Digital Lathe Tachometer Project

I'm sure some of you won't care for this but it's been a fun project. The tachometer on my lathe jumps around quite a bit. Thought I'd come up with something a little more accurate.

YouTube

More info would be nice,
you need to make a much longer video with some how, whys, the workings and parts involved.
With this computer power out there could this be expanded to be a sort of encoder for thread engagement on a manual? A blinking light for when to throw the lever?
Maybe even a countdown to the flip it moment?
Bob

I didn't really think there would be much interest in it but decided to post that quick video since it was much easier than explaining what it looked like and how it functioned. I'll take another video when I'm back in front of it.

Functionally, the display in the middle shows the RPM in hundreds of RPM anytime the RPM is greater than 100. When it's less than 100 it shows the actual RPM and the "x100" disappears. The ring of lights around the display indicates how close you are to the next multiple of 100 RPM's. If the LED at the 6:00 position was lit then you'd be at 50, 3:00 would be 25, 9:00 75, etc. So if the display said 2 and the 6:00 LED was lit it would be 250 RPM.

The ring of LED's also denote acceleration. You'll notice as I speed it up and slow it down there is a "tail" of LED's lit behind the LED. The size of that tail indicates how quickly the spindle is accelerating.

Where this differs from any other similar project I've seen is that it implements multiple tachometers. Even though the video didn't show it (since I'm just bench testing at the moment) you'll notice 3 boxes below the main spindle RPM. These boxes will show motor RPM, reeves drive RPM, and feed per revolution based on feed rod RPM and spindle RPM.

It probably could be expanded as you asked by using an encoder on the spindle as opposed to a simple hall or IR sensor to get its data. There are many projects floating around on the Internet where people have tried to do just that. I've seen one successfully done using an arduino and one with a development board from TI. Neither of those have posted parts and code however.

Here's the videos from the guy that used the TI board...

YouTube

Here's the guy that did it with an arduino...

YouTube

The tachometer is one of those projects that can easily fall into the trap of "just one more thing." It can get to the point that you almost should have just converted to CNC.

So yes, most anything is possible even with something as low powered as an arduino.

For implementing a tachometer there are two primary methods I've seen used. The first uses the concept of frequency and the second uses the concept of period.

For frequency, the code measures the number of pulses in a given time. For period, the code measures the time between pulses. You can calculate the RPM from either since one is just the inverse of the other, but as a practial matter frequency is to "coarse." If you'd like further explanation why, let me know.

So when you ask what's capable, the simplest way to answer is to say that what I'm measuring (with the sensors I'm using) is the time it takes for any given rotating part to make one revolution. From there it's all math.

To specifically answer your question about a blinking light or a countdown to throw the lever, yes, I think that could be implemented using only the sensors I'm currently using.

Eidt: Oops! I didn't see your second post. Sorry.
Second edit: I meant I din't see your 3rd post!

I've seen one successfully done using an arduino and one with a development board from TI. Neither of those have posted parts and code however.

Click to expand...

Have you seen the project below? It is the most detailed Electronic Leadscrew project I have seen on the internet. I think he has close to a dozen videos on the project. He has released the code on Github and I believe he may have a list of the hardware parts too.

I know it is not the main point of your own project but if you are talking about adding an encoder to the spindle to synchronize whatever, then you should definitely take a look there.


Jacques

Jacques, yes, that's the one I posted above. Didn't know he had posted the code. Little beyond the average tinkerer. Very slick implementation though.

That TI board he uses definitely has some interesting features. It's just beyond my current capabilities. I like using the arduino because it's relatively simple. Of course that also means relatively limited.

Hi Travis,

By the way, he posted a new video last night or this morning about the manufacturing of the boards for the project. I have followed that project from afar but I plan to come back and study the whole thing in detail once he releases everything, all hardware and source code.

I may or may not decide to implement it, but it seems to me like this is one of the good ones!

Good luck with your project.

Jacques

Agreed, best I've seen.

Thanks. Will post another video when I have it hooked up to the lathe.

Arduinos can be usefull, but have you consided switching to 32bit ARM microcontrollers? I made the switch a while ago and can't imagine going back. STM32 processors are available in similar demo boards to the arduino if you are weary of soldering 0.5mm lead spacing quad flat packs. Eclipse is free & supported by STM and the debugger is built-in, something the AVRs don't have. They are a bit more complicated than AVRs but if you are used to reading data sheets and setting/clearing register bits, it's not too high a hurdle. I have been working with STM32F7s for a while now and the speed (216Mhz), peripherals, DMA and FPU are a wonder compared to AVRs.

This is interesting to me. I don’t know enough about instrumentation to build my own, but I would love a readily adaptable tach kit that could be installed on lathes and mills. There are multiple postings on the web about building your own, but no simple kit that I have ever found.

My machines are 3 phase running off of VFDs and the original manual speed controls. Now I calculate in my head for approximate rpm and comvert to sfm as needed. That's a bother.

henrya said:

This is interesting to me. I don’t know enough about instrumentation to build my own, but I would love a readily adaptable tach kit that could be installed on lathes and mills. There are multiple postings on the web about building your own, but no simple kit that I have ever found.

My machines are 3 phase running off of VFDs and the original manual speed controls. Now I calculate in my head for approximate rpm and comvert to sfm as needed. That's a bother.

Click to expand...

I assume you mean the tachometer is interesting and would like a tach kit? Not the electronic lead screw/electronic threading device?

Putting a tachometer together that you can adapt to a machine is pretty darn simple. I can tell you what parts to buy and give you the code to make it run, and tell you how to wire it up. To me the hardest part is just finding or building some kind of housing to put it in. 3D printer makes that pretty easy though.

Keep in mind a simple tachometer can't tell you SFPM since it doesn't know the diameter of the part.

You can get a digital tachometer kit on amazon for $15-20 that will easily read manual lathe/mill speeds. The rub is they require a little magnet mounted for the pickup. On my lathe it was easy though, file a flat spot on one of the spindle nuts and stick that magnet on. No need for adhesive or anything up to 1800 rpm. A project box and wall wart power supply round out the list of needed parts.

JCByrd24 said:

You can get a digital tachometer kit on amazon for $15-20 that will easily read manual lathe/mill speeds. The rub is they require a little magnet mounted for the pickup. On my lathe it was easy though, file a flat spot on one of the spindle nuts and stick that magnet on. No need for adhesive or anything up to 1800 rpm. A project box and wall wart power supply round out the list of needed parts.

Click to expand...

Here's just such an animal...

https://www.amazon.com/DIGITEN-Digital-Tachometer-Proximity-Switch/dp/B00VKAT8A2/ref=sr_1_3?keywords=tachometer+kit&qid=1569530726&sr=8-3

At least on a non-VFD driven lathe, there are a distinct number of speeds so the need to have a tach is minimal.

Any RPM over 100 is shown in 100 RPM increments? Why bother having a tach, then? There's a massive jump from 101RPM to 202RPM...it'd be nice to know just where you lie therein. Then...if you have a tach that's too sensitive it's also near-useless as all you get is a jumble of numbers constantly.

GregSY said:

At least on a non-VFD driven lathe, there are a distinct number of speeds so the need to have a tach is minimal.

Any RPM over 100 is shown in 100 RPM increments? Why bother having a tach, then? There's a massive jump from 101RPM to 202RPM...it'd be nice to know just where you lie therein. Then...if you have a tach that's too sensitive it's also near-useless as all you get is a jumble of numbers constantly.

Click to expand...

On my lathe, the speed is infinitely variable, there are not a distinct number of speeds and there is no VFD. It has a reeves drive.

‘Any RPM over 100 is shown in 100 RPM increments ON THE DISPLAY. That’s the reason for the LED ring. That ring indicates RPM between the 100’s down to a resolution of 4.16 (since there are 24 LED’s.). Take a look at the video and this should be clear. I can easily tell if I’m at 100 if only the 12:00 LED at the top is lit, 104.16 if the 2nd LED from the top is lit, etc. If the display says “1” and the LED at the 3:00 position is lit then I know the RPM is 125. If the LED at the 6:00 position is lit it’s 150. 9:00 is 175, etc. Make sense?

It’s not a matter of too sensitive, it’s a matter of the update frequency. And you’re right, if the display is updated too frequently you’d get a jumble of numbers at least in the 1’s and 10’s digits of the RPM. Hence the reason I use the ring to represent the 1’s and 10’s digits. It gives you a visual representation of the RPM between the 100’s.

In addition to that the RPM is dynamically smoothed. I don’t remember the exact values I used but basically at less than 100 RPM it displays the real time value (calculated from the only the very last pulse width measured) both on the display and the ring. From 100 to 200 RPM it averages and displays the last two readings. From 200 to 400 it averages and displays the last 3 readings. From 400 to 800 it averages and displays the last 4 readings, etc. Those aren’t the exact numbers but it works similarly. At 4000 RPM it averages and displays the last 10 RPM values/pulse durations read.

Where this differs from other projects I’ve seen for smoothing is this. Remember that all we’re doing at the heart of it is measuring the the time it takes to make one revolution and calculating RPM from that. Most others that smooth dynamically like I’m doing will actually wait for x revolutions (depending on the RPM/time between pulses) then calculate the smoothed RPM based on the average and then display it. So for a smoothing factor of 10 the display is updated only once every 10 revolutions. What I’m doing is storing the last 10 pulse durations and doing the averaging at each pulse. So for mine the display is still updated at every resolution no matter how many values are being averaged. But it is using and average and thus the values displayed are smoothed.

Once I get this off the bench and onto the lathe I’ll do another video and try to make this more clear.


Sent from my iPad using Tapatalk

Nice project. For my use i would prefer a simple setup. The tach on my Monarch EE is a perfect example of utility and function.
When running any machine tool , you need to concentrate on the work.
Flashing and blinking lights are not helpful at keeping your focus on the work, the tool and how it is working....
Having a tach that displays the speed of the spindle directly with perhaps a buffer to keep the display form flashing between numbers would be how
i would want it...Direct reading over its entire reading range...no multiplication factors, just the needed info.
The count up and down LED's are cute , but for my money they are a pure distraction and really have little value other than in the fun of making the device to start with.
That has value for sure, just not sure it has a place on a running machine tool...

I fitted a digital tach to a knee mill that was converted to a VFD controlled motor and it works well.. it shows the spindle speed directly,which is all i need.

Cheers Ross

Ross, I’m not familiar with the tach on the EE but the tach on my lathe isn’t very good. At best I’d say can estimate speed to +-100 RPM.
My thought is that you set the speed before you begin working and the light would give a quick indication of where you are between hundreds of RPM.

But you may be right. Might be an unnecessary distraction. Will find out when I get it hooked up and use it. May end up changing the design. Might be that a plain old 7 segment direct reading display is better.

On bench testing the smoothing seems to work well.


Sent from my iPhone using Tapatalk

You can also get tachs that use a reflective spot on the rotating part. They supply some reflective tape with them. Or you could use a spot of white or silver paint. Or a black spot on a bright shaft. These are widely available on the internet.

I purchased one of these and it works pretty well.

https://www.amazon.com/dp/B001N4QY6...Non Contact Tach RPM Meter&sb-ci-a=B001N4QY66

JCByrd24 said:

You can get a digital tachometer kit on amazon for $15-20 that will easily read manual lathe/mill speeds. The rub is they require a little magnet mounted for the pickup. On my lathe it was easy though, file a flat spot on one of the spindle nuts and stick that magnet on. No need for adhesive or anything up to 1800 rpm. A project box and wall wart power supply round out the list of needed parts.

Click to expand...

I said I'd post another video once I had this thing actually running on the machine instead of the test bench. So here it is. I think it clarifies some of the earlier questions and confusion.

YouTube