Custom Keyboard?

[Clive66]

Having tried a touch panel my conclusion was it was not viable for manual work as the feel is inferior to a keyboard.
I note that custom keyboards are available on the internet and would like advice as to how I might obtain one with the correct number of keys and layout required for H5.
My reason for wanting to go this route is space saving as I find the current compact keyboard could be improved on.
Any ideas and suggestions apreciated.
Thanks.

[GeckoSub]

Hi @Clive66
A few thoughts on this but honestly, it's just from me going into this theoretical rabbithole myself recently - as I haven't actually finished my build:)

I agree that tactile, real button keys are a real benefit on a machine tool. Also, I personally, do not want to have that (almost) full size keyboard around, though it's fine for debugging or starting on.
As I see it, there a few different paths to take:

• Buy a keypad with a number of keys that more or less fit our requirements as you're proposing. Most of these will be USB these days and the FW code can't read that, as far as I understand. I don't know what it will take to make that happen. Maybe not a biggie but there could/should be some PS/2 kaypads around, too which would make things easier on the code side. I looked into this approach a while back but then ended up thinking I would do the following:
• Make your own keyboard with "old school" keyboard keys. Look up "hand wiring keyboard" and tons of tutorials will show up online. I actually bought the key switches and caps and was going to do this but then changed direction. Keep in mind, you will need a subplate to mount the switches on, but this can be 3D printed on a good printer (or CNC milled, laser cut, etc). But it would get you the exact layout you'd want. You'd ideally also need to find a way to mark the key caps, though. (I still think it would be a cool way to do it, but I abandoned it because I have too much other stuff going on).
• The third option as I see it is to keep the touchscreen and add a few simple push buttons where it makes the most sense. This is the route I am taking.

Here's a screenshot of where I am heading:

The screenset in the render is one that I am working on these days but it's intended to work seamlessly with Max's latest touch FW - with a few tiny additions/alterations. But then I have - in this panel itiration - six hard buttons that can be mapped in code to do the same thing as some of the touch buttons would do. You can add less or more buttons, of course.
Since I have not yet used any NanoELS on the machine, I am still undecided on which functions makes the most sense to have on the hard keys. Start/Stop and possibly jogging could be some. That said, stuff like setting limits, changing modes and inputting values can safely be done on the touchscreen as there are no time constraints on those.

Actually, and this is different than most users, since my machine wont have an X-motor I am looking into getting a semi-automatic threading mode coded and whatever buttons end up controlling that function will be the most important on my panel (it will likely be the left/right jog buttons).
BTW, those push buttons are 12mm in size and cheap as chips. Also, I should mention that I have a 7'' Nextion so that may help the touch user experience over a 5'' model. But there are ways to improve the screensets, either mine or Max's or anyone else's, by e.g. removing the numpad from the main page and making it a pop-up when needed. That would free up more space for other functions.
Also, on my panel, I have the lathe spindle motor controls, too - so the direction and speed to the very right. No, e-stop, yet admittedly. I may put one on the side of the panel. But my goal is to have everything consolidated into one panel. I am thinking that others will likely keep the spindle controls on the lathe so that would add back some space on the panel to add more push buttons. Same if you don't have or need the handwheel(s). Also, the push buttons could go on a separate panel, too of course.

But enough about me, back to you;). Since you mention your two main goals being having tactile feel and saving space, I think that keeping the touch FW and then adding a few push buttons might be the ticket?

I'll tag @gueee here as he might be able to share some valuable input. He actually made a few custom keyboards already, maybe he can drop us some pics here:)

@kachurovskiy I asked in another thread about mapping push buttons to functions in the code. You dropped me some code in that thread, but am I correct in assuming it should be doable to map all touch functions to push buttons? I am asking since my proposal to Clive above depends on it, haha.
With my non-existant coding skills, I still haven't been able to make the scrolling buttons (like mode) and hold down buttons (like jog) work, yet but I do have limits working on push buttons - as a test, I don't need them on real buttons. But that's likely just a matter of getting the code sorted, right?
(Of course, without wiring hard keys/push buttons in a matrix, there's a limit to how many push buttons we can wire straight to their own input depending on what other peripherals are being used).

[kachurovskiy]

assuming it should be doable to map all touch functions to push buttons?

Yes, you can map any touch code to any button code and we only use button codes in controlling logic so keyboard always have same or more features than touch.

[GeckoSub]

assuming it should be doable to map all touch functions to push buttons?

Yes, you can map any touch code to any button code and we only use button codes in controlling logic so keyboard always have same or more features than touch.

@Clive66
See Max's reply,
This means, we can indeed map a push button to any keyboard/touch function. It's "just" a matter of changing the FW code a bit, but I will get that sorted for my own needs soon enough - with or without help, we will see:).

There is a limit to how many buttons you can wire to the ESP32 without having to wire them in a matrix (not smart enough to explain what that is, but again it'll be explained in the "hand wiring keyboards" tuturials if need be).
Personally, I don't need many push buttons and since I am dropping a lot of the other peripherals - no X/Y motor, just one handwheel, no joystick - I should have enough available inputs on the ESP32 for my buttons.

As for the code for the push buttons, if enough of us take this on, perhaps Max will either integrate it in a later revision or help us with some working examples of the code:).

One thing that's obvious for the guys who know about this but that I just realied is that these buttons need some debounce in the code. I found a library that does it - and one that ChatGPT knew well enough to get the code right - so my test button works super well now. Never gets a push wrong:).
On that same note, there's another library that I really like the idea of using as it does long press, double click, release event, etc and it names these commands in this exact lingo and not ProgrammerSpeak but sadly, ChatGPT sucks at using this particularly library and since that's how "I" do coding, I am struggling with that library. Either way, for others, it would be a walk in the park:).

Point being, maybe this stuff will get incorporated into the FW at which point setting up push buttons will be as easy as defining pins and binding them to the key codes which is already in a table in the FW.

[GeckoSub]

Point being, maybe this stuff will get incorporated into the FW at which point setting up push buttons will be as easy as defining pins and binding them to the key codes which is already in a table in the FW.

Actually, @kachurovskiy sorry to bother you - especially on a Sunday/vacation.
But when you find the time: Can we map the push button to the key code? In my working example (left limit), I am having to use buttonLeftStopPress(&z) and honestly it hurts my non-programmer brain.

I wish I could somehow just map it to a number in your PS/2 table like 65 in #define B_STOPL 65 // a - sets left stop
or
[40] = B_STOPL, from the touch ID table.

Is there any way, the push button code could be made as simple as just using either 65 or 40 in this example instead of buttonLeftStopPress(&z)?

[Clive66]

Thanks for all the info it is much apreciated. Having read your post it would seem a combination of touch keys and manual would suit my application perfectly.
Personally, I was happy with all of the touch screen controls but would really like manual (push button or keys) control of L-R & U-D for the Z and X axis.
I have the 5 inch Nextion display and do not plan to change it at this time so that will save some work.
Not being a coder I will watch this space , again, thanks for the replies.

[GeckoSub]

Thanks for all the info it is much apreciated. Having read your post it would seem a combination of touch keys and manual would suit my application perfectly. Personally, I was happy with all of the touch screen controls but would really like manual (push button or keys) control of L-R & U-D for the Z and X axis. I have the 5 inch Nextion display and do not plan to change it at this time so that will save some work. Not being a coder I will watch this space , again, thanks for the replies.

Yeah, I came to the same conclusion for my own machine. A few tactile buttons for the most critical functions do make a lot of sense.

I will try to get L-R working one of these days as I need them, too and if I don't crack it, I am sure we can ask someone to come to our rescue:). Also, Gueee may well implement it in a branch of his code, or he might already have done so, so I will check there, too.

Anyhow, I will circle back here once I find a working solution.

[GeckoSub]

@Clive66
It's Working
Good news from one not-a-coder to another: The buttons are working (I think)!

I spent yet another half day yesterday with ChatGPT trying to get the push buttons to work to no avail.
So, I threw the towel in and asked Max for help and he said to try another AI engine and, I think, agreed on my general idea on how to go about this - but put it in better words.

Long story short, I just spent a bit of time with that other AI and it nailed it:). I have the left jog button working now on the bench. It does continous jog and incremental just as the keyboard does:). Something I never managed to do before. I will wire up a few more buttons later today and then test a bit more. But I am pretty confident this works.

Furthermore, I did some "pin counting" and if you don't need any Y axis stuff at all (I don't) and you don't plan on ever using a joy stick, it looks like we have 10/13 pins that can be re-mapped to use for our buttons without having to resort to any kind of matrix wiring. Also, this still leaves the X/Z scale ports untouched for future DRO use and also keeps the keyboard port which would make sense in terms of having a backup.
(I say "10/13 pins" as three of them can't be messed with during boot up as it might make the ESP32 confused. So, 10 pins are super safe to use and if you need more, then just don't touch any of those three extra buttons connected to the last three pins during boot;))

So, I think in a day or two, I can share a modified FW here - just remind me if I forget.

Which Button Functions
In terms of which buttons, I am thinking that since you want L, R, U, D you will be using them to jog to set zeros, right? So, I think it might make sense to add the increment button, too. That way you can jog coarsely and precisely without having to go back and forth between the buttons and the touch screen. That said, on my custom screenset, I did increase the size of the increment button for ease of use.

Personally, I may add +/- push buttons, too. For two reasons: I think - keep in mind, I still haven't used this ELS ever on a real machine - that when in general turning mode along the Z, like power feed basically, we will be using the +/- to fine tune the feed, right? I don't see how else to do it. Again, it's not super time nor safety critical so it can also just stay on the touchscreen. But the second reason is that I have managed to make +/- work with hold down mode, too. So, you can now use them to scroll through feed values rather than single tapping the key to no end. This should make using it to adjust feed on the fly more viable.

(I don't actually need you to tell me which buttons, you want - as mentioned, you will be able to pretty easily change it in the code yourself and I will make sure you can find where exactly to do this pretty easily, too.)

[Clive66]

Like you, I have no need of a joystick. I use my machine manually mostly hence the need for L. R. U. D. buttons. I agree adjusting the feed on the fly with buttons could also be very useful .
If you are able to share the fruits of your hard work (modded FW) that would be most apreciated. The ELS has really helped me and makes machinig tasks, especially thread cutting so much easier. Max is a diamond!
A combination of touch controls and buttons will be the icing on the cake.
Again thanks for your hard work.

[GeckoSub]

@Clive66
OK, so I have a custom FW now. But keep in mind this is my first foray into coding and AI did the heavy lifting. So, please be very careful as I have only tested this on the bench myself - not on the lathe, yet. And please read the ReadMe file in the zipped folder, too.

First off, push buttons should be wired between ground and the signal pin you choose and define in the code.
If you don't want to change any pins in the code, then the jog/arrow buttons (LRUD) all get wired to the joystick connector (JSIGNAL). But you need to get ground from either the port just below it (JPOWER) or anywhere else you see fit.
I also added two buttons for increment and cycling through the modes in the FW already.

Here's how'd you connect the buttons unless you want to change the pins in the code:

• JOG LEFT connects to "JB" on the PCB's JSIGNAL connector
• JOG RIGHT connects to "JY" on the PCB's JSIGNAL connector
• JOG UP connects to "JX" on the PCB's JSIGNAL connector
• JOG DOWN connects to "JZ" on the PCB's JSIGNAL connector
• INCREMENT connects to "IO10" on the PCB's MISC connector
• MODE CYCLING connects to "IO11" on the PCB's MISC connector

If you do wanna mess around with the code, then read the first few lines in the code to see the search phrases so that you can find the actually place to change the pins but in this version of the FW it starts on line 974.
Also, the ReadMe has a table of what the different commands/functions are called - you would need that, too.

Also, I will add the tft files for the latest screenset to my own "build thread" a bit later and send you a link here when done. That said, the custom FW is made to work with Max's code so his screenset should be more than fine, too. It's just a matter of taste. But I tried to clean mine up a bit, get rid of all the Z-stuff and enlarged the coordinates and numpad.

h5_V11_NO_DRO_GS15.zip

P.S.
One issue I tried to tackle is the one where it's very hard to get any button - touch or physical - to just do one step/increment at a time. If you don't tap super fast, it will go 2-3 increments. It's in the original FW and Max knows of it and I think he will be working on fixing it.
I spent hours on it today and it's very hard for the AI to crack, too so not much progress. The rough fix makes scrolling too slow.

[Clive66]

Perect! This will give me something play with on my test rig. All being well I will test on the lathe.
Thanks again for sharing.

[GeckoSub]

You're most welcome:)
And great to hear you have a test rig. Actually, just from playing around with the jog buttons on my bench, I do think you might want to try that 5th Increment button. On my small CNC mill, I use the increment switch on the pendant all the time so I think it will be nice to have right next to the ELS jog buttons, too.

There are a few other small improvements in the FW:
• +/- now has hold down mode so you can scroll through feed values
• Decimals are stable now, always 2 in metric, always 3 in imperial (sort of, I hope).
• If you use my custom screenset with a mono space font this means the values don’t jump left and right when jogging
• Long pressing the Left Stop (Limit) deletes all limits.

As for the speed/timing of the hold down and long pressing, you can mess around with those in the code, too. Again, look at the top of the code for search phrases on how to get to these values.

Anyhow, dinner now (I am in Asia) and then I will compile the newest screenset for you.

[Clive66]

"I am in Asia" A part of the world I know very well! I will get started on the new FW tomorrow customising is all part of the fun.

[GeckoSub]

@Clive66
The screenset is here if you wanna take it for a spin:
#345 (comment)

[gueee]

@Clive66

I slaughtered an old ps/2 keyboard and repurposed the PCB, printed a base plate, a front plate and the key membrane from tpu.
hand wired cherry knock off key switches to the PCB, used the debug function to print the key codes to the display, deal done.

If you like I can share the print files.

[GeckoSub]

@Clive66 I slaughtered an old ps/2 keyboard and repurposed the PCB, printed a base plate, a front plate and the key membrane from tpu. hand wired cherry knock off key switches to the PCB, used the debug function to print the key codes to the display, deal done.

If you like I can share the print files.

Ah, I finally get it. You have regular MX style key switches under the membrane. That was what I never figured out, I thought you were talking about two different keyboards and assumed your membrane version had those small short travel SMD buttons.
I think the membrane makes a lot of sense if you want your keypad that close to the action. If you want it up and away, e.g. near the headstock, you can probably get away with regular key caps.
Personally, I am beginning to really like the MX switches I bought and then abandoned - so they might be making a come back. Also, I only need six to eight of them, maybe only four. So, wiring them is much easier as they can each go to their own pin.

As for my own panel - I thought I was only days away from machining it as the plastic "billets" were supposed to be here by now. But I just realized there was a hickup and it never made it to my freight forwarder in China. So, it never went on the cargo ship to me. And even if we reissue the order today, it'll be another 3-4 weeks before it'll get here:(.
This hardly ever goes wrong but my forwarder recently changed warehouses and I think there's when the mishap happened.
Anyways, I might go look for some scrap plywood, layer them up and cut a dummy in the meantime.

[Clive66]

@Clive66 I slaughtered an old ps/2 keyboard and repurposed the PCB, printed a base plate, a front plate and the key membrane from tpu. hand wired cherry knock off key switches to the PCB, used the debug function to print the key codes to the display, deal done.

If you like I can share the print files.

Yes please the print files would be useful

[GeckoSub]

Custom Keycaps
Since this thread is titled as it is, I thought I should share this little teaser:

Fresh from the "lab" as I call my little workshop: 3D resin printed keycaps, colored with nail polish;)
They are for classic MX style mechanical keyboard switches.
This was just a proof of concept but it's promising enough that I will drop the round, stiff push buttons on my build and go with keyboard switches and homemade caps instead.

More of an informative "build post" here

[GeckoSub]

Setting Feed With A Potentiometer
@Clive66
I have another feature kinda working now, but that one might be risky: I got a potentiometer to work for setting feed:)

The idea would be to only use it for general turning, but as a faster way to set and adjust the feed on the fly.
That said, it may not be needed, maybe the numpad and +/- buttons are fine. Especially if you know your material and often cut the same stuff, then I guess you will find "your" values easily and not stray much from them.

The feature works (only in Gear mode for now) within a smaller range that I defined as 0.0 - 0.15mm/rev for now. So basically, the potentiometer can set any value in that range in increments of 0.005mm/rev. So, it would go e.g. 0.070 -> 0.075 -> 0.080. I also made the number of decimals stable at three so it looks nicer on the display when scrolling through values. The range, the increments and the number of decimals can be changed in the code pretty easily.
It does have "Last-Action-Wins" coding so if I type a new value on the numpad that takes over until I use the pot again.

But since it's an analog signal it is susceptible to noise so the risk is that the value can jump in the middle of an operation. I got it to a point in the code now, where noise doesn't make it jump - but that's on the table at home. And the electricity in my workshop is super, super noisy as we don't have any proper ground at all (at least, I think that's the main culprit but I am not an electrician, either). And then the BLDC motor (plus perhaps a servo) will be noisy, too on the real machine.
I will see what happens once it's mounted.

If noise is an issue, then the solution could be to move the feature to an encoder instead of a potentiometer. Or pair the pot with a button so that any set value gets locked in and basically disables the pot until it gets turned again - less elegant with the extra press though.

I guess I just wanted to ask if it's something you think people/you would actually use?

[Clive66]

Personally I would be happy with your custom firmware as is. I am guessing other users may well find this feature very useful however.

[GeckoSub]

Personally I would be happy with your custom firmware as is. I am guessing other users may well find this feature very useful however.

Yeah, but also it's one of those ideas that sounds OK but may not see that much real life use:). Anyhow, I thought about it a bit more and I think I want to try it with an encoder, too. That also makes sense as then people could possibly use one of their handwheels in double duty mode, if they have any on their build.

[gueee]

So, it would go e.g. 0.070 -> 0.075 -> 0.080.

if you do it like that, why don't you use a rotary encoder instead of a potentiometer?
pretty much the same form factor and you could even use a push button rotary encoder to confirm the setting.
this way you would avoid any accidental pitch change.

[GeckoSub]

So, it would go e.g. 0.070 -> 0.075 -> 0.080.

if you do it like that, why don't you use a rotary encoder instead of a potentiometer? pretty much the same form factor and you could even use a push button rotary encoder to confirm the setting. this way you would avoid any accidental pitch change.

Yep, I agree - already on the to do list of things to try and was looking at encoders for this exact thing last night incl. the push button:)
Also, with an encoder I can make it always start at 0.00 or any value that seems a good starting default value.

[GeckoSub]

Custom Keycaps - Take Two
I made a new batch of keycaps, a full one this time. And I think I will settle on 8 keys:

This time, I made the top of the caps flat so it was easier to sand off the overflow of paint. I thought I wouldn't like the look of that much, but I think it's fine. Almost a more utilitarian look in a nice way;)

Also, I switched to pink for the stop button as the red color doesn't show well on those dark caps.

As before, the longer write-up is on the other thread and once I have settled on the final design, I will share some STLs and the project file so you can make your own text engravings.

[Clive66]

Today I had free time so uploaded the firmware and .tft files. Fantastic job!! Thanks to everyone that contributed.
After testing I have decided on using buttons for the Step key in addition to UD LR as this arrangement would suit me perfectly.
If the tactile feel of buttons is inferior I will attempt to make up a keypad as described above.
Again thanks to all.

[Clive66]

Good job! I am thinking of using one of these (https://ebay.us/m/Spqll8) for tactile keys as I already have a working bluetooth to ps2 adapter.
I just need to figure out how to progam the keys.

[GeckoSub]

Good job! I am thinking of using one of these (https://ebay.us/m/Spqll8) for tactile keys as I already have a working bluetooth to ps2 adapter. I just need to figure out how to progam the keys.

Don't know much about those little pads except I keep thinking I should get one for all my photoshop and nowadays Fusion work (I am freelance photographer with too much time to tinker these days...).
For our needs as long as the PS/2 signal comes out fine without delay I guess you can use that debug function Gueee mentioned. If I understand correctly, it's in the code already, you just have to uncomment it and reflash and supposedly, it will then show the key press codes on the Nextion screen. That way, you can see what your keypad actually sends out and then go back in the code and change it to "bind" your pad.

Or maybe one of those apps that you can program these pads with can also just change the key code to match what the code already has - then you don't need to touch the code.

[EDIT]
Found the debug line. At least I think this is the one that Gueee mentioned.
This is how it looks in the code and the AI lords left it in "my" code, too - so search for "hex codes" and you'll land right at it:

// Uncomment the line below to see the key hex codes on screen. For debugging
// setText("t3", (isPress ? "Press " : "Release ") + String(keyCode));

[Clive66]

I have now used your latest screenset .tft on my machine with Max's firmware. Will test your firmware when my buttons arrive from China.
All is working well and functions so far with either touch screen or pS2 keyboard. I love the green numeral display option this is very useful for me when machining.
Just one minor suggestion . The RPM, Angle and Turns display were originally cycled by the Tilda key and now all three are permanently displayed. Personally, I would prefer a bigger RPM display box and retain the cycle feature.
From your experimentation how difficult would this be to achieve?

[GeckoSub]

I have now used your latest screenset .tft on my machine with Max's firmware. Will test your firmware when my buttons arrive from China. All is working well and functions so far with either touch screen or pS2 keyboard. I love the green numeral display option this is very useful for me when machining. Just one minor suggestion . The RPM, Angle and Turns display were originally cycled by the Tilda key and now all three are permanently displayed. Personally, I would prefer a bigger RPM display box and retain the cycle feature. From your experimentation how difficult would this be to achieve?

Thanks so much for the feedback:)
Personally, I am all for a larger RPM box, too. And yes, I can make a revision for you (and me).
I will load Max's tft back in a day or two and see how he did it. But I thought that in his touch screenset they were all displayed at the same time, right? Maybe in the non-touch version it was cycled. Would you need them at all?
The easiest would be to just have RPM, I can do that without changing the code, but only the tft, I think. That would be the quick fix.

Because toggling through them and having them display in the same box might need some deeper code surgery as Max split them up into three different text box outputs in the new touch code. So, maybe first step would be only RPM and then I can see if I can get the cycling back later on.

Or if you only need one of the remaining two let me know which one. And if I can make it even smaller to make room for a larger RPM readout.

Also, do you work in metric or imperial? There are few areas where imperial is bit neglected in my screenset/code;). Nothing in terms of functionality, I think - mostly just tiny aesthetic stuff.

[Clive66]

Personally, I use Metric and imperial. With regard to my suggestion, the three options all have uses. For example the angle feature can be used for diagnostic purposes.
It was only a minor request useful if only a samll code tweak was needed.
Overall I am extemely happy with the latest improvements.

[GeckoSub]

Double Duty Numpad
In related news, I got the double duty numpad working today.
I was actually about to say that if you thought you can find room for 10-11 buttons, then it might be worth it - because then we can make a screenset with no numpad on it, which would help grow e.g. the RPM readout.

Did the hardware in the workshop earlier today:

(superglue for work holding - works well on one offs where you need access all around the part. Just be wary of the temps)

And I literally just finished the first version of working code for this. Took battling the AI more than I would have liked, but it's almost there and I have to remember that it's pretty crazy that a non-coder can even get anything like this done:)

[GeckoSub]

@Clive66
I made it happen:). One box only for larger RPM/Turns/Angle values...

But there's a "but". It took a fairly deep code dive so it will only work with a new version of my custom FW. So, be extremely careful, please!
More here, including the files (and a video if you wanna see me babble on).

[GeckoSub]

Will test on machine when buttons arrive. Thank you!

Great:)
Also, unless my FW has some nasty bugs in it, I think you will find that the screenset behaves even better paired with it. I really like the fact that the coordinate readouts in terms of decimals are stable and don't jump left and right when changing values. It's easier on the eye:)

In other news, I think I may just have cracked the DRO challenge which was the very reason I came to this forum in the first place. When I first posted, I had no idea that I could even try to do it myself so I am honestly pretty stoked right now;)

[Clive66]

I have just posted an issue with the new touch screen software which seems to be present in the official software and also your screenset.
Maybe Max can advise.

[Clive66]

Max figured out my mistake instantly! Sorry for wasting your time, both screensets are working perfectly.
I guess we can put this down to a senior moment on my part!

[GeckoSub]

Max figured out my mistake instantly! Sorry for wasting your time, both screensets are working perfectly. I guess we can put this down to a senior moment on my part!

Yeah, I saw that - that's cool.
I didn't even know that part of the forum existed. There's some good stuff over there - and I guess it's where we can reach Max if he's more urgently needed:)

You will find one issue, but Max is aware and I mentioned it somewhere else, too - but it's very hard to do just a single increment jog - most often you end up getting 2-3 out of a single touch of the button. So, just be careful and perhaps switch over to a finer increment step earlier than you would have normally when sneaking up on a part.

It's in the stock code and the issue is the same whether we are using the PS/2 keyboard, touch display or external buttons. Just have to live with it for now.

[Clive66]

I am used to touching off using the fine increments so that is not a problem. Once you get used to using the ELS for actual machining it really makes life easier.
I would never go back to a manual machine. Hopefully the order from China will arrive in the next few days and I can road test the new developments.

[GeckoSub]

Do Not Use Stepper Pins For Buttons!
Since I just wasted a whole lot of time battling the AI to get push buttons to work on the left over Y-stepper pins I guess I should share that you can't, haha...
It's probably something about the level converters only allowing one-way traffic - so we can't repurpose any of the stepper pins for push buttons.
It's funny how not long ago I was sure I would have more than enough pins for my buttons, but now I feel constrained;).

Maybe it has something to do with the this little thing growing:

It has an encoder now and I am already thinking about squeezing in two more keys... Feature creep. I should really, really stop. Now. Stop. Haha.

[Clive66]

After waiting for some deliveries I have now completed my set up. I already had a Bluetooth to PS2 adaptor that I made for a Bluetooth keyboard. I have now added two progammable mini keyboards and have set the left keyboard to activate limits plus off and on.
The right keyboard controls Z axis and X axis manual moves also pitch and mm/imperial.
Using this system any key can be programmed to the function you require ,
All that remains for me to do is make some custom labelled keycaps.

[GeckoSub]

@Clive66
Granted, I am not sure what kind stuff you make with your machines and your interests but for many of us, a 3D printer is one of those things that once you have it, you find more and more uses for it. And they have improved vastly in the past 2-3 years to almost work perfectly out of the box whereas before you could spend as much time tinkering with the printer and setup as actually printing. Also... depending on where you are in the world, there might be enough people who bought one and still felt in over their heads leading to good deals on used ones.

I have a 7-8 year old FDM printer - of the kind that needs constant attention - which I will probably upgrade once I find a good deal on a more modern, used one and then I have two resin printers but that's mostly because I sell a product I make on them. For most things, FDM is a more versatile, cleaner process.

[gueee]

as a heads up regarding 3D printers....

I just bought a formbot xplorer kit, a very compete kit, with precrimped wires and everything in the box you'd need, except the printed parts.

if you have that money to spare, I can highly recommend this printer.
once built, the experience is far out, I finished it 4 or 5 days ago, run already 4 kg of filament (ASA-GF and ABS-GF) through it and haven't had a single fail.

I run a small print farm and reliability is key for me.

[GeckoSub]

@gueee
Congrats - looks like a super sturdy and capable machine:)

But... since you mentioned print farm, I have to ask and drift this thread, haha: Do you have any experience with off the shelf offerings enclosed printers that you could (almost) recommend? Qidi, Bambu, Creality?
I wont run a farm so don't need the farm app stuff or crazy hours between maintenance. Just asking if you know of something that runs fine for a while and has no big issues.

I will, as with most of my gear, probably get a secondhand offering sourced directly from China hence the Chinese brands mentioned. Before the latest gen of enclosed printers I was very close to getting a large delta. They used to have a speed advantage but maybe not so much anylonger.

[edit] Just looked up your printer. I am not in the market to spend that much, BUT that looks to me like a very fair price considering what e.g. the H2 series is going for now. And good to see a Chinese company seemingly doing proper open source.

[gueee]

there are a bunch of good printers that I could recommend.

for the most part you can't go wrong with a qidi plus 4,a reality K series (k1c, or later k1, or even k2), if you need a bigger one a sovol SV 08 is a good candidate, just look for one that already has the enclosure added, cause from factory it comes without it, which isn't that much of a problem, since it's basically a voron 2.4.

if you want a broader range of filaments to choose from, I'd go for the QIDI plus 4, it's got an active chamber heater and reaches quite high nozzle temperatures.

[GeckoSub]

there are a bunch of good printers that I could recommend.

for the most part you can't go wrong with a qidi plus 4,a reality K series (k1c, or later k1, or even k2), if you need a bigger one a sovol SV 08 is a good candidate, just look for one that already has the enclosure added, cause from factory it comes without it, which isn't that much of a problem, since it's basically a voron 2.4.

if you want a broader range of filaments to choose from, I'd go for the QIDI plus 4, it's got an active chamber heater and reaches quite high nozzle temperatures.

Thanks for these reccos.
Yeah, the Qidi Plus 4 was high on the list. If I was buying new, that would likely be what I would get. They aren't on the used market, yet though. But the Q1 Pro is. And there should be loads of Creality ones out there and possibly Sovols.
Anything will be better than my old Anycubic i3 bedslinger even though it's klipperized and has some nice extruder and hotend upgrades.
If I buy very soon, it will be a Q1 Pro or a Creality K1C/K2 and if I can wait a few months, perhaps the Plus 4s start turning up.

I just got a Saturn 4 Ultra resin printer for a bit more than half price with just 20 hours on it and all the original packaging - which was awesome for shipping. But the resin printers can be had used for a steal as too many people still get fooled into thinking they are super easy to run or are fine in their small studio apartments. So, they buy them, get surprised and then dump them on the used market. Not so much with FDMs.

Anyways, thanks for your thoughts:)