the UI. You can use a ADG3257 from Analog Devices for example. If you are in plain cut mode, the controller switches only one bit on the select line of the ADG3257 and the handwheel ist active instead the encoder on the spindle. The same as the lathe is stopped and a menu button is pressed. It adds only a few dollars to the bill but adds a lot of useability to the project. It is in my opinion worth it to invest some extra dollars.
The encoder can be really great, but it can also be a liability because it's easy to bump accidentally. Careful design of the user interaction would be required to make it work well. I have encoder menu inputs on my 3D printers, and they're not very precise. I would have to get the feel right to be happy with it.
I did this just last week! I was tearing my hair out over why my board wasn't working for ages, finally broke down and soldered some fly wires to vias so I could hook up the logic analyzer. Saw the glitch, went "whaaa?" and reread my code. Oops. As far as your refresh rate goes, it's massive overkill. 15Hz is the "fluid motion" threshold for humans, and even 4Hz is basically unnoticeable for number updates. It could even help, because it lets you _read_ the numbers instead of just having them all blur together. Giving yourself that extra margin at low clock rates will also give you better noise immunity, which might be important with the big lathe motor going and the stepper having to exert a lot of torque. You'd be amazed at how much interference those steppers can generate. It might be overkill, but one other thing you could put before your input filter is a discrete TVS (transient voltage suppression) diode. This can clamp transients of thousands of volts down to tens of volts, which is much easier for the input filter to eat.
Great project! :) That software bug was a bit surprising, but educational . I would have went down the power supply rabbit hole...and probably lost my mind it the process
Good for you! Anyone who has been through a transformation like this will tell you that it's hard, but it's also easy in a weird way. If you're debating in your head every day whether to have a burger and fries for lunch, it's a horrible grind. But if you come to the place mentally where you have decided that being healthy is more important to you than the enjoyment you used to get from food, something changes and it becomes a lot easier. I don't know how to teach someone to flip that switch, but once you do, it's just a matter of time.
@@Clough42 I totally know what you mean. I think for me it was just time to do it. My clothes fit better, climbing out of the hatch into the garage where the machines (lathe/mills) are is easier, no end of benefits. One day at a time. On a side not I'm looking forward to upgrading my 602 with your kit. I bought the replacement motor years ago, well before your channel started, but never got round to doing the deed. Have all the parts coming in the next few weeks and looking forward to getting that project done too. Thanks for all your hard work.
Fabulous video - as are all in this playlist. This is one of the best videos an tracking down a bug that I have seen! Everyone so far is a gold mine of well explained analysis. I will continue watching the rest in this series. I do not have a Grizzly but the principles explained here have many applications.
Really excellent discussion of the electronic debug process. 20 years ago, I has almost the same exact problem bit-banging Dallas Semi 1-wire protocol. I had a very similar Tektronix scope on my bench (except it was almost $5k!), and plowed through it. Great stuff! Thanks for this entire series. I have a project in mind for your design...
For some reason I always knew these bidirectional converters as "nifty two-way but inherently too slow for anything that is actually fast", so I was surprised to find zero mention of this aspect anywhere from Ebay to Banggood and even Sparkfun, including the linked many-page Philips app note analysing the thing from top to bottom. Not a peep. One has to go all the way to Adafruit to find this immediately addressed in their product description: "the 10K's do make the interface a little more sluggish than using a TXB0108 or 74LVC245 so we suggest checking those out if you need high-speed transfer". So, yeah... nice little gotcha about not wiggling things too fast, apparently not nearly as commonly advertised by those who sell this as it should be.
Great project. Thanks for sharing. The pulse problem has probably to do with the voltage. The phone charger has probably a higher voltage around 5.2 volts. The USB has usually around 4.9 volts. When the chip triggered around half of the power supply voltage the pulse is high enough to trigger at USB and just to low at 5.2 volt.
These USB phone chargers actually fluctuate between 4.5 and 4.75V. I use them to run esp8266 devices that log voltage, and it's really erratic. Won't be buying any more of those.
I have often wondered about the i2c levels and how much the waveform can be distorted and still work. I didn't think it could function looking like that, so now I know that I don't have to be so picky about it. Thanks!
Makes me wonder if the pro touch screen crowd has ever worked a lathe. Oily hands with metal shavings and dust isn't going to fair well with a touch screen. I like your push button box the best so far. Good eye on catching the glitch. Brilliant!
Hi James, thanks for the explanation (it's right on the outside edge of my electronics understanding) and the hard work. Please don't get distracted by comments and come up with an end result that no one can afford. Unless of course you decide that's the direction you want to go. It's your project. I would prefer something that does its job. If I try a deep cut in stainless with a binding leadscrew that's on me. It isn't the job of a hobbyist leadscrew controller to make up for my every bad practice. There are already cnc controllers with servos that do those things and they are too costly for me. Just another opinion by the way. Keep up the great work.
so far a great build. Need to get some red transparent film, I picked up red green. and blue sheets off ebay , makes a world of difference. I prefer surface mount now even though I cannot see without magnifiers, just use a stencil, place and bake
James, there is one thing that you have to keep in mind if you plan to make this commercially available and that is patent infringement. There was a commercially available version of the electronic lead screw on the market in the past but it was withdrawn from the market and the patents were sold to the German company that makes Wabeco lathes. Also there was a Canadian fellow who was developing an electronic lead screw and selling kits. I believe he has a Google groups site and he may or may not have patened his idea.
Fantastic idea for starters. I really hate my lathe banjos and gears. Major concern is of course any asynchronicity between spindle and lead screw due to, not SW or HW, but physical forces strong enough to throw the stepper motor off. In a physical gear connected configuration, lathe motor could halt but would always keep spindle and lead screw in sync.
The effect of the pullups is less than expected because the MOSFETs are on for a good part of the rising edge and allow the driving end to charge the capacitance quickly initially. It looks like the SPI outputs are tri-state rather than true open drain. If the MOSFET has a low gate threshold it can be conducting all the way to the switching threshold of the receiver and the pullup resistor only has to pull it the rest of the way to 5V.
James Great idea making the lead screw electronic! There is one more thing that the project needs. Use the encoder to give degrees of rotation to be used as an indexer, please!!! You might even consider adding another motor to rotate the head x degrees on key press. Looking forward to your thoughts on this.
*I Changed My Mind* I was one of the ones pleading for touchscreen but I have been thinking about it and honestly, I now think the most important thing is a tactile feel. Bear with me as my lingo may be incorrect, but I am hoping you will make it possible to work with the half nuts always engaged through a threading operation. On the very small mini-lathes, that's how a lot of us thread (keep half nuts engaged and reverse the lathe before each threading cut). Now, with your controller, I imagine a new way to do it where buttons on the controller stops the leadscrew, so you would hit a "Stop" button when your cutter got to the relief cut. Then back out on the cross slide and hit the "Reverse" button on the controller (which could be set to a faster return speed to move the cutter back towards the tailstock) and then hit "Stop" again. Finally, for the next cut wind the cross slide back in to cutting depth and hit the "Forward" button on the controller and let the box act as the thread dial indicator and engage at the right time. If this sort of programming would be possible, then the touchscreen would not be a good control interface. Tactile buttons would be needed to so you could keep your eyes on the part and not the screen.
Seeing you use the Bud boxes gave me the nudge to get one for my recently-acquired drill press. I want to make a safer and more versatile control for it besides the ancient flaky toggle switch. The investigation into the power supply issue was cool too, with the evidence shown in the scope traces. I didn't realize your typical wall wart was that noisy.
This project is so great. I really appreciate the careful explanations. You obviously have a sound understanding of the underlying principles. Thank you so much.
Yes, I need one! Agree about using a touch screen, but personally, I'd like to build a more sophisticated display for my needs. I have a Hardinge turret lathe, and you have most of what I'd like to make into electronic carriage control for manual operation. Phenomenal work - I'm looking forward to catching up on your project videos, looking at the code (nicely commented btw), and making it work for my lathe.
By the look of the datasheets, you shouldn't need level shifters, just a 3.3V pullup. The TM1638 is low voltage interface friendly by interpreting anything higher than 0.7V as logic high and anything below 0.3V as logic low. Just ensure that there is no pullup on the side of the display/button board, or delete it if present. Then again with moderate baud rate and decent delays, level shifting is fine. Good one on finding the STB bug. Nice generous STB drive now :D. Conveniently, most of the flank distortion of the signals introduced by capacitance sits well above the 0.7V threshold voltage that i mention above, so the signal is really more legible than it looks at a glance. For static protection, you can add diodes from signal to both ground and supply of the microcontroller. 50pf or 200pf is not a real static protection measure.
I had wondered if I could get away with driving the display at 3.3v. The hangup is that pullup resistor on the board. I am really trying to make this something that anyone could put together, and requiring a tiny resistor to be removed from the stock board is something I wanted to avoid. TIme will tell if that was a good tradeoff.
Great Video, James. I am on the outskirts of understanding, as am a retired toolmaker. I am ready to start buying components, and making this project happen though.
Looks like it's moving along well, nice work. One note, the 138 has a much lower working voltage than the 7000's series mosfets. You might want to check the data sheets and gate transfer characteristics, it might make a difference in your level shift.
Really enjoying this series thank you. I was thinking of fitting something to my lathe when I saw Stefan Gotteswinter (another UA-camr) add a commercial German made ELS to his lathe. However it was many hundreds of Euros so I lost interest but what you are building is probably less. Great work and well explained !
Same here. It was also German-language-only (at least at the time). That's when I lost interest and started thinking about building something of my own.
Good sleuthing. Since the interface is only being refreshed every 50ms and you don't have that many bits of data to send & receive during that time and it doesn't have anything to do with keeping up with the main loop you can certainly utilize a lot slower baud rate and take advantage of extra timing margins. An LCD touchscreen is an option. The LEDs are nice and bright, my preference. If you integrate the controller and display in the same enclosure then the only long wires you'll have to contend with are the stepper drive signals. Perhaps put a resolver on the leadscrew and look for missed steps. So far, I love the digital concept of controlling the leadscrew. My bet is that lathes will probably soon move in this direction. Singer sewing machines moved into this digital direction a long time ago. Just keep in mind minimum setup, hold times and levels and you'll do all right. Keep up the great work.
If I were designing a commercial product from scratch, I would put the microcontroller and the display on a single PCB and put it all in the control box. This particular dev board is a little bigger than I wanted for that kind of construction, but it's an option.
I worked at a TV studio for a few months in the analog days and an engineer explained to me that on long cable runs they would put a pre emphasis/eq on the video signal so when it got to the receiver the loss was compensated for and you got the signal you wanted. You don't do it on receive as you get noise problems.
Unfortunately on the data line, both ends are sending and receiving data. It would also be highly unwise to anyhow drive data, clock and burst lines in a different manner, because just as long as signal timing is generous and the distortion is equal, it will work fine, but as soon as you introduce differences in distortion, you can get into trouble.
I just stumbled into this series and I have encountered similar problems when trying to convert an elector-mechanically controlled machine to a computer controlled system because electronic switches are so much faster than mechanical switches. Time has to be a consideration. The computer chip sees the switching as instantaneous compared to mechanical contact switching. We also encountered it when trying to add a proximity switch to stop an upstream feed conveyor and alert a machine operator when a component stopped turning. The short interval resulted in many false stoppages as the component speed varied by load. If the it slowed for an instant, it stopped the conveyor. All it took was increasing the time allowed to get the required counts.
I would love to see a motor on the encoder and a tachometer on the stepper to validate the math, possibly add in a method to offer a mechanical load to the stepper to see at what point the full feedback loop would be necessary. Just a thought. I can appreciate the issues of signal integrity as it relates to serial clocking and the edge to state alignment. Nice work!
Great job using an oscilloscope to check signal propagation. Suggestion: You should flash the RPM display to remind people to stop the lathe when switching directions.
Almost as soon as I turned off the cameras, I thought of just decelerating and accelerating the leadscrew smoothly. That would save a step and make it much easier to use in the case of "oops, it's going the wrong way".
Loving how this is developing! Would absolutely buy this as a kit when ready and available ❤ I think the simple display and buttons is fine - touchscreens can be nice, but I'm not convinced it's necesary or advantageous.
*Touch Screen Thoughts* I know you mentioned in a previous video, that a touch screen would be a bad idea in a shop environment. I still think that it could possibly work. It's supposedly not that uncommon on high end CNC controllers (I think Mazak has one with touch screen). I'll admit that my iPhone is horrible if my fingers are just a tiny bit moist or greasy, but I have had that exact same phone in a super cheap zip lock style bag for diving trips and even with wet hands and gloves on, for some reason I have been able to jot down GPS points when I got back on the boat. Just saying, there must be some touch screens they work well even in an workshop environment. Maybe even some cheap solutions. Either way I, can't wait to get my hands on your board and make this upgrade:). Great to hear you will test servos also. Since you are making the rest of the controller so "sturdy", I think a closed loop for the leadscrew makes a lot of sense. Actually, apologies up front for the crazy curve ball here, but is there any way this could be made to work as an APP on, say, Android...? I did put scales on my little lathe and used Yuri's TouchDRO firmware on a cheap TI Launchpad. How amazing it would be if there was a way to merge his and your features into one APP. I know you might have plenty of very good reasons for not doing that, but no harm in trying to respectfully plant a seed.
The smartphone app idea is definitely possible, but probably out of scope for the work I'm doing. Priority one is getting the core controller and a simple UI working. It could eventually be configurable to use different display hardware, and a bluetooth dongle with an app for the display seems like a reasonable option.
I like that black plastic better. You should be able to print your lettering into the front and even do so in another color. You just need a simple 5-12v transformer that isn’t an actual charger.
Nice progress! One of my interns has been using 3d printed molds to make silicone buttons using hardware store silicone. They feel great and are individually collapsible buttons in a sheet format. I can send you OpenSCAD files to play with if you'd like.
The issue is the sink current of the microcontroller output, which is 3 or 4 mA. There are already three 10K pullups in parallel (one on the 3V3 side, one on the 5V side and one on the display board. So we're already at half that. I may play with it a bit.
@@Clough42 Since these lines may be uni-directional why not use opto-isolators to achieve the level shift. You prevent the micro-board from being fried too.
For the enclosure put a thin membrane (rubber/silicone?) between the buttons and switches. It will stop oil and dirt from entering the PCB... Maybe sandwich it between the enclosure and some support structure.
I believe a conformal coating (which is a specially formulated acrylic for PCBs) wouldn't go amiss either, if oil or some detritus ever did make it's way into the box, the coating will prevent a short circuit from occurring.
The cure for capacitance on I2C and SPI signal and clock lines is to make certain all your inputs are configured in Schmitt trigger mode. Or, if you're bit-banging the input (or using an FPGA that doesn't support Schmitt trigger mode), use a software glitch filter (for example, sample the input at 16x the clock and ignore transitions that aren't steady for at least 8 samples). If you think about it, an I2C or SPI signal line with a lot of capacitance (either distributed or bulk), which goes off-board (or all over a board to different power domains) isn't really a digital signal line. It's an analog signal which might be very noisy or have transition times which are too slow for reliable detection with non-Schmitt inputs.
This is good advice. The TI DSP chip I'm using has a sampling qualification mode, but I think in this case, the SPI peripheral is actually conditioning the input.
The signals are OK for development version but I would suggest a driver and receiver chip for the signals because eventually you will get caught out. That has been my experience at least.
You're not wrong. In this case, since it's a bidirectional signal, I would need a driver at each end. If I'm going to spin a PCB for the control panel end, it's worth thinking about just doing a whole control panel board.
The lost signal by Volatage drop is a really old problem. The Voltage breaks down when the thyristor/transitor is switched. A good power supply can handle that. But cheap charger don't compensate, small surges of power can make the voltage drop in the hole system. Two ways to counteract that. 1. Use over-voltage 5.5V by example. 2. put big caps ( 5v to ground ) near paces where surges can happen. By example near the thyristors. I'm designing/building circuits for over 20 years now and had such problems often in the past. Modern power supplies have much better voltage regulation so the problem is not so prevalent anymore. The first time i had such a problem it took me weeks to figure it out.
Hi, that was a very good and well presented masterclass in debugging. I'd like to suggest, if only for my own benefit (!), that at the end you consider either making the final (is it ever?) compiled code available, or link to or demo how to go about doing a compilation. At the moment that part looks like a 'black art' to me! I've looked at the reference for the source code - which I can sort of read - but would have no idea how to put it all together. I do have a handy son who understands these things though, but not everyone will have such a handy child or knowledgeable friend. I'd like to confirm that should you choose to make a through hole pcb board or a full kit (less the heavier shipping items like the motors, but perhaps with an assembled smd board) available at anything resembling a reasonable price - including recovering some of your time and development costs - I would order one. This is perhaps the most useful add on to any lathe - even one with a change gear box - that is available. The ability to instantly switch between US, imperial and Metric thread standards is invaluable to much of the model making world outside the States for reasons that are a bit too long to go into here. If you do go down that route please endeavour to make sure that your selected supplier or fullfillment centre can both cope with international orders and do so without silly shipping costs - some sadly price themselves out of the market. BobUK.
Hello, project looks great! I do not know, if you mean real servo or a closed loop stepper like this: JMC 2-Phase iHSS86-80-70. I would be more then happy to build one and test along with you on my lathe (TSB-20 polish small lathe). Also I have an idea for additional feature: add another encoder input to to replace carriage hand wheel. Button or switch to disable spindle encoder input, this way you could replace lead screw with ball screw i.e. SFU1605 or even SFU2005. You could wire alarm output from motor drive to detect motor stall due to collision or use serial port in JMC drive.
how are you going to manage threading passes? it looks as if you would be able to just reverse the spindle to reverse and let the carriage come back to the start of the thread, but it would be pretty cool to just hit a return button, or the reverse button, and have the controller keep track of the spindle position and carriage position, and move the carriage independent of the spindle but still keep them in sync for threading. I guess a bit like electronic half nuts. Automatic stopping at the end of the thread would also be pretty slick
If I remember correctly, the LED&KEY Board need some modification to work correctly. A trace need to be cut and re-soldered. I think graphic LCD will requires probably another dedicated microcontroller board as it is memory and CPU intensive. I don't think it is wise to share with the real-time processing tasks. Since the application only requires limited amongst of TPI to be cut, I won't be going for the graphic LCD version.
SPI bus is not at all meant for long distance communications (even 2-3 feet) especially in noisy environments. I would use RS232 instead or (better) some kind of differental signaling bus such as RS485 between the microcontroller and the HMI interface... This is also the reason why RS232 intelligent displays, such as a Nextion + RS232-TTL converter, are a pretty good match for these usecases.
it does seems from the datasheet that the inputs have 0.35V hysteresis, otherwise those rounded edges would be be sensitive to noise going thought the threshold so slowly
Very interested in your project. You have done an excellent job in presenting your project. Well Done. I have a Grizzly G4003 that I am thinking about adding CNC conversion. I would be interested in a kit and/or boards (with holes, eyesight is waning with age). I have ordered a TI board, encoder, key pad, etc. based on what I have seen so far. Was wondering if you would publish a circuit diagram and pin out for the project (I have not been able to find it) and maybe a photo of your breadboard wiring. Also, in the future, it would be helpful to me to have a video showing how to upload the sketch code onto the TI motherboard, although I may be able to figure out that myself since I assume it is similar to Arduino. Also, want to mention, I am planning on putting my Z drive on the tail stock end by machining a section out of the end bearing assembly to allow a timing belt gear to be installed on the shaft. I would then modify the end bearing assembly to fit a roller bearing (for belt tension load)
Another superb episode in your fascinating journey. A great piece of troubleshooting. Everything is perfect. I hate you. Hehehehe - just kidding. I'm looking forward to the next one.
Don't let people convince you that a touchscreen is needed. The best interface if the most simple one that will do the job. Fancy touchscreens sell products but aren't necessary a lot of the time.
I'm cheap. I don't want this to cost any more than it has to.😊
Could agree more, when your hands are often covered in cutting fluid and dirt, who needs something which will have problems with this environment.
A touchscreen does not have any character ;-)
@@oetken007 well not until you program it....
@@DoRC 😂 Absolutely!
I don't see why the fascination with Touch Screens. You have designed a beautifully simple and robust User Interface so stick with it. Remember - KISS!
I agree. Plus he only needs to display numbers. Although a touch screen would be very cool, it is such an overkill for this project.
@@ohammouda ,I agree as well buttons and knobs= more robust
I good touchscreen will have an equal or better life span then a mechanical button. But as a machinist the tactile feeling of buttons. Is very very important it lets you find and work the controls without taking your eyes off the work piece i like the other guys ideas about rotary switches. I also like the idea of a hand encoder for detail work.
Yeah touchscreen is nonsense here.
LED displays are much more readable from a glance and real clicking buttons make for a much more satisfying and reliable interface.
A touchscreen would make that thing only more annoying. You can't use it with greasy fingers, gloves or anything of the sort.
This is actually better than a school lesson because you can actually see what's happening and how to solve the problem or at least a way to solve it. I'm definitely going to watch all the series and I don't know yet but I'll give it a thought about making an ELS for my lathe.
i'm making my way through these videos and i have to say, i am fully team No Touch Screens. so thank you for not taking the "easy" way out.
Excellent progress!
Please consider the following:
Do make kit considerations. I am too old with eyesight gone and memory shot to re-develop the work you are doing so well. I am sure there are few others who, for their own reasons, will join me here.
Once you agree, I will swamp you with several "requirements" 😊
Thanx!
Amen, amen and AMEN!
Make that 3.
DITTO!
SheemOn Google Double ditto on the comment and the kit request!
pat dee b
Here's how good the videos in this series are; I don't have the slimmest clue what you're talking about other than speeds, feeds, TPI and it's for a lathe but, I watched through the WHOLE thing and found it thoroughly enjoyable!
This electronics is pretty much all new to me. But I could watch you all day. Fascinating Teacher! That is a rare gift. Genius for sure.
It´s just perfect, all this tinkering is what a good process is all about. I´m really blown away how good you are able to explain quite complexe stuff. Cheers
All the “”wall wort” power supply give you noise are topically half wave rectified, with no filtering. Battery’s under charging don’t need capacitance. “Because the battery essentially is a big capacitor”
So take that nice scope and look at the out put of the wall charge it’s a lot of bumps all positives but humps. Every rising or falling part of the wave is a count.
Put a 500-100 uF. Across the output of supply, it will get better.
Keep up the good work
J Hicks
Since you have innovated this yourself and most of the commenters are sitting on their hands I would not worry too much about their criticism. You obviously know exactly what you are doing and as you finish it and use it you will figure out what works well and what doesn't. Ignore the comments. Oh wait, then that means you should ignore this too. Ok, so ignore all comments EXCEPT mine cause I'm always rite.
Very nice job! I just bought an old mini lathe, and the leadscrew wheels are missing. Your project is exactly what I need to bring it back to life. Thank you very much for sharing your job.
"Absolutely briliant". You are an awesome genius dear person. Cant wait for each new episode. If you will indulge me with a bit of trivia:
When "transistors" first came into "electronics", I learned and then later taught the theory behind it. This was decades ago of course; when ALL we had was germanium PNPs. Then we all went crazy when silicon NPN's came in, long before chips. And of course, as technology got on a "faster than sound" jet plane and has not landed since; we have gone into a totally new era in mankind. Amazing for sure.
But of all the education and training I got; it was NEVER as good as what you do naturally when you teach. A tip of the hat to you kind sir; for you are truly gifted. Keep it coming.
If a person is in electronics, and really wants to know 'HOW", these are the videos you should watch and listen to. I would give ANY thing just to ask you just a "few" billion questions LOL. I have NO doubt you would KNOW every answer. and teach it in a way that "even" I could understand Oh yes Siree "bobwizer"!
WOW!
Great job with the debugging, real critical thinking 🤔 👌
Whelp...I used to think I was smart.
Amazing work. Thank you for posting these videos. I hope this device bears much fruit for you.
Thank god, you made it metric too!
I think I will actually build 2 of these gadgets.
One for my lathe and one for a gear wheel cutter
I'm very glad there are EE people like you, because I was about to take a very long walk on a short pier. Long story short, software engineer downloads Eagle, makes a ESP32 based irrigation system, order PCB from china, solder stuff on, connects full rect bridge supply and I2C bus spits software dude in the face. There is hope now. I just have to call an EE guy from uni. Thanks :-)
I have a couple of EE guys who answer my questions. :)
I get the feeling you don’t need to listen to the armchair experts - great work, really keen to see your project up and running. I’m sure this will become a commercial option very soon. Good on you for pioneering it.
Wow. Really nice catch. I'd have pulled what's left of my hair out and gone with a specified power supply and the glitchyness. Well done!
There should be min. CS hold time called in the data sheet of the display driver. I would reduce the bit rate because the signals do not get to full HIGH. Great project!
Clough42 =>
Rotary switches for the interface would be nice. Then you have an interface that looks and feels like a gearbox. ;-) Also you can easy switch the buttons with gloves full of dirt.
One for inch / mm; One for thread / plain cut; One for direction; An encoder (with heavy wheel for velocity) 2 select parameters / jogging; The LED display is perfect.
I agree, encoders are nice and easy.
Negative on the encoders. While I agree with your comments that they are nice, it would require even more components to interface with them and I think the indication from most comments I’ve been reading are that he should try to keep this as simple and cheap as possible, and I agree with that as well!
@@EdAgers110 I understand that the second encoder sounds complicated, but it is not. It only must / should work when the lathe runs in plain cut mode (and not in thread mode) or when it stands still (for settings). You can simply switch between the endoder that is connected to the spindle and the encoder on the UI. You can use a ADG3257 from Analog Devices for example. If you are in plain cut mode, the controller switches only one bit on the select line of the ADG3257 and the handwheel ist active instead the encoder on the spindle. The same as the lathe is stopped and a menu button is pressed.
It adds only a few dollars to the bill but adds a lot of useability to the project. It is in my opinion worth it to invest some extra dollars.
The encoder can be really great, but it can also be a liability because it's easy to bump accidentally. Careful design of the user interaction would be required to make it work well. I have encoder menu inputs on my 3D printers, and they're not very precise. I would have to get the feel right to be happy with it.
...It only must / should work when the lathe runs in plain cut mode (and not in thread mode) or when it stands still (for settings)....
I did this just last week! I was tearing my hair out over why my board wasn't working for ages, finally broke down and soldered some fly wires to vias so I could hook up the logic analyzer. Saw the glitch, went "whaaa?" and reread my code. Oops.
As far as your refresh rate goes, it's massive overkill. 15Hz is the "fluid motion" threshold for humans, and even 4Hz is basically unnoticeable for number updates. It could even help, because it lets you _read_ the numbers instead of just having them all blur together. Giving yourself that extra margin at low clock rates will also give you better noise immunity, which might be important with the big lathe motor going and the stepper having to exert a lot of torque. You'd be amazed at how much interference those steppers can generate.
It might be overkill, but one other thing you could put before your input filter is a discrete TVS (transient voltage suppression) diode. This can clamp transients of thousands of volts down to tens of volts, which is much easier for the input filter to eat.
Great project! :) That software bug was a bit surprising, but educational . I would have went down the power supply rabbit hole...and probably lost my mind it the process
You didn't see that part because I didn't turn the cameras on until I figured out what it was. :)
Very cool. Proud of you. I'm also losing weight and taking better care of myself. Seeing your transformation is an inspiration to me. Take care!
Good for you! Anyone who has been through a transformation like this will tell you that it's hard, but it's also easy in a weird way. If you're debating in your head every day whether to have a burger and fries for lunch, it's a horrible grind. But if you come to the place mentally where you have decided that being healthy is more important to you than the enjoyment you used to get from food, something changes and it becomes a lot easier. I don't know how to teach someone to flip that switch, but once you do, it's just a matter of time.
@@Clough42 I totally know what you mean. I think for me it was just time to do it. My clothes fit better, climbing out of the hatch into the garage where the machines (lathe/mills) are is easier, no end of benefits. One day at a time. On a side not I'm looking forward to upgrading my 602 with your kit. I bought the replacement motor years ago, well before your channel started, but never got round to doing the deed. Have all the parts coming in the next few weeks and looking forward to getting that project done too. Thanks for all your hard work.
Very good video, and good production quality too ! top tier content
Thanks!
What's truly impressive is your rate of progress considering this is a hobby.
I tend to focus pretty well. :)
Thank you, an excellent explanation of the fault and a lovely example of de.bug technique
Damn you're a clever chap! Thanks for sharing
Thanks!
Fabulous video - as are all in this playlist. This is one of the best videos an tracking down a bug that I have seen! Everyone so far is a gold mine of well explained analysis. I will continue watching the rest in this series. I do not have a Grizzly but the principles explained here have many applications.
Its always a software bug - ask any hardware engineer ;-) . Loving the level of detail you are going into, have to say its a tempting project.
It's worse than you know. I'm essentially a programmer with a screwdriver.
Really enjoying this series especially thank you!
Really excellent discussion of the electronic debug process. 20 years ago, I has almost the same exact problem bit-banging Dallas Semi 1-wire protocol. I had a very similar Tektronix scope on my bench (except it was almost $5k!), and plowed through it. Great stuff!
Thanks for this entire series. I have a project in mind for your design...
Very cool. PC boards are coming.
Hey just wanted to say that your interface looks really organized and clever
Thanks!
Learning while entertained. Thank you.
...and a peace of red translucent plastic over the seven segment displays will enhance readability even more.
For some reason I always knew these bidirectional converters as "nifty two-way but inherently too slow for anything that is actually fast", so I was surprised to find zero mention of this aspect anywhere from Ebay to Banggood and even Sparkfun, including the linked many-page Philips app note analysing the thing from top to bottom. Not a peep. One has to go all the way to Adafruit to find this immediately addressed in their product description: "the 10K's do make the interface a little more sluggish than using a TXB0108 or 74LVC245 so we suggest checking those out if you need high-speed transfer". So, yeah... nice little gotcha about not wiggling things too fast, apparently not nearly as commonly advertised by those who sell this as it should be.
Great project. Thanks for sharing. The pulse problem has probably to do with the voltage. The phone charger has probably a higher voltage around 5.2 volts. The USB has usually around 4.9 volts. When the chip triggered around half of the power supply voltage the pulse is high enough to trigger at USB and just to low at 5.2 volt.
These USB phone chargers actually fluctuate between 4.5 and 4.75V. I use them to run esp8266 devices that log voltage, and it's really erratic. Won't be buying any more of those.
I have often wondered about the i2c levels and how much the waveform can be distorted and still work. I didn't think it could function looking like that, so now I know that I don't have to be so picky about it. Thanks!
Makes me wonder if the pro touch screen crowd has ever worked a lathe. Oily hands with metal shavings and dust isn't going to fair well with a touch screen. I like your push button box the best so far. Good eye on catching the glitch. Brilliant!
Hi James, thanks for the explanation (it's right on the outside edge of my electronics understanding) and the hard work. Please don't get distracted by comments and come up with an end result that no one can afford. Unless of course you decide that's the direction you want to go. It's your project. I would prefer something that does its job. If I try a deep cut in stainless with a binding leadscrew that's on me. It isn't the job of a hobbyist leadscrew controller to make up for my every bad practice. There are already cnc controllers with servos that do those things and they are too costly for me. Just another opinion by the way. Keep up the great work.
I'm still considering options. On the stepper vs. servo question, the interface is the same, so they can be swapped out, depending on needs.
so far a great build. Need to get some red transparent film, I picked up red green. and blue sheets off ebay , makes a world of difference. I prefer surface mount now even though I cannot see without magnifiers, just use a stencil, place and bake
I'm reaching the same conclusions. A decent reflow oven can be kit-built for $250 or less.
Very informative video, you have put an enormous amount of work into this project and I'm sure it will pay off.
James, there is one thing that you have to keep in mind if you plan to make this commercially available and that is patent infringement. There was a commercially available version of the electronic lead screw on the market in the past but it was withdrawn from the market and the patents were sold to the German company that makes Wabeco lathes. Also there was a Canadian fellow who was developing an electronic lead screw and selling kits. I believe he has a Google groups site and he may or may not have patened his idea.
Fantastic idea for starters. I really hate my lathe banjos and gears. Major concern is of course any asynchronicity between spindle and lead screw due to, not SW or HW, but physical forces strong enough to throw the stepper motor off. In a physical gear connected configuration, lathe motor could halt but would always keep spindle and lead screw in sync.
Most likely, the spindle motor will stop faster than the lead screw, but both are possible. Maybe tachymeter sensors on both can fix this?
Great troubleshooting, coming along great.
Thanks!
The effect of the pullups is less than expected because the MOSFETs are on for a good part of the rising edge and allow the driving end to charge the capacitance quickly initially. It looks like the SPI outputs are tri-state rather than true open drain. If the MOSFET has a low gate threshold it can be conducting all the way to the switching threshold of the receiver and the pullup resistor only has to pull it the rest of the way to 5V.
James Great idea making the lead screw electronic! There is one more thing that the project needs. Use the encoder to give degrees of rotation to be used as an indexer, please!!! You might even consider adding another motor to rotate the head x degrees on key press. Looking forward to your thoughts on this.
Creeping Specifications? Are you with the Gov't? :-)
Odrive is way to go.
Thank you for sharing your project, excellent video.
*I Changed My Mind*
I was one of the ones pleading for touchscreen but I have been thinking about it and honestly, I now think the most important thing is a tactile feel. Bear with me as my lingo may be incorrect, but I am hoping you will make it possible to work with the half nuts always engaged through a threading operation.
On the very small mini-lathes, that's how a lot of us thread (keep half nuts engaged and reverse the lathe before each threading cut). Now, with your controller, I imagine a new way to do it where buttons on the controller stops the leadscrew, so you would hit a "Stop" button when your cutter got to the relief cut. Then back out on the cross slide and hit the "Reverse" button on the controller (which could be set to a faster return speed to move the cutter back towards the tailstock) and then hit "Stop" again. Finally, for the next cut wind the cross slide back in to cutting depth and hit the "Forward" button on the controller and let the box act as the thread dial indicator and engage at the right time. If this sort of programming would be possible, then the touchscreen would not be a good control interface. Tactile buttons would be needed to so you could keep your eyes on the part and not the screen.
Seeing you use the Bud boxes gave me the nudge to get one for my recently-acquired drill press. I want to make a safer and more versatile control for it besides the ancient flaky toggle switch. The investigation into the power supply issue was cool too, with the evidence shown in the scope traces. I didn't realize your typical wall wart was that noisy.
Been waiting for this!
Hi James, nice update for us.
Lance & Patrick.
This project is so great. I really appreciate the careful explanations. You obviously have a sound understanding of the underlying principles. Thank you so much.
Yes, I need one!
Agree about using a touch screen, but personally, I'd like to build a more sophisticated display for my needs. I have a Hardinge turret lathe, and you have most of what I'd like to make into electronic carriage control for manual operation.
Phenomenal work - I'm looking forward to catching up on your project videos, looking at the code (nicely commented btw), and making it work for my lathe.
By the look of the datasheets, you shouldn't need level shifters, just a 3.3V pullup. The TM1638 is low voltage interface friendly by interpreting anything higher than 0.7V as logic high and anything below 0.3V as logic low. Just ensure that there is no pullup on the side of the display/button board, or delete it if present.
Then again with moderate baud rate and decent delays, level shifting is fine. Good one on finding the STB bug. Nice generous STB drive now :D. Conveniently, most of the flank distortion of the signals introduced by capacitance sits well above the 0.7V threshold voltage that i mention above, so the signal is really more legible than it looks at a glance.
For static protection, you can add diodes from signal to both ground and supply of the microcontroller. 50pf or 200pf is not a real static protection measure.
I had wondered if I could get away with driving the display at 3.3v. The hangup is that pullup resistor on the board. I am really trying to make this something that anyone could put together, and requiring a tiny resistor to be removed from the stock board is something I wanted to avoid.
TIme will tell if that was a good tradeoff.
Great Video, James. I am on the outskirts of understanding, as am a retired toolmaker. I am ready to start buying components, and making this project happen though.
Looks like it's moving along well, nice work. One note, the 138 has a much lower working voltage than the 7000's series mosfets. You might want to check the data sheets and gate transfer characteristics, it might make a difference in your level shift.
Do you have another favorite transistor for this?
Great project
Really enjoying this series thank you. I was thinking of fitting something to my lathe when I saw Stefan Gotteswinter (another UA-camr) add a commercial German made ELS to his lathe. However it was many hundreds of Euros so I lost interest but what you are building is probably less. Great work and well explained !
Same here. It was also German-language-only (at least at the time). That's when I lost interest and started thinking about building something of my own.
OK. Good luck with the project, following with interest. Regards UK.
Awesome. Very educational. Would love to have one of these on my lathe, but a DRO is first priority! :-)
Good sleuthing. Since the interface is only being refreshed every 50ms and you don't have that many bits of data to send & receive during that time and it doesn't have anything to do with keeping up with the main loop you can certainly utilize a lot slower baud rate and take advantage of extra timing margins. An LCD touchscreen is an option. The LEDs are nice and bright, my preference.
If you integrate the controller and display in the same enclosure then the only long wires you'll have to contend with are the stepper drive signals.
Perhaps put a resolver on the leadscrew and look for missed steps.
So far, I love the digital concept of controlling the leadscrew. My bet is that lathes will probably soon move in this direction. Singer sewing machines moved into this digital direction a long time ago.
Just keep in mind minimum setup, hold times and levels and you'll do all right.
Keep up the great work.
If I were designing a commercial product from scratch, I would put the microcontroller and the display on a single PCB and put it all in the control box. This particular dev board is a little bigger than I wanted for that kind of construction, but it's an option.
I worked at a TV studio for a few months in the analog days and an engineer explained to me that on long cable runs they would put a pre emphasis/eq on the video signal so when it got to the receiver the loss was compensated for and you got the signal you wanted.
You don't do it on receive as you get noise problems.
I wish I could remember the proper words
Unfortunately on the data line, both ends are sending and receiving data. It would also be highly unwise to anyhow drive data, clock and burst lines in a different manner, because just as long as signal timing is generous and the distortion is equal, it will work fine, but as soon as you introduce differences in distortion, you can get into trouble.
I just stumbled into this series and I have encountered similar problems when trying to convert an elector-mechanically controlled machine to a computer controlled system because electronic switches are so much faster than mechanical switches. Time has to be a consideration. The computer chip sees the switching as instantaneous compared to mechanical contact switching. We also encountered it when trying to add a proximity switch to stop an upstream feed conveyor and alert a machine operator when a component stopped turning. The short interval resulted in many false stoppages as the component speed varied by load. If the it slowed for an instant, it stopped the conveyor. All it took was increasing the time allowed to get the required counts.
I would love to see a motor on the encoder and a tachometer on the stepper to validate the math, possibly add in a method to offer a mechanical load to the stepper to see at what point the full feedback loop would be necessary. Just a thought. I can appreciate the issues of signal integrity as it relates to serial clocking and the edge to state alignment. Nice work!
im enjoying your work. you help me "bootstrap" my understanding of cnc. poppy
Excellent debugging with a scope. Thanks for sharing.
Very cool Sir great work
I'm following your project intently. I hope to god you will make a kit in the end.
Great job using an oscilloscope to check signal propagation.
Suggestion:
You should flash the RPM display to remind people to stop the lathe when switching directions.
Almost as soon as I turned off the cameras, I thought of just decelerating and accelerating the leadscrew smoothly. That would save a step and make it much easier to use in the case of "oops, it's going the wrong way".
This is a really good project
Loving how this is developing!
Would absolutely buy this as a kit when ready and available ❤
I think the simple display and buttons is fine - touchscreens can be nice, but I'm not convinced it's necesary or advantageous.
Great video, thanks!
*Touch Screen Thoughts*
I know you mentioned in a previous video, that a touch screen would be a bad idea in a shop environment. I still think that it could possibly work. It's supposedly not that uncommon on high end CNC controllers (I think Mazak has one with touch screen). I'll admit that my iPhone is horrible if my fingers are just a tiny bit moist or greasy, but I have had that exact same phone in a super cheap zip lock style bag for diving trips and even with wet hands and gloves on, for some reason I have been able to jot down GPS points when I got back on the boat. Just saying, there must be some touch screens they work well even in an workshop environment. Maybe even some cheap solutions.
Either way I, can't wait to get my hands on your board and make this upgrade:).
Great to hear you will test servos also. Since you are making the rest of the controller so "sturdy", I think a closed loop for the leadscrew makes a lot of sense.
Actually, apologies up front for the crazy curve ball here, but is there any way this could be made to work as an APP on, say, Android...? I did put scales on my little lathe and used Yuri's TouchDRO firmware on a cheap TI Launchpad. How amazing it would be if there was a way to merge his and your features into one APP. I know you might have plenty of very good reasons for not doing that, but no harm in trying to respectfully plant a seed.
The smartphone app idea is definitely possible, but probably out of scope for the work I'm doing. Priority one is getting the core controller and a simple UI working. It could eventually be configurable to use different display hardware, and a bluetooth dongle with an app for the display seems like a reasonable option.
I like that black plastic better. You should be able to print your lettering into the front and even do so in another color. You just need a simple 5-12v transformer that isn’t an actual charger.
Thanks again. Enjoying all you are doing.
Great content!
Great job spotting the issue!
Acme threads? Amazing work though, thank you!
Outstanding work and presentation!
Nice progress! One of my interns has been using 3d printed molds to make silicone buttons using hardware store silicone. They feel great and are individually collapsible buttons in a sheet format. I can send you OpenSCAD files to play with if you'd like.
please post a link.
drop the pullups by a factor of 2. you are well within the operating region of the fet. look at the data sheet.
The issue is the sink current of the microcontroller output, which is 3 or 4 mA. There are already three 10K pullups in parallel (one on the 3V3 side, one on the 5V side and one on the display board. So we're already at half that. I may play with it a bit.
@@Clough42 Since these lines may be uni-directional why not use opto-isolators to achieve the level shift. You prevent the micro-board from being fried too.
Awesome!
For the enclosure put a thin membrane (rubber/silicone?) between the buttons and switches. It will stop oil and dirt from entering the PCB... Maybe sandwich it between the enclosure and some support structure.
I believe a conformal coating (which is a specially formulated acrylic for PCBs) wouldn't go amiss either, if oil or some detritus ever did make it's way into the box, the coating will prevent a short circuit from occurring.
Really good clear explanations.
The cure for capacitance on I2C and SPI signal and clock lines is to make certain all your inputs are configured in Schmitt trigger mode. Or, if you're bit-banging the input (or using an FPGA that doesn't support Schmitt trigger mode), use a software glitch filter (for example, sample the input at 16x the clock and ignore transitions that aren't steady for at least 8 samples).
If you think about it, an I2C or SPI signal line with a lot of capacitance (either distributed or bulk), which goes off-board (or all over a board to different power domains) isn't really a digital signal line. It's an analog signal which might be very noisy or have transition times which are too slow for reliable detection with non-Schmitt inputs.
This is good advice. The TI DSP chip I'm using has a sampling qualification mode, but I think in this case, the SPI peripheral is actually conditioning the input.
Great video Thanks
This series has been fantastic ... I'd love to see how you document and keep track of everything. Word doc? Knowledge base? Legal pad? Thanks!
The signals are OK for development version but I would suggest a driver and receiver chip for the signals because eventually you will get caught out. That has been my experience at least.
You're not wrong. In this case, since it's a bidirectional signal, I would need a driver at each end. If I'm going to spin a PCB for the control panel end, it's worth thinking about just doing a whole control panel board.
Amazing. Thank you.
Great video, well explained - I learned something too! 👍
James, you explain things so practically, did you used to teach?
Negative. I'll take that as a compliment, though. :)
The lost signal by Volatage drop is a really old problem. The Voltage breaks down when the thyristor/transitor is switched. A good power supply can handle that. But cheap charger don't compensate, small surges of power can make the voltage drop in the hole system.
Two ways to counteract that.
1. Use over-voltage 5.5V by example.
2. put big caps ( 5v to ground ) near paces where surges can happen. By example near the thyristors.
I'm designing/building circuits for over 20 years now and had such problems often in the past. Modern power supplies have much better voltage regulation so the problem is not so prevalent anymore. The first time i had such a problem it took me weeks to figure it out.
Yeah, bulk bypassing often helps. I did add the recommended bypass caps to the display board, but the timing was the underlying issue in this case.
Hi, that was a very good and well presented masterclass in debugging. I'd like to suggest, if only for my own benefit (!), that at the end you consider either making the final (is it ever?) compiled code available, or link to or demo how to go about doing a compilation. At the moment that part looks like a 'black art' to me! I've looked at the reference for the source code - which I can sort of read - but would have no idea how to put it all together. I do have a handy son who understands these things though, but not everyone will have such a handy child or knowledgeable friend. I'd like to confirm that should you choose to make a through hole pcb board or a full kit (less the heavier shipping items like the motors, but perhaps with an assembled smd board) available at anything resembling a reasonable price - including recovering some of your time and development costs - I would order one. This is perhaps the most useful add on to any lathe - even one with a change gear box - that is available. The ability to instantly switch between US, imperial and Metric thread standards is invaluable to much of the model making world outside the States for reasons that are a bit too long to go into here. If you do go down that route please endeavour to make sure that your selected supplier or fullfillment centre can both cope with international orders and do so without silly shipping costs - some sadly price themselves out of the market. BobUK.
Hello, project looks great! I do not know, if you mean real servo or a closed loop stepper like this: JMC 2-Phase iHSS86-80-70. I would be more then happy to build one and test along with you on my lathe (TSB-20 polish small lathe). Also I have an idea for additional feature: add another encoder input to to replace carriage hand wheel. Button or switch to disable spindle encoder input, this way you could replace lead screw with ball screw i.e. SFU1605 or even SFU2005. You could wire alarm output from motor drive to detect motor stall due to collision or use serial port in JMC drive.
how are you going to manage threading passes? it looks as if you would be able to just reverse the spindle to reverse and let the carriage come back to the start of the thread, but it would be pretty cool to just hit a return button, or the reverse button, and have the controller keep track of the spindle position and carriage position, and move the carriage independent of the spindle but still keep them in sync for threading. I guess a bit like electronic half nuts. Automatic stopping at the end of the thread would also be pretty slick
If I remember correctly, the LED&KEY Board need some modification to work correctly. A trace need to be cut and re-soldered.
I think graphic LCD will requires probably another dedicated microcontroller board as it is memory and CPU intensive. I don't think it is wise to share with the real-time processing tasks.
Since the application only requires limited amongst of TPI to be cut, I won't be going for the graphic LCD version.
SPI bus is not at all meant for long distance communications (even 2-3 feet) especially in noisy environments. I would use RS232 instead or (better) some kind of differental signaling bus such as RS485 between the microcontroller and the HMI interface... This is also the reason why RS232 intelligent displays, such as a Nextion + RS232-TTL converter, are a pretty good match for these usecases.
Felicitaciones Veo que es todo un profesional en infinidad de tareas Yo aqui me encuentro como cucaracha en vaile de gallinas.
it does seems from the datasheet that the inputs have 0.35V hysteresis, otherwise those rounded edges would be be sensitive to noise going thought the threshold so slowly
Very interested in your project. You have done an excellent job in presenting your project. Well Done. I have a Grizzly G4003 that I am thinking about adding CNC conversion. I would be interested in a kit and/or boards (with holes, eyesight is waning with age). I have ordered a TI board, encoder, key pad, etc. based on what I have seen so far. Was wondering if you would publish a circuit diagram and pin out for the project (I have not been able to find it) and maybe a photo of your breadboard wiring. Also, in the future, it would be helpful to me to have a video showing how to upload the sketch code onto the TI motherboard, although I may be able to figure out that myself since I assume it is similar to Arduino. Also, want to mention, I am planning on putting my Z drive on the tail stock end by machining a section out of the end bearing assembly to allow a timing belt gear to be installed on the shaft. I would then modify the end bearing assembly to fit a roller bearing (for belt tension load)
Did not disappoint!
Another superb episode in your fascinating journey. A great piece of troubleshooting. Everything is perfect. I hate you. Hehehehe - just kidding. I'm looking forward to the next one.