• Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.

CX704 Lathe conversion

lucsimoneau

Well-Known Member
Hi all,
At the school where I teach we are in the process of doing the conversion to CNC of 10 units of the Craftex CX704 mini-lathe.
On the first prototype unit, the X and Z axis are already modeled up and 3D printed, ball screws with bearing blocks are installed and it's going well.
Students will have the possibility of returning the machine to it's original manual state if needed.

Now the question, in the past we have done multiple iterations of a mini CNC mill using the Mach3 USB Breakout board shown below. We have DC spindle RPM control directly from Mach3 using the 0-10V input of the potentiometer that came with the spindle motor (500W model). Spindle control for tapping is not in the scope of this build.
1733606399939.png

Going with the CNC mini-lathe, I would like to be able to do single point threading by controlling the Z axis feed with the spindle RPM.
We have installed an optocoupler with a magnet glued to the spindle and manage to get the index reading on Mach3 but it is very inconsistent.
Was trying to reproduce what Franco does in this video:

Looking for another solution that wouldn't be too expensive to adapt for these little machines.
Would even be willing to install an encoder like this one to get the job done:
1733606711547.png

The problem is the pricing for an Ethernet Smooth Stepper or Centroid Acorn board.

Any of you have some sort of suggestion? Trying to keep the cost down for my students.

As of now, plan A is to not have the threading capability.

Thanks!
 
Hi all,
At the school where I teach we are in the process of doing the conversion to CNC of 10 units of the Craftex CX704 mini-lathe.
On the first prototype unit, the X and Z axis are already modeled up and 3D printed, ball screws with bearing blocks are installed and it's going well.
Students will have the possibility of returning the machine to it's original manual state if needed.

Now the question, in the past we have done multiple iterations of a mini CNC mill using the Mach3 USB Breakout board shown below. We have DC spindle RPM control directly from Mach3 using the 0-10V input of the potentiometer that came with the spindle motor (500W model). Spindle control for tapping is not in the scope of this build.
View attachment 55585
Going with the CNC mini-lathe, I would like to be able to do single point threading by controlling the Z axis feed with the spindle RPM.
We have installed an optocoupler with a magnet glued to the spindle and manage to get the index reading on Mach3 but it is very inconsistent.
Was trying to reproduce what Franco does in this video:

Looking for another solution that wouldn't be too expensive to adapt for these little machines.
Would even be willing to install an encoder like this one to get the job done:
View attachment 55586
The problem is the pricing for an Ethernet Smooth Stepper or Centroid Acorn board.

Any of you have some sort of suggestion? Trying to keep the cost down for my students.

As of now, plan A is to not have the threading capability.

Thanks!
You may want to reach out to Expatria Technologies (a Canadian small business) as they built an excellent board which runs grblHAL for the PrintNC project. I don’t know if grblHAL will run a lathe but it’s worth a discussion.
 
Options:

1) LinuxCNC is free, you will need a $150 MESA board say 7i96S and a PC or RaspberryPi, requires a little bit of computer savvy to configure, but is hands off, no manual intervention required. Read up on "lathe macros". This is about as cheap as it gets, but demonstrates the concept:
There is a great video somewhere? that shows about 10-15 of the macros on a real machine, tapers, threading facing etc. I can't seem to find the link.

2) Clough42 method is pretty cheap and allows basically any thread type without any gear changes, but you still need to manually engage and disengage for each pass

Both are proven for single point threading.

I use the Omron encoder shown in your post above
 
Last edited:
Now I could get behind that. Keep the whole electro-mechanical conversion enchilada off the manual lathe & have a separate mini machine. Even at that scale I could put it to work. OK I'll have to start watching that channel. Thanks for the rabbit hole. See you in a year LOL
 
This is the first I'm hearing of Expatria but their offering seems very interesting. If I'm reading correctly, the FlexiHAL ($140CAD) along with a Raspberry Pi4 running LinuxCNC (say$80 CAD for 4GB version) is one option, and the other option is the PicoBOB-DLX ($63) and a basic PP breakout board ( $15) plus a computer to run LinuxCNC is another? I'm not fully clear on the system architecture - I need visual representation as the written words are failing me (I blame the long work day but I probably wouldn't get it any day).
The second one seems like the way to go for a bare bones project.
 
For the encoder, on my EMCO Compact 5 it has a disk with 100 slots in it for giving better feedback resolution on spindle speed, along with an extra index notch on the outside. 2 very inexpensive optical sensors and use to count the pulses, and LinuxCNC does the rest. works very well. You could 3D print the encoder wheel.
The sensors were something like this: https://www.digikey.ca/en/products/detail/tt-electronics-optek-technology/OPB492T11Z/1636971 I'll see if I can find what I used exactly.
IMG_3123.jpg
 
This is the first I'm hearing of Expatria but their offering seems very interesting. If I'm reading correctly, the FlexiHAL ($140CAD) along with a Raspberry Pi4 running LinuxCNC (say$80 CAD for 4GB version) is one option, and the other option is the PicoBOB-DLX ($63) and a basic PP breakout board ( $15) plus a computer to run LinuxCNC is another? I'm not fully clear on the system architecture - I need visual representation as the written words are failing me (I blame the long work day but I probably wouldn't get it any day).
The second one seems like the way to go for a bare bones project.
Yes those are two paths to the same destination.
 
I used that board for a 3d printer, it was good for that because it supported all the bells and whistles of the TMC drivers in the early days without additional wires.
Are TMCs a thing with bigger drivers? Like are there skookum high powered silky smooth cnc drivers where you can set current draw in software and such?
With servos and their drivers yes but not with run-of-the-mill stepper drivers.
 
As to why you are getting inconsistent results - are you referring to cutting the thread or to the signal from the magnetic pickup? For the magnetic pickup:
1 - the signal from it is too weak to be reliably read ( going through a cheap break out board with maybe its own isolator is killing the signal?), this would present as being less than the expected number of pulses from it. Could also be a Mach3 limitation on its read rate?
2 - or its too "noisy" and you need to add debounce to it.

Maybe do a test where with a tach you measure the spindle speed, and figure out ho many pulses should be seen over a sufficiently long period of time - say 1 minute to make math easy. 1 pulse per rev. I'm pretty weak electronics wise but this type of trouble shooting I've had to do before.
 
Can you not use one of the opto-isolated inputs as an encoder input? I'm guessing that is what is being done with Mach3 and the board Luc has now.
 
Back
Top