Gallery, Projects and General > Project Logs
Electronic Leadscrew for the New Lathe
vtsteam:
I can open all of the files. Which files can't you open, BB, and I'll tell you what they are, and how to open them. (The ODG file is an Open Office/Libre Office spreadsheet). Most others are code files, and stock images of the display and ATMega board.
The connections are between existing Arduino board, some switches, encoders, and stepper driver, so it's just given as a pin list referencing the Arduino's numbered pins. There is no schematic or circuit board info. You'd have to hard wire it. Maybe write to him -- he seems to speak English on the video.
Here is the Pin List:
--- Code: ---Pins - Funktion RJ45 (Encoder A,A' / B, B' / Z,Z')
8- GND BR
7- 5V BR/W
6- B+ GR
5- Z- BL/w
4- Z+ BL
3- B- GR/W
2- A- OR
1- A+ OR/W
---------------
Pins - Funktion RJ45 (Encoder A/B/Z und Eingänge (Endschalter) )
8- GND
7- 5V
6- B+
5- Arduino 41
4- Z+
3- Arduino 39
2- Arduino 38
1- A+
--------------
Pins - Funktion RJ45 (Endstufe)
8- GND
7- Arduino 11 (Status Eingang)
6- 5V
5- Arduino 8 Dir Pin
4- 5V
3- Arduino 9 Step Pin
2- GND
1- Arduino 10 Enable
----------------
S1- Arduino 42
S2- Arduino 43
S3- Arduino 44
S4- Arduino 45
S5- Arduino 46
S6- Arduino 47
S7- Arduino 48
S8- Arduino 49
S9- Arduino 50
S10- Adruino 51
S11- Arduino 40 (Encoder Taster)
S12 - Reset
Encoder A- Arduino 19 Int4
Encoder B- Arduino 20 Int3
Externer Encoder A- Arduino 2 Int0
Externer Encoder B- Arduino 3 Int1
Endschalter - 38
Endschalter - 39
Endschalter - 41
LED Status 1- Arduino 53
LED Status 2- Arduino 52
LED Status 3- Arduino 30
LED Status 4- Arduino 31
Summer- Arduino 32
--- End code ---
Edit:
On his website there is this in the Downloads section:
--- Quote ---ZyklenAutomatik b2.02 comming soon
--- End quote ---
In his video he shows an older version (probably the pin list above) and a newer version of his board. The newer version is probably what he's working on.
vtsteam:
That project is for automatic thread cutting, I believe, and so it's kind of a subset of a CNC program, requiring a couple axes, and acceleration deceleration, a display and lots of button functions.
What I'm thinking of doing is MUCH simpler (or more primitive, depending on your point of view) --- simply replacing the change gears.
Cutting threads exactly as you would on a manual lathe, otherwise. No second axis, no acelertion deceleration factors, etc. You engage the half nut with the spindle and leadscrew already turning, just as you would with change gears. You feed the slide in as multiple passes, in the manual way.
PekkaNF:
I like the direction you are heading. Keep it clean and simple.
Very long time ago I used EPROM as a look uptable - without using a uP. Really simple. Incremental encoder to address lines (all extras pulled), data lines were output. Eprom was mostly wasted but small partion of was this lookup table. You need reset and clock, but they could be generated from index and quadrature count pulses?
Few posts back:
--- Quote from: vtsteam on March 31, 2015, 09:45:46 AM ---.....Ideally a rotary switch with a big metal knob would be a great "old school" solution to fit a traditional looking lathe, I think. You'd have to make one with a PC board and wipers, but it could be done as a later project in itself!
Thank you all for your replies! :beer: :beer: :beer: :beer: :beer:
--- End quote ---
Are you thinking of binary code rotary switch? Make the disc/code GRAY-code. You probably know this, but unsuspected person could find very unstable readings with ordinary binary code. Reason being that two (nearly) simultaneous bit changes are often required to happen. In real world on transition you get whatever code.
http://en.wikipedia.org/wiki/Gray_code
In this application software filtering is possible to see that the input is stable.
CNC Hand Wheel MPG uses similar type encoder than you are planning to use on spindle. It is handy feeding in all sort of numbers. I had camera that had that sort of handwheel, only circumfere sticking out and number or menu item was scrolled by rotating that hand wheel on thumb and then the hand wheel was pressed to lock/choose value. This will ofcourse require rather coarse resolution and definate detent, but for most of the people this is desirable on menu selection. On feed you want fine resolution.
Bit of OT sometimes goes far. :lol:
Pekka
RussellT:
I'm not familiar with commercial rotary encoders so I've had a quick look on ebay. It seems that for not much more you can get 2000 pulses/rev. Presumably whichever encoder you choose for your purpose you could double the number of pulses by using two inputs and counting both sets of pulses - or possibly all three, A B and Z. I'm not sure whether they've already done that to claim 2000p/r.
Here's an example http://www.ia.omron.com/product/item/e6b27090f/index.html
Looking at the output diagram you might be able to count 5000 p/r which ought to improve accuracy. I've tried playing with your spreadsheet but adding to the encoder disc sectors doesn't affect the accuracy. I must be missing something there I think doubling the pulse output ought to halve the error.
You can program PICs in basic. I use mikrobasic which has a free version.
Russell
PekkaNF:
--- Quote from: RussellT on April 01, 2015, 05:12:20 AM ---I'm not familiar with commercial rotary encoders so I've had a quick look on ebay. It seems that for not much more you can get 2000 pulses/rev. Presumably whichever encoder you choose for your purpose you could double the number of pulses by using two inputs and counting both sets of pulses - or possibly all three, A B and Z. I'm not sure whether they've already done that to claim 2000p/r.....
Russell
--- End quote ---
I used work some time with rotary encoders (incremental and absolute) and my experience is that at some point in the system more P/R does not produce any gain. Specially when you start counting leadin/trailin edeges as a pulses you are stepping on thin ice. A little interface mismatch, threshold, not that crisp edges, not to mention little mechanical vibration when pulse is setting between 0/1 states and sometimes the count drifts even when nothing moves. Same goes to single channel encoders. They can't detect directional change. Some might think that in this case it does not matter. But it does. Even when you think that you rotate spindle on one direction only, but funny stuff happens when you start/stop at pulse transition.
I tried to shed some light on this issue few posts back.
Pekka
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version