I'm concerned about the e-stop being a NC to ground. If your conductor shorts to ground at any point, the control won't react. That's usually why you have it as an NC passing HIGH, so any disconnect, being via the switch or a break in the conductor, will force an e-stop condition.
100% correct, also a saftey relay monitoring the zones should close normally open auxiliary contacts for releasing the enable signal for the inverters. But at the end of the day it all comes down to how advanced does he need to make a custom machine, that doesn't have to comply with international saftey system standards since it's presumably none commercial, if he ends up selling this as a kit to customers it would be a different story of course.
Yeah, in this configuration it's more of a "feed hold" that is dependent on the normal operating system to work correctly. He must have a reason, though - he's very sharp.
While not ideal, there definitely needs to be a physical pull up resistor at the controller side. Absolutely do not rely on the software controlled internal pull up
You made the right choice with your purchase of the Knipex ferrule crimpers. I've had the same ones for a number of years now and I have to say I have never been disappointed with them at all. Having said that, all my pliers and grips, etc, seem to be Knipex now.
23:40 for the 10V max input you could consider a 10 or 12v TVS diode between pin 1 and 4, in case of a resistor failing or solder crack it limits to a safe input voltage to the controller
I once worked on a huge old CNC lathe with a separate control cabinet connected to the machine by a 6" wire duct. Instead of taking the wires out of the terminal strip, the riggers just cut the duct in half with a sawzall. I would guess there were 250+ wires inside, all blue, only labeled at each end. It took two electricians about a week to sort that out. Nowadays that duct would have a few powers and grounds and one Modbus cable.
My father spent decades as lead mechanic/electrician in a tool & die shop, and eventually went into business for himself retrofitting/upgrading CNC controls. I eventually lost count of the number of times he'd come home with that same story -- hundreds of identical wires that had to be rung out and reconnected because someone bought a new-to-them machine (or had to move an old one) and didn't know any better. Even better when the wires weren't labeled at all anymore. I seem to recall you working on a backhoe or some such with a similar problem (in a video on your channel) -- a thick harness with identical wires, and you sorting it out. Must've given you some PTSD. 😱
@@mikepetersen2927 It was a JCB Telehandler, with all the wires being yellow if I remember correctly. I just watched that one like a week or so ago. Funnily enough, I came to this video right after finishing Watch Wes Work's latest video
It's the same with live sound. We used to have raceways with a hundred shielded balanced strands and racks of patch panels and processing gear. It's now just a single CAT5 cable and everything is in software. Technically this HMI has GPIO with hardware quadrature decoders, and I could run everything over the single serial cable, but that would require a custom PCA to break out the flat flex and translate to 3.3V.
The ceramic SMD resistors have no strain relief mounted like that, thermal changes and shock/vibration of the switch contacts could crack the ceramic or peel the metallisation off the substrate. Opinion: through-hole resistors wired to the switch pins might be a safer choice in the final implementation.
As a general rule when one implements resistive multiplex switches one avoids putting valid states at VCC or GND so that your code can detect short/open faults. This brings up another point, if your GND opens the whole matrix is going to pullup to 24v. Is your analog input ok with that?
I suggest to change the resistor devider: Pos1 = 2,5 Volt Pos2 = 5 Volt Pos3 = 7,5 Volt Pos4= 10 Volt Because if analog wire gets disconnected, it will most likely be detected as zero Volt by the microcontroller. If you do it like i suggest, then you can detect open wire fault
@Clough42 10 Volt analog inputs are most likely fairly low impedans. Possible 100K ohm, it will then detect zero Volt if floating. Not many ADC can handle above 5 Volt, thats why i expect the input having some kind of a resistor devider. Also 10 Volt analog signal is an industrial standard, where long wires often are connected. And here a low impedans input is better to suppress noise on the wires. Im sure your design will also work, at least until the analog signal wire or the rotary switch goes bad. In that case I still think there is no dangerous situation. The ZXY will detect OFF and the other rotary switch with feedrate will move faster (if it detects a faulty zero volt). BTW great videos you make, I watch all off them and enjoy every second👍
Been avoiding cable management generally for quite some time now. But things have come home to roost recently since I've been using micro-controllers and breadboards. I now have a ghastly mess on my hands, and its fun and games trying to figure out where to reposition what. This was an interesting watch and I'm grateful for the tips you gave. Anyway, man, the wiring on your control panel is now a thing to behold.
Nice design, shows your attitude to not only have it functioning but also have a clean design. Just add a 10V zener diode over the 10V pin, if one of the resistors break, you get 24V to the analog input.
Great video! One small way to make the rotary switches a little bit safer, is to forbid the values 10V and 0V from being valid, so that you can detect if the 24V or the ground are broken. The current has to flow in the ladder to get non-extremal values. The 24V =10V statement impliedin my previous remark might seem a bit weird at face value, but the analog input is probably protected with a diode that would peg the max voltage at the pin at 10V.
I don't know why you don't have more subs but I'm sure the ones you do have are devoted to your channel as I am. I'll never have your skills and ability but it's such a pleasure being able to enjoy these things vicariously as if I did!
Also, consider some wire loom covering. Can be purchased split for retrofit or solid for custom harnesses. There's also wore loom tape. Either would pretty up the umbilical from the connector panel to the boards & terminal strips
One thing I would suggest is routing the run command through contacts on the e-stop switch so that the e-stop must be released to pass the run command to the controller.
@clough42 Excellent job! I have been an engineer for the better part of a half century. I find you to be a cut above in the fact that you do not stay in your own swim lane. (software I believe). I have been having discussion with some of the Senior level engineers and the consensus is that with the younger generation they stay in their own lane never venturing out, but then they complain that guys like us make so much more. I over heard the exact conversation one time and I explained it was because we would have to hire 4 of you to do the same job as Joe over there who has been there for 40+ years. When you can do what he can do you will make what he makes. The looks I received were rictus. Keep up the awesome job! Thanks again. BTW technically there is no such thing as a DB-9. The 9 pin connector has an E size shell on it but because the DB-25 which has a B size shell was out first in the PC market the term caught on and the actual terminology DE-9 has been lost into obscurity. I have the actual data sheet to back this up someplace but obviously it can't be posted here.
I’m building a small CNC milling machine. My controller is very similar to yours. I’m going to use your wiring tricks for mine! Thanks for the instructions.
Looks amazing - I love it. Having one of the surface mount resistors upside down compared to the other two did wind me up a little bit, but I suspect it annoyed you more once you saw!
I recently sat drooling over the Amphenol Aerospace connector catalog, with their MilSpec connectors, and immediately thought about this project of yours. They have connectors up to 20 pins in there, with serious IP ratings. But, its probably overkill for you (and very expensive). Also came across a RUclipsr who made an ingenious cable wrapping gismo, with a large C-shaped wheel driven by a stepper motor and belt. Quite simple, but very effective. One could make seriously good harnesses with the 2 above. Probably not for prototyping though. ATB!
@@roflchopter11 Its a video from 'Well done tips" channel. Video title: "DIY Wire Harness Wrapping Machine". But there are many similar videos with some good designs on YT.
Definitely worth the time imo. I know connectors are expensive, and can cause issues overtime in humid environments. But if its a prototype, or a beta.... I am putting connectors on it. I don't want to have to desolder stuff, or go fishing in a loom if i can avoid it. These days I also go to the trouble of printing labels for my wiring. I find it makes my life that much easier, when stuff breaks down the road. Should I have to replace switches/components, etc. Sometimes I even go to the trouble of embossing the wiring connecting list and pin numbers, on the inside of project lids.
Yes, I hate DB-9 connectors as well. In the ‘80s, I had to build a bunch of cables for Texas Instruments minicomputer terminals with a custom wire consisting of 7 conductors plus coax. We used crimp DB-9 connectors, so it wasn’t the worst thing ever, but it was close.
Back in the day my friend called the thumbscrews for d-sub connectors "goddamnits" due to how fiddly they are. "You forgot to do up the goddamnits on this cable." "We need to order another bag of goddamnits for these breakout connectors." etc
Just a tip for wire maintenance, using fishing wire or dental floss wrapped around and down the loom cleans it up without the use of zipties. Automotive motorsports wiring harnesses use this technique and it looks super clean
I'm gonna be that guy from Napoleon Dynamite that answers Napoleon's question "do the chickens have large talons" and he says "son, I don't understand a word you just said" 😂 Regardless James, I find your content relaxing, enjoyable, entertaining, and perhaps someday, useful to me because it truly interests me. Cant wait for to have to do nearly any work to surface grind something (knowing you have done a $h!t-ton of work to get to that stage! Stay AWESOMENESS!
By this stage, James has printed so many versions of the case... how long until he just makes two units, one with each type of base, complete with duplicate electronics, and then he never has to swap back and forth between grinder and software development? (Also, it's good to have a duplicate for troubleshooting.) Of course -- thumbs up as usual for recounting all the detailed thinking!
Great job. Personally I would have gone with thro-hole resistors (mechanically stronger and more flexible than 0805) with a 14K + 3k3+ 3k3+ 3k3 chain. Also, a 10uF capacitor at the wiper to ground as a low impedance path for any noise current. Also 14.0K is in the E96 series.
Nice trick with the analog voltage steps as the switch positions. I would avoid using gnd as one of the positions in that case. Instead add another resistor. That way you can distinguish between a valid position of the switch and it being in a transitional state between positions.
I think the main and the biggest drawback of having a 3d printer is you would never be 100% satisfied and get lost in the cycle of improving and printing, again and again
A few comments: You might want to add a small series resistor along with a cap to ground for noise suppression (debounce) and (open circuit) protection when the switch is rotated. It most likely is a break before make switch thus the input voltage to the controller will be open circuit when the switch is rotated. Additionally, depending upon the controller the 10V selection could possibly exceed the maximum input voltage at the ADC. Ideally you would make the max input voltage be
Nicely done. I would have estimated/calculated the current that the circuit needs and set the current limit of my PSU at 20% above that before "first power up". Less chance of magic smoke that way.
beautiful work james... the wiring, the enclosure... the care you put into your work. all beautiful... except for that god-awful label on top of your jog dial :) thanks for filming and sharing.
Nice and informative video. But why not wire the plugs the other way around? Now you have to plug the connectors in the appropriate enclosure bottom-box and then plug the computer or machine to the box. You could chose to have the connectors fixed in the bottombox and already wired to the machine or the computer and plug in the connectors when you mount the top... or am i missing something? 21:06
if the ferrils sometimes fall of then its the force the crimpers put on them that is to low, mine had a screw i could loosen to turn a little wheel making them krim much harder onto a wire ( that could also dammage wires if to heavy )
20:51 I would suggest replacing the RJ-45 patch cable with a slightly longer one, so it's about the same length as the other wires. Other than that, beautiful job!
Nice. I'm used to doing this stuff for a living, in the LA area (HAC MTS3). I have always liked this stuff but there is no more work for it. Besides, at 70, I am long retired anyway. ;) Nice project. The software will be the fun stuff ;)
Interesting the lights on the board were dimmer at different voltages. I would have hoped the LEDs where driven by a high impedance (transitor, mosfet, etc), but given they are dimming as soon as the voltage starts to drop, does that mean the LEDs are trying to pull typical LED current (10-20ma) from the source feed rather than the board power feed?
The SMD resistors are very brittle and will with time break if the switch terminals move at all. Speaking from experience. 1/8W axials are small and reliable for this purpose.
Just so you know it’s bad practice to twist the conductor strands before insertion into the terminal [ in your case bootlace ferrule] according to IPC/WHMA-A-620D.
Was thinking the same thing, but if you look closer it seems that he's not actually twisting them, but rather "tidying up" some stray strand with a very light twisty-turny brushing motion.
When I use ferrules on small wires I like to remove more insulation off of the wire so I can double over the wire before putting it in the ferrules . It makes a strong connections and I can use a slightly larger ferrule.
Careful with surface mount resistors they very susceptible to cracking due to mechanical stress I'd be quite nervous having them floating in mid air between switch terminals.
Maybe you've talked about it already in another video but what digital notebook are you using? Do you like it? Would you recommend it or another one? I've been looking into making a similar purchase. Thanks
You are 10x smarter than me, so hard to have a comment but I would suggest always using female connectors on a bulkhead panel and use male-to-male cables because if you break a pin, repair or replacement is way easier.
Am I only one finding ABS-GF parts to cause itchy feeling on my skin? If not how to deal with that? Aslo for projects like that, maybe modular plugs would be good idea? ODU is expensive but having 3D printer parts except pins and inserts chould be done inhouse. Cheers
I’m concerned that if your resistor chain is varying the intensity of the on board LED, the current in the chain may be wrong and hence the voltages will not be what you expect. Measure and see!
My $15 Amazon ferrule crimper works amazingly well, even down to very thin wires. Never had the ferrules come loose. If you can afford to buy Knipex, sure, but definitely not necessary.
make sure to check if its tied into a cloud subscription. allot of them sell you the digital 'pad' and then you need to pay for a cloud subscription or the pad no longer functions. pretty shitty.
Why did you not just connect the piece with the connectors to the newly printed housing instead of the base. If the connectors are part of the upper part then it is way easier to move around.
This is some really impressive work, James. The hardest part of digital control for me anyway, is the interconnection between the software and the machine. This really helps me wrap my head around the entire process.
kind of pushing the wattage on those 0805 resistors in total, since they are split up its not that big of a deal but I would keep an eye out on the 1.4k combination (yea I know you mentioned it but run of the mill no name 805's are only 0.1 watt and your at 0.1 watt with the math)
Looks nice James. I will be patiently awaiting you next installment on this series. I like the use of the SMD resistors on the switch, it makes for a clean simple installation. Some thoughts for consideration. I have used spiral wire wrap to help protect and clean up an installation of multiple wires on some control panels I have made before. Personally, I like that stuff more than just using wire ties (zip ties). Wire ties alone can pinch wires and possibly cause damage, especially if they are cinched too tight. Admittedly, the spiral wrap can be a pain in the rear to use when the wire harness gets really long, but the end product does look clean and tidy.
Not that any of the beautiful work you did is invalid, but I’m my world we just make two (or more) protos, one for software dev and the other for hardware. Swapping between platforms takes too much time. Of course, that does require an easy way to update the firmware in the hardware proto, but that is commonplace these days.
21:23 Piano hinge, with a loop bent into the end of the hinge wire, is a convenient means of attaching similar panels. I don't mean that you should change what you have now, it's an idea to tuck in a grey cell for something down the road.
I often wonder why industrial applications like this don't fully "modernize" connections to USB-C. If you are going to be programming anyway, surely a single cable to carry power and all the data you could possibly use would be an ideal solution?
great video! maybe put a tvs or zener diode across the switch resistors in case the power supply puts out a spike in voltage. might save your clear core, but it also might already have protection built in ^^
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
I'm concerned about the e-stop being a NC to ground. If your conductor shorts to ground at any point, the control won't react. That's usually why you have it as an NC passing HIGH, so any disconnect, being via the switch or a break in the conductor, will force an e-stop condition.
100% correct, also a saftey relay monitoring the zones should close normally open auxiliary contacts for releasing the enable signal for the inverters.
But at the end of the day it all comes down to how advanced does he need to make a custom machine, that doesn't have to comply with international saftey system standards since it's presumably none commercial, if he ends up selling this as a kit to customers it would be a different story of course.
Yeah, in this configuration it's more of a "feed hold" that is dependent on the normal operating system to work correctly.
He must have a reason, though - he's very sharp.
He explains why around 18 minutes in.
Keep in mind he is using a microcontroller to interpret these signals.
While not ideal, there definitely needs to be a physical pull up resistor at the controller side. Absolutely do not rely on the software controlled internal pull up
You made the right choice with your purchase of the Knipex ferrule crimpers. I've had the same ones for a number of years now and I have to say I have never been disappointed with them at all. Having said that, all my pliers and grips, etc, seem to be Knipex now.
It's an infection that usually starts with a pliers wrench set. Looking in my toolbox, I see the same thing.
@@Clough42 Ha haa. And don't our wallets know it.
@@billdoodson4232 It's likely both of you got what you paid for. A "Caddy" never legitimately comes at a "Chevy" price.
23:40 for the 10V max input you could consider a 10 or 12v TVS diode between pin 1 and 4, in case of a resistor failing or solder crack it limits to a safe input voltage to the controller
Very nicely done James!
ATB, Robin
I once worked on a huge old CNC lathe with a separate control cabinet connected to the machine by a 6" wire duct. Instead of taking the wires out of the terminal strip, the riggers just cut the duct in half with a sawzall. I would guess there were 250+ wires inside, all blue, only labeled at each end. It took two electricians about a week to sort that out. Nowadays that duct would have a few powers and grounds and one Modbus cable.
My father spent decades as lead mechanic/electrician in a tool & die shop, and eventually went into business for himself retrofitting/upgrading CNC controls. I eventually lost count of the number of times he'd come home with that same story -- hundreds of identical wires that had to be rung out and reconnected because someone bought a new-to-them machine (or had to move an old one) and didn't know any better. Even better when the wires weren't labeled at all anymore.
I seem to recall you working on a backhoe or some such with a similar problem (in a video on your channel) -- a thick harness with identical wires, and you sorting it out. Must've given you some PTSD. 😱
@@mikepetersen2927 It was a JCB Telehandler, with all the wires being yellow if I remember correctly. I just watched that one like a week or so ago. Funnily enough, I came to this video right after finishing Watch Wes Work's latest video
It's the same with live sound. We used to have raceways with a hundred shielded balanced strands and racks of patch panels and processing gear. It's now just a single CAT5 cable and everything is in software. Technically this HMI has GPIO with hardware quadrature decoders, and I could run everything over the single serial cable, but that would require a custom PCA to break out the flat flex and translate to 3.3V.
@@Beef4Dinner22 Winnebago RVs are done the same way: dozens of yellow wires all with printed id numbers on the insulation.
Seeing your nicely routed and cable-tied down wiring inside the case was a lovely sight to see.
The ceramic SMD resistors have no strain relief mounted like that, thermal changes and shock/vibration of the switch contacts could crack the ceramic or peel the metallisation off the substrate. Opinion: through-hole resistors wired to the switch pins might be a safer choice in the final implementation.
I second that
Several interesting and worthwhile comments, but James, you always delight me with your videos!
Thank you.
As a general rule when one implements resistive multiplex switches one avoids putting valid states at VCC or GND so that your code can detect short/open faults. This brings up another point, if your GND opens the whole matrix is going to pullup to 24v. Is your analog input ok with that?
Totally agree. Cracked SMD resistors failing open are at the risky end for me and 24v up an input potentially isn't good.
I suggest to change the resistor devider:
Pos1 = 2,5 Volt
Pos2 = 5 Volt
Pos3 = 7,5 Volt
Pos4= 10 Volt
Because if analog wire gets disconnected, it will most likely be detected as zero Volt by the microcontroller.
If you do it like i suggest, then you can detect open wire fault
I've been thinking the same thing. It would be easy to do by swapping the resistors. Though I think it might go high when it's open.
@Clough42 10 Volt analog inputs are most likely fairly low impedans. Possible 100K ohm, it will then detect zero Volt if floating.
Not many ADC can handle above 5 Volt, thats why i expect the input having some kind of a resistor devider.
Also 10 Volt analog signal is an industrial standard, where long wires often are connected. And here a low impedans input is better to suppress noise on the wires.
Im sure your design will also work, at least until the analog signal wire or the rotary switch goes bad. In that case I still think there is no dangerous situation. The ZXY will detect OFF and the other rotary switch with feedrate will move faster (if it detects a faulty zero volt).
BTW great videos you make, I watch all off them and enjoy every second👍
Been avoiding cable management generally for quite some time now. But things have come home to roost recently since I've been using micro-controllers and breadboards. I now have a ghastly mess on my hands, and its fun and games trying to figure out where to reposition what. This was an interesting watch and I'm grateful for the tips you gave. Anyway, man, the wiring on your control panel is now a thing to behold.
Nice design, shows your attitude to not only have it functioning but also have a clean design. Just add a 10V zener diode over the 10V pin, if one of the resistors break, you get 24V to the analog input.
Great video! One small way to make the rotary switches a little bit safer, is to forbid the values 10V and 0V from being valid, so that you can detect if the 24V or the ground are broken. The current has to flow in the ladder to get non-extremal values. The 24V =10V statement impliedin my previous remark might seem a bit weird at face value, but the analog input is probably protected with a diode that would peg the max voltage at the pin at 10V.
I don't know why you don't have more subs but I'm sure the ones you do have are devoted to your channel as I am. I'll never have your skills and ability but it's such a pleasure being able to enjoy these things vicariously as if I did!
"I'll deal with it later." - SURPRISE! It's later! 😀
Also, consider some wire loom covering. Can be purchased split for retrofit or solid for custom harnesses. There's also wore loom tape. Either would pretty up the umbilical from the connector panel to the boards & terminal strips
I like that you design with the future in mind. I tried to do that in all systems I designed.
One thing I would suggest is routing the run command through contacts on the e-stop switch so that the e-stop must be released to pass the run command to the controller.
@clough42 Excellent job! I have been an engineer for the better part of a half century. I find you to be a cut above in the fact that you do not stay in your own swim lane. (software I believe). I have been having discussion with some of the Senior level engineers and the consensus is that with the younger generation they stay in their own lane never venturing out, but then they complain that guys like us make so much more. I over heard the exact conversation one time and I explained it was because we would have to hire 4 of you to do the same job as Joe over there who has been there for 40+ years. When you can do what he can do you will make what he makes. The looks I received were rictus. Keep up the awesome job! Thanks again. BTW technically there is no such thing as a DB-9. The 9 pin connector has an E size shell on it but because the DB-25 which has a B size shell was out first in the PC market the term caught on and the actual terminology DE-9 has been lost into obscurity. I have the actual data sheet to back this up someplace but obviously it can't be posted here.
I’v just passed the half century!
So excited for the rest of this series! Appreciate the time you put in to these!
Very impressive James, really nice design and attention to detail with the wiring.
I’m building a small CNC milling machine. My controller is very similar to yours. I’m going to use your wiring tricks for mine! Thanks for the instructions.
*_Another excellent video that went right over my head.. Again.._* 😆🤣
It's absolutely amazing to follow this project :)
Terrific work there James. Well planned and executed Thanks for sharing.
Great job James, looking forward to seeing the programming video.
Looks amazing - I love it. Having one of the surface mount resistors upside down compared to the other two did wind me up a little bit, but I suspect it annoyed you more once you saw!
I recently sat drooling over the Amphenol Aerospace connector catalog, with their MilSpec connectors, and immediately thought about this project of yours. They have connectors up to 20 pins in there, with serious IP ratings. But, its probably overkill for you (and very expensive). Also came across a RUclipsr who made an ingenious cable wrapping gismo, with a large C-shaped wheel driven by a stepper motor and belt. Quite simple, but very effective. One could make seriously good harnesses with the 2 above. Probably not for prototyping though. ATB!
What's the video title? Link probably won't work, sadly.
@@roflchopter11 Its a video from 'Well done tips" channel. Video title: "DIY Wire Harness Wrapping Machine". But there are many similar videos with some good designs on YT.
I love DB-9s , milling out spots for them on broadcast back panels kept me employed for a while.
I once milled openings for a DB9 and an IEC inlet receptacle in a die cast traffic light housing.
Great job, James.
Definitely worth the time imo. I know connectors are expensive, and can cause issues overtime in humid environments. But if its a prototype, or a beta.... I am putting connectors on it. I don't want to have to desolder stuff, or go fishing in a loom if i can avoid it. These days I also go to the trouble of printing labels for my wiring. I find it makes my life that much easier, when stuff breaks down the road. Should I have to replace switches/components, etc. Sometimes I even go to the trouble of embossing the wiring connecting list and pin numbers, on the inside of project lids.
Hi James nice system design it's also neater to put the cable in nylon braid with heat shrink on the ends thanks for the video.
Yes, I hate DB-9 connectors as well. In the ‘80s, I had to build a bunch of cables for Texas Instruments minicomputer terminals with a custom wire consisting of 7 conductors plus coax. We used crimp DB-9 connectors, so it wasn’t the worst thing ever, but it was close.
Place I worked had me making custom 26 pin cables with Amphenol connectors on both ends. All connections were soldered. That was... an experience.
Back in the day my friend called the thumbscrews for d-sub connectors "goddamnits" due to how fiddly they are.
"You forgot to do up the goddamnits on this cable."
"We need to order another bag of goddamnits for these breakout connectors."
etc
I was somewhat shocked when I looked up the Knipex prices. But, a good tool that keeps on working is probably worth it. Love your videos.
I was also shocked. Butt it has performed admirably and I haven't had a failed crimp since.
Just a tip for wire maintenance, using fishing wire or dental floss wrapped around and down the loom cleans it up without the use of zipties. Automotive motorsports wiring harnesses use this technique and it looks super clean
I'm gonna be that guy from Napoleon Dynamite that answers Napoleon's question "do the chickens have large talons" and he says "son, I don't understand a word you just said" 😂
Regardless James, I find your content relaxing, enjoyable, entertaining, and perhaps someday, useful to me because it truly interests me.
Cant wait for to have to do nearly any work to surface grind something (knowing you have done a $h!t-ton of work to get to that stage!
Stay AWESOMENESS!
By this stage, James has printed so many versions of the case... how long until he just makes two units, one with each type of base, complete with duplicate electronics, and then he never has to swap back and forth between grinder and software development? (Also, it's good to have a duplicate for troubleshooting.) Of course -- thumbs up as usual for recounting all the detailed thinking!
That's an excellent idea.
Great job. Personally I would have gone with thro-hole resistors (mechanically stronger and more flexible than 0805) with a 14K + 3k3+ 3k3+ 3k3 chain. Also, a 10uF capacitor at the wiper to ground as a low impedance path for any noise current. Also 14.0K is in the E96 series.
Nice trick with the analog voltage steps as the switch positions. I would avoid using gnd as one of the positions in that case. Instead add another resistor. That way you can distinguish between a valid position of the switch and it being in a transitional state between positions.
“good job James”……yup!!!! good job James❤
I think the main and the biggest drawback of having a 3d printer is you would never be 100% satisfied and get lost in the cycle of improving and printing, again and again
A few comments: You might want to add a small series resistor along with a cap to ground for noise suppression (debounce) and (open circuit) protection when the switch is rotated. It most likely is a break before make switch thus the input voltage to the controller will be open circuit when the switch is rotated. Additionally, depending upon the controller the 10V selection could possibly exceed the maximum input voltage at the ADC. Ideally you would make the max input voltage be
Nicely done. I would have estimated/calculated the current that the circuit needs and set the current limit of my PSU at 20% above that before "first power up". Less chance of magic smoke that way.
Super satisfying
Rather than using chocolate blocks (screw terminals), it's fairly trivial to create mounts for Wago inline connectors.
Thanks for Sharing ... Stay Safe and Well ...
Looking beautifully over designed…love it
The quality of those prints and the etched(?) surface is insane, I constantly thought it was going to segue to a JLC ad for sls printing haha
Sooooo satisfying! 🤘
beautiful work james... the wiring, the enclosure... the care you put into your work. all beautiful... except for that god-awful label on top of your jog dial :) thanks for filming and sharing.
Just awesome!
Great job!!
Thanks for sharing 👍
Ferrules are like chamfers for wire. This is the way
Beautiful work. I don't like the screw connector blocks. I would want to look at a Wago-based set of connectors. Might be a bit big though.
Nice and informative video.
But why not wire the plugs the other way around?
Now you have to plug the connectors in the appropriate enclosure bottom-box and then plug the computer or machine to the box. You could chose to have the connectors fixed in the bottombox and already wired to the machine or the computer and plug in the connectors when you mount the top... or am i missing something? 21:06
Beautiful! SMOP😏
if the ferrils sometimes fall of then its the force the crimpers put on them that is to low, mine had a screw i could loosen to turn a little wheel making them krim much harder onto a wire ( that could also dammage wires if to heavy )
I was very pumped for the first power on just being lights blinking in the right way.
20:51 I would suggest replacing the RJ-45 patch cable with a slightly longer one, so it's about the same length as the other wires. Other than that, beautiful job!
Nice. I'm used to doing this stuff for a living, in the LA area (HAC MTS3). I have always liked this stuff but there is no more work for it. Besides, at 70, I am long retired anyway. ;) Nice project. The software will be the fun stuff ;)
Interesting the lights on the board were dimmer at different voltages. I would have hoped the LEDs where driven by a high impedance (transitor, mosfet, etc), but given they are dimming as soon as the voltage starts to drop, does that mean the LEDs are trying to pull typical LED current (10-20ma) from the source feed rather than the board power feed?
No, the LEDs are driven by the MCU thorough another channel. They aren't electrically connected to the input.
I see you are all about that base, that base. Very good...
The only thing that could make that cable management look better would be some old fashioned cable lacing.
The SMD resistors are very brittle and will with time break if the switch terminals move at all. Speaking from experience. 1/8W axials are small and reliable for this purpose.
What gauge are the screen power supply cables in the new wiring scheme ?
Just so you know it’s bad practice to twist the conductor strands before insertion into the terminal [ in your case bootlace ferrule] according to IPC/WHMA-A-620D.
Was thinking the same thing, but if you look closer it seems that he's not actually twisting them, but rather "tidying up" some stray strand with a very light twisty-turny brushing motion.
When I use ferrules on small wires I like to remove more insulation off of the wire so I can double over the wire before putting it in the ferrules . It makes a strong connections and I can use a slightly larger ferrule.
Careful with surface mount resistors they very susceptible to cracking due to mechanical stress I'd be quite nervous having them floating in mid air between switch terminals.
My ocd is going nuts those wires not having sleeving on.
Maybe you've talked about it already in another video but what digital notebook are you using? Do you like it? Would you recommend it or another one? I've been looking into making a similar purchase. Thanks
You are 10x smarter than me, so hard to have a comment but I would suggest always using female connectors on a bulkhead panel and use male-to-male cables because if you break a pin, repair or replacement is way easier.
Am I only one finding ABS-GF parts to cause itchy feeling on my skin? If not how to deal with that?
Aslo for projects like that, maybe modular plugs would be good idea? ODU is expensive but having 3D printer parts except pins and inserts chould be done inhouse. Cheers
I’m concerned that if your resistor chain is varying the intensity of the on board LED, the current in the chain may be wrong and hence the voltages will not be what you expect. Measure and see!
The LEDs are driven by the MCU. They aren't electrically connected to the input.
My $15 Amazon ferrule crimper works amazingly well, even down to very thin wires. Never had the ferrules come loose. If you can afford to buy Knipex, sure, but definitely not necessary.
James, I don't think you explained why that ClearCore board isn't just integrated into the control panel in the first place.
what printer was that printed on? the quality looks AMAZING
Not sure if someone else covered this, but I expect the 0805 resistors will crack and fail in operation
That notepad looks really cool. I'm on the verge of springing for one.
make sure to check if its tied into a cloud subscription. allot of them sell you the digital 'pad' and then you need to pay for a cloud subscription or the pad no longer functions. pretty shitty.
I have a remarkable like he has, its pretty nice, we aren't a part of any subscription, but I do think it requires and account be made.
Why did you not just connect the piece with the connectors to the newly printed housing instead of the base. If the connectors are part of the upper part then it is way easier to move around.
Wait a minute. That's not regular graph paper! What electronic device is that? Looks like a very handy device...
ReMarkable 2.
James, what is the electronic pad you are writing on?
Also want one. Someone in comments said it was a remarkable
This is some really impressive work, James. The hardest part of digital control for me anyway, is the interconnection between the software and the machine. This really helps me wrap my head around the entire process.
It's too bad you are not in charge of design for the god-awful mess in automobile design these days. It is nice to see perfection.
Good luck James.
kind of pushing the wattage on those 0805 resistors in total, since they are split up its not that big of a deal but I would keep an eye out on the 1.4k combination (yea I know you mentioned it but run of the mill no name 805's are only 0.1 watt and your at 0.1 watt with the math)
Use a string of Zeners for the voltage ladder
I was thinking that you gave up on this project, what a nice surprise
Looks nice James. I will be patiently awaiting you next installment on this series. I like the use of the SMD resistors on the switch, it makes for a clean simple installation. Some thoughts for consideration. I have used spiral wire wrap to help protect and clean up an installation of multiple wires on some control panels I have made before. Personally, I like that stuff more than just using wire ties (zip ties). Wire ties alone can pinch wires and possibly cause damage, especially if they are cinched too tight. Admittedly, the spiral wrap can be a pain in the rear to use when the wire harness gets really long, but the end product does look clean and tidy.
Not that any of the beautiful work you did is invalid, but I’m my world we just make two (or more) protos, one for software dev and the other for hardware. Swapping between platforms takes too much time. Of course, that does require an easy way to update the firmware in the hardware proto, but that is commonplace these days.
So the trolling was putting the resistors in mixed orientations?
26:30 resistors not all aligned the same way up ... for shame.
21:23 Piano hinge, with a loop bent into the end of the hinge wire, is a convenient means of attaching similar panels. I don't mean that you should change what you have now, it's an idea to tuck in a grey cell for something down the road.
I'd have used a 1k5 or 1k6. There's no need to chase the precision with the error margin you have in the voltage steps.
I often wonder why industrial applications like this don't fully "modernize" connections to USB-C.
If you are going to be programming anyway, surely a single cable to carry power and all the data you could possibly use would be an ideal solution?
Ahhh! .....no blue smoke!❤
great video!
maybe put a tvs or zener diode across the switch resistors in case the power supply puts out a spike in voltage. might save your clear core, but it also might already have protection built in ^^
What material did you use for the support on the adapter plate thing
ah, that typo on the remarkable confused me for a bit, its 14v that needs to be dropped, not 10v
James, what setup do you use to get those steady overhead shots?