I did a search and all the discussions I found on duel endstops are either old or kind of incomplete. So starting a new thread to see what the community thinks.
In that last one @vicious1 says that 99% should be fine with serial but that was in September 2019 and there doesn’t seem to have been much since. Is that still the case? Anyone start series then go to duel?
Repeatability and auto-squaring certainly seem like worth the extra firmware and hardware build efforts to me but then there are other threads saying that no one should need it so I’m not sure which to think.
Thanks for posting this. I’m wrapping up the printing of my A color parts. And I’m starting to buy the other parts needed so this is something I started to think about. Hence your post is very timely for me as I basically have the same questions.
If I had to pick one solution for everyone, I would make everyone get the hardware for dual endstops, and leave the switches in the bag until they a few projects worth of experience at least.
There is a lot to learn out of the gate. The “checklist” for setting up a job can have a lot of caveats and you need to understand what you are doing with CAM and in gcode to set up a job. Dual endstops makes that a bit more complicated (not a ton, but some) and it would be better for people to do the first few projects with as few ways to screw it up as possible.
After that, you really can do the same thing, for the most part, without dual endstops. But when you get some experience, and especially if you are using the machine a lot, it can be very helpful.
The alternative that people don’t see is that if you start the two steppers square, they will just stay square. I have little blocks I pull my gantry against when I start a job and I am confident those blocks are pretty square. I have a low rider, and the gantry is 5’ long. I will notice if either end is even 0.5mm away from the stop blocks. 0.5mm over 5’ is more accurate than wood is in general, so I haven’t ever even installed endstops on my LR.
So, if you are in a situation where you want to reduce the barrier of entry, and save a few bucks, you can still open up the CNC world with either a mini rambo or a grbl controller and serial wiring. You really won’t be unhappy.
If you want The Best , then you can buy a dual controller now and either set it up right away, or wait until you have some experience.
When I got into this I knew NOTHING about electronics, mechanics and programming microcontrollers. (I still don’t know much honestly).
I got the Rambo with dual end stops, but did wire in series first, learning to build and use the machine. After a while I upgraded, and it was a pleasant experience. I really appreciate the end stops, especially for jobs when using different tools and splitting the jobs in smaller portions.
But - I’m glad that I waited with them in the beginning, it might have been too much to cope with. Figuring out the MPCNC was a challenge, and for me it was good to take it step by step.
I did the same as Turbinbjorm and highly recommend it. Keep everything simple and stock at first then make the upgrades. This was good advice from Ryan when I first started and it still hold true.
I love the dual end stops on my machine. I like the homing capability and to know I’m square on something. I also install the z-stop and love it too.
I built mine with dual end stops. Initially I had a hard time with the fact that it didn’t do negative numbers and I didn’t like working with offsets. I ended up tweaking the vcarve configuration file to disable end stops after homing and squaring is complete, this way I get the benefits of the end stops without the complications… As others suggested, order it with, build it without.
First off, thanks for all you do in this forum Jeff. I’ve been very successful building my first CNC because of you and Ryan’s design.
This post of yours is basically exactly what I did. I wanted to keep the initial build simple but I bought the end-stops and full Rambo board so that I could upgrade later. The only thing I don’t have is the wiring harness to power the motors individually.
Could you point me to a write-up or documentation that details what I would need to do to upgrade at this point? I bought the full rambo from Ryan so would I need any firmware upgrade or is it pre-flashed with all the correct code?
You will need to change the firmware (unless you are using dual and don’t know it). You can check what you have in two ways:
M115:. Shows the versions number. We added D to dual a while ago. So something like 509 is serial and 509D is dual. In 510+, we will have a 510S for serial and 510L for Low Rider, etc.
M119: dual endstop versions will show status for things like x2_min. Serial will just show x,y,z.
You will need to rewire the steppers. Each motor gets its own 4 wires.
At this point, you will be working again, but just moving around. It will be exactly the same, functionally but each motor will have its own driver.
After that, you can wire the 4 endstop microswitches (if you dont have any, make sure you get the kid from the shop. The red pcb ones are trouble). They need to be wired NC and C. You can plug them in and test them with M119.
At that point, you can try homing X and Y. Then adjust the squareness by moving the endstops or using M666 to adjust the offset for one endstop.
There are a lot more details in the docs. But those are the basic steps.
I am a total fan of dual endstops. My MPCNC (Primo) machine is my 3rd build – I have dual endstops on all my machines (LowRider, MPCNC Burly and Primo). I agree with keeping things simple at first, but plan to add them when the time is right. I hope this is helpful.
I could not get my steppers to work in series, so they were paralleled, although this increases speed performance it causes a big power loss though heat dissapation. I’m not a fan of either series or parallel, for me individual drivers seems more technically correct, obviously plenty of users run these configs so it comes down to choice. I just like the fact that the gantry squares itself.
It’s not a matter of feelings though, and you are answering a question for a beginner. Series wiring is not rocket science, and it works just as well as dual wiring. I would say it is identical after you leave the endstops, unless you are trying to go really fast (like really fast).
In a single driver for a single motor wiring, the driver is off for a huge portion of the time. In a parallel configuration, it is off just as long, but the power is split between the two motors (and if you just crank up the setting, it does overheat the driver). In a serial configuration, the driver sends the same amount of current as with the single motor, but it stays on for longer (and the heat is proportional to the current, so it stays just as cool). The only difference comes when either: a) you are moving fast enough that the driver needs to stay on longer than 100% or b) you are wanting to move the motors independently (with dual endstop homing).
There has been a long history where we all started with parallel wiring (like on most 3D printers with 2 Z) and we tried doubling the drivers (in hardware) and Ryan (I think) found the serial wiring and we all went to that. A ton of problems went away and then Ryan figured out how to do dual endstops in the firmware and we gained the ability to auto square. But from a power perspective, at the speeds we mill at, the series and dual are identical. Parallel sucks.
Thought I’d throw my two cents in here. At first I was really intrigued with the dual endstops and investigated getting another controller board since I have the miniRambo. Then I realized that the only reason I’d have them is for squaring the axes, a problem I’ve already solved. Blocks on the rails with adjustable screws get me perfectly square by holding the gantry in place and turning on the motors. However, I have found that for many projects I don’t need absolute square either. When I’m cutting out a box bottom and lid, or the contour of a wooden toy I’ve never seen a problem.
But, I have given up on detailed cabinetry with the MPCNC because it is not rigid enough for the kind of precision with different grains and cut loads, at least for the large bed I have. I have a much more rigid machine for that stuff and both stay reasonably busy for what they’re good at.
Having repeatability seemed nice for tool changes, etc, but I have found that the only thing I need to do on tool changes is to reprobe Z, since I don’t turn the motors off.
So, endstops are great for 3D printers, but not so much with the relative zero needed for a CNC. Just my opinion
, then you can buy a dual controller now and either set it up right away, or wait until you have some experience.
