Electronic Leadscrew for the Myford ML7
I have been interested in converting the ML7 to an electronic leadscrew ever since I started watching the YouTube videos by Clough42 where he describes the design, developement and construction of the ELS for his own Grizzly G0602lathe. I hate messing around with changewheels!
So what is an electronic leadscrew? I would recomend watching James's videos to learn all about it but basically, instead of driving the leadscrew using gears the leadscrew is driven by a stepper motor or servo motor. The lathe spindle drives a rotary encoder which monitors the rotation of the spindle and outputs pulses as it does so. These pulses are fed into a 'computer' which then outputs signals to drive the stepper motor/servo motor. The ratio between the rotation of the stepper motor/servo and the rotation of the lathe spindle is determined by the software installed on the 'computer' and the required relationship is entered into the 'computer' by a simple keypad and display console. The console displays the threads per inch chosen when using the lathe for screwcutting and also the feed per revolution of the spindle when using the ELS for fine feeds when turning.
One thing that is useful is that the lefthand part of the display reads the rpm of the spindle. This reads all the time, even when the ELS part of the electronics is turned off. I was intending to fit a rpm meter to the lathe at some point but now I won't need to.
The 'computer' used is a standard microprocessor board available from Texas Instruents. The display board is available as a kit from James who has an Ebay shop in the USA. All the other components are off the shelf items available everywhere but I ordered most of them direct from China as that is a much cheaper option and the delivery time was not too bad.
I chose to go with a servo motor rather than a basic stepper motor as the servo motor will compensate for any lost steps that might occur with a stepper. That's not too important for a mill table feed or a lathe saddle feed but not good for screwcutting as the thread will be damaged or deformed if the leadscrew gets out of sync with the spindle.
Here's a photo of all the major components:
Top left is the rotary encoder for the lathe spindle, middle top is the servo motor, then a 5 volt 'Wall Wart' to power the electronics. I had a few of these lying around so didn't need to buy one.
Under the encoder is the fascia panel for the control box and in the middle is the microprocessor board with the electronics for the control panel underneath. The box on the right is a standard 48 volt power supply for the servo motor.
All it needs is some mounting hardware for the encoder and the servo (when I've decided how to fit them!), wiring and a box to put it all in.
The total cost so far is about £260, the most expensive item being the servo motor. The servo motor does have the driver electronics built in however so no extra driver unit is required unlike with a stepper motor.
For anyone interested, here is the link to the Github page that James has made which will help if you decide to follow a similar route. You can also download the software for the processor board from here:
I don't know how much time I will be spending over this project as I would like to get the Green Arrow completed and the ELS is not urgent. However, I did assemble the control panel and the Boost Board.
The control panel PCB requires the 5 pin connector changing to a right angled job to make it easier to attach the connector cable but that is no problem. A new connector is included in the kit. Then it's a case of bolting the board to the front panel and adding the 3D printed buttons.
The display has a red filter over it to make it stand out better when lit. The panel is meant to fit into a diecast aluminium box mounted within easy reach on the lathe but can be mounted anywhere convenient. It does recommend to keep the connecting cable as short as possible though to reduce noise pick up.
The Boost Board needs all the connectors soldering on but again no problem.
I think I will wire it all up on the bench first to check that everything works but I need to order some connectors before I can do that.
I thought I would have a play with the software and try downloading it onto the microprocessor board. I made up a temporary lead to connect the display board to the microprocessor board and then downloaded the required software to flash the ELS software onto the mcu board. The required software is called Code Composer Studio and is downloaded from the Texas Instruments website. You need to make sure that you have version 9 which was used to create the ELS software. There is now a version 10 but the ELS software will not work with that at the moment. James has an excellent video about how to download all the software and use it on his Youtube channel.
The ELS software has two versions - A 'Debug' version and a 'Release' version. The Debug version runs slowly so that you can hopefully find any problems with it (if you know what you are doing, which I don't!) and the Release version is the faster, trimmed down working version that is loaded permanently to the mcu board. The debug version is not flashed permanently onto the mcu board and is lost if you remove the power from the board.
I managed to get the debug version to work fine but the Release version failed and not knowing anything about code I didn't know why. I did put a question on the Github page to see if anyone could help but I eventually got it to work by putting the software files into a different folder on the computer. I don't know why that solved the problem but it did which is all that matters! I could now flash the ELS software permanently onto the mcu board.
When the board boots up it displays 'Clough42', then the release version of the software before finally going into operational mode which initialises at the fine feed mode.
I also connected up the encoder to check the RPM feature on the display and that worked when I spun the encoder (or at least it did once I had got the connections the right way around!)
The next stage is to connect up the servo motor and get that working. The servo motor that I'm using has different connections to those shown on the wiring diagram on the Github page so I've got to work out how it should be wired up.
The ELS software has a onfiguration file where you can change the settings to suit your particular installation e.g. the tpi of your leadscrew and the gear ratio between the servo motor and the leadscrew but I'll tackle that when I've decided how the servo will drive the leadscrew. I'm not sure where to mount the servo yet. I think it willl be very difficult to mount it neatly at the headstock end and still keep the changewheel cover as it is so I may mount it at the tailstock end with a pulley mounted where the leadscrew dial is at the moment. I'm fitting a DRO to the lathe as well so I won't need the leadscrew dial to actually measure the travel of the saddle but I'm hoping that it will still be possible to turn the leadscrew by hand with the servo connected so I want to keep the handle if possible.
Fun times ahead!