MEW 283 Electronic Leadscrew link

The link in a letter on page 18 of MEW is a bit ambiguous. It's actually a capital 'I' at the end not a lower case 'l',

www.youtube.com/watch?v=7QaQrqn4yeI

Neil

Thanks, Neil

The use of ambiguous fonts [in many publications, not just MEW] is irritating in the extreme.

There are plenty of legible ones to choose from.

MichaelG.

Most sans-serif fonts are poor at distinguishing between distinguishing between I and l.

Neil

Luckily the digital viewer does not have a problem with the font and you can just click the url in the viewer and be taken straight to Youtube.

[responding to Neil]

... and some of the worst add the numeral 1 to that list !

Edited By Michael Gilligan on 11/07/2019 13:11:43

By strange coincidence my Rotary Encoder arrived from China while I was watching the video.

I'm messing with this via Joe Noci's Electronic Lead Screw, and Duncan Webster who is looking at building one. My interest is more in programming microcontrollers than making an ELS but ELS has some nice challenges. Joe in another thread got me investigating ST Nucleo boards rather than the usual Arduinos I mess with.

For £13 the Nucleo F446RE board is considerably more powerful than an Arduino, in particular it should have no trouble counting 4000 positions per revolution as shown in the ELS video. (Be careful buying encoders - I accidentally ordered a 1000 positions per revolution encoder due to not reading the small print...)

Arduino fans can program Nucleo boards from the Arduino IDE and treat them as a kind of super-Arduino. Compatibility is good, even though it uses 3.3V logic, so not all modules will work directly. Also not all libraries are compatible, though quite a few are. Relatively painless conversion for anyone already familiar with writing code for Arduino.

Not all Nucleo features are available from the Arduino environment, for example it comes with a built-in Real Time Operating System supporting threads and other advanced features that Arduino don't understand (yet). At the cost of learning a new interface these can be accessed using an online IDE called mbed. If you are already into C++, likely you might find mbed cleaner to code than Arduino, but not so well supported in terms of hobby libraries.

I found compiling mbed online to be rather restricting, but it's fairly straightforward to download the environment and use it locally either with an IDE like Eclipse (several other alternatives available) or as a raw editor / make project. The latter is particularly suited to experienced programmers. As a Linux fanboy I've not tried mbed Studio because it's only available for Windows and Mac. Looks good though and it's nice they don't force you to use it.

The Rotary Converter comes with a warning label I've not seen before:

Dave

Edited By SillyOldDuffer on 11/07/2019 14:10:40

If you print a correction in the mag can you add a warning that it is very very very boring and mostly a view of the left hand end of a lathe. Well I only skipped through the first 15 minutes but it should have been condensed to 15 seconds of actual content.

Can you modify your code so that it reads every edge of the quadrature pulses? Then the 1000 ppr magically transforms into 4000 ppr.

Yep, Duncan is building one, and progressing nicely! But lets not steal his thunder...

From SillyOldDuffer

(Be careful buying encoders - I accidentally ordered a 1000 positions per revolution encoder due to not reading the small print...)

And from DC31k (???)

Can you modify your code so that it reads every edge of the quadrature pulses? Then the 1000 ppr magically transforms into 4000 ppr.

Exactly how it works in my ELS implementation - Actually, I used a 1024PPR encoder for the spindle, and generate an interrupt on the rising AND falling edge of each of A and B channel, giving 1024 X 4 counts. Important for the spindle encoder though, you must use an encoder with a Z channel ouput, ie, an index pulse per revolution. For the spindle encoder in my ELS code, within limits, the higher the PPR the better. Less than 4000 pulses ( 1000PPR encoder) is not recommended..

Joe

Hello Dave,

I'm interested in building an ELS for my EMCO Compact 8 lathe, so I started looking for an incremental rotary encoder that can be used in the project.
As I just know - from what you wrote in your post - that the rpm must be at least 4000 - I searched and found a rotary encoder **HERE**
Could you please tell me if, in your opinion, his specifications are compatible with my objective?

Grateful
Dias Costa

Hi Dias.

Your unit is the same type as mine. It's a 'B E6B2-CW26C', which defines it's shape and size. They're good for 6000rpm, so OK on a lathe.

They come in different resolutions though, ie the customer can choose the resolution (pulses per turn) he wants.

The resolution is tucked away in the other number:

Your ebay photo above says: B38S6G2-F100B526. I have a B38S6G2-C1000B526

The code is:

B = Product Type

38 = Maximum Diameter in mm

S = Solid shaft

6 = diameter of shaft in mm

G = ?

2 = Cable length in metres

F = Output is Complimentary Mode

100 = Pulse Count per Rotation !!! Too low !!!

B = Output Phase ( 2 signals)

526 = Power input Minimum 5Vdc, Maximum 26Vdc

Two things make your choice unsuitable: 100 pulses per revolution is too low, and its complimentary output makes the interfacing the unit to the microcontroller a little inconvenient.

My B38S6G2-C1000B526 has a collector low output and outputs 1000 pulses per revolution, which is OK. Ideally I'd have ordered a 'B38S6G2-C4000B526'. However Joe's comment above very helpfully points out I can write code to read the edges of the quadrature signal, thus getting 4000 pulses per revolution. I hadn't thought of that but it's easy to do!

This is where I got mine - but I see the 4000 position version is no longer available.

Dave

Just in case Joe misses this thread, the encoder you need for the spindle has to have A, B & Z outputs, the Z is one pulse per rev. The one I've fitted is B E6B2-CWZ6C 1000P/R

Indeed, and back in the win95/98 days there was a trojan virus which installed itself as rundIl.exe, which looked exactly like rundll.exe and so was hard to spot.