When machining metal, it is important to know how fast the cutting tool is traveling in relation to the surface of the part being machined. This amount is called the ‘Surface Speed’. There are Surface Speed standards for cutting different types of materials and it is good practice to stick with those standards in order to end up with a good surface finish as well as maximizing tool life. On a lathe, for example, having a known target Surface Speed in mind as well as a part finish diameter, it is possible to calculate the necessary spindle speed.
Hobbyist [Paul] wanted a method of measuring his lathe’s spindle speed. Since spindle speed is measured in RPM, it made complete sense to install a tachometer. After browsing eBay for a bit he found one for about $20. His purchase came with the numeric LED display, a mounting bezel and the all important hall effect sensor. The Hall effect sensor measures changes in a magnetic field and in turn varies its output voltage. [Paul] fabbed up an aluminum bracket that supports the sensor just off of the rear of the lathe spindle. A magnet was then glued to the outside diameter of the spindle below the sensor. The once per revolution signal is generated every time the magnet passes the sensor while the lathe is running. The display was mounted to the lathe near eye height by means of another aluminum bracket and case.
After a little work, [Paul] can now keep a close eye on his spindle speed with a quick glance over at his new tachometer display while he’s turning those perfect parts! If this project tickles your fancy, you may want to check out this fantastic DIY tachometer or this one that uses a soundcard.
When you started talking about surface speed I though that it was at least going to take the distance of the tool from the centre line and calculate surface speed – perhaps even automatically vary the spindle speed to keep a constant surface speed.
A bit disappointing (if not surprising) that buying something and using it for its intended purpose is considered HaD worthy these days.
Yeah really… I guess knowing your motor base RPM and working back though the belt or gear drive ratio is just too complicated, or if using a vfd, knowing that 30hz is half as fast as 60hz base speed, and calculating from there.
I too was disappointed that the diameter of the work piece wasn’t part of this system, because ultimately where the metal meets the appropriate cutting tool is where it’s important, not at the shaft.
I was equally disappointed when it was revealed the measurement was simple spindle speed… That’s like saying it’s important to know how many miles per gallon you are getting so you don’t run out of fuel, and then incorporating a speedometer to measure how fast you’re going.
Would have been more interesting if some method was provided to adjust the readout for the varying radius, and much more interesting if the spindle speed was compensated as a result.
Even using an Arduino to read the input and drive the display would have been an improvement ;-)
Almost all machinists don’t care. If it chatters your speed is wrong.
The CNC guys don’t care either, the software sorts it out for them.
People who don’t make much (usually beginners) on their lathe care about such things. And nerds, because well why not. LEDs are good, more (even if useless) are even gooder.
?! I would agree with “LEDs are good”.
It would be cool to tie the tach to the DRO’s X-readout to get an instantaneous surface speed. Even though, umm, ‘nobody cares’…
It’s true, no-one cares. Really.
Cutting speeds are so variable anything might work, and if it’s wrong you’ll soon know about it.
The ‘feeds & speeds’ stuff you see are aimed at production work and tend to be on the high side to maximise throughput. You could halve the ‘recommended’ rates and have it work just fine.
I have to agree, speed is done by feel. I’m not too concerned about tool wear because all it takes is a quick hit with a stone to sharpen some HSS back up and ultimately that takes less time than machining at a slower speed. Maybe that attitude will change someday if I start using carbide inserts, but I seem to recall those had excellent wear properties when I was in school.
Reminds me of competing in VICA, they asked us to mind our speeds as they had a limited number of inserts and a lot of competitors to run through that day. I may have turned some heads when I just randomly cranked the speed and started cutting. Still, I got 100% on the lathe and it only took me 30-40 minutes out of the 60 allotted. Mission accomplished!
For RPM, I look at the setting on the gear change. I guess those numbers could be off but never bothered to check. 70 is slow, there’s a few steps to 360 which is about the fastest I use. And then there are stupidly fast speeds after that.
I mostly use tooling with inserts. The sound of my lathe changes with loading from the cut… I crank up the force until the lathe starts to moan then I back off. Speed? I vary it to manage the chip or minimize the smoke from the manually applied coolant.
Yeah, most lathes are the gear change type. I’ve got a small 7×10 at home that has speed controlled with a potentiometer and the lathes at VICA were controlled the same way, but are the only time I’ve seen lathes that large with such a control.
Another application where I think this would be handy is an old Monarch where they had a mechanical tach and a variable speed drive.
Agreed. I work as a machinist, manual and CNC. You rarely worry that much about the speed- especially on belted equipment, which is much manual equipment. The pulleys keep the speeds you can do at set speeds- you match your feed by hand to get it to cut well without chatter. Same with CNC- except no belts. I can hear from across a room if something is cutting well, even if I have never seen the cutter or material, just by sound.
Things like the Machinery’s Handbook give “best case” numbers for test cutting a specific steel as a reference, called 12L14, a heavily leaded free machining steel. Every number is a reference on tooling from cutting that. You are supposed to realize that, and adjust accordingly depending on the type of tooling and geometry your tooling uses. I used to think 8 or so years ago when I was in school learning that everything had to match the book really close to cut well. It all seemed like unless every tool angle I ground dead on, every speed and feed had to perfectly match or things wouldn’t cut. After a while, I realized I was very wrong.
Once you actually do enough manual machining, you learn really quick most materials have a lot of leeway- speed alone isn’t that important- speed is matched to an appropriate feed by feel. That will give you a background of general idea where to machine materials when you go to CNC. The only time numbers are more critical is in tapping, reaming, and trying to achieve really fine surface finishes.
I was hoping this was a constant surface speed adjuster for a manual lathe- something I’d really like to see and could really use- tons of us would! Dissapointed :(
I guess I would put myself in “cares about it” category.
Spindle speed and surface speeds are critical to know in the dynamics of tool cutting. IF you’re doing high speed machining or anything with thin walls it is critical to know your surface speeds. The CAM software gives you a decent starting point, but if you’re doing high speed stuff you’re going to need to do a tap test to get the real dynamics and harmonics of your tool combination.
But for most home and hobby machinists if you’re cutting steel, if the chip turns blue it’s in the ballpark.
A similar sensor could be mounted to the tool-holder advance wheel, zeroed to the spindle, and a quick calculation made to display the actual surface speed at a given diameter (roughly, depending on bit height relative to center and angle of travel from perpendicular). Its resolution would depend on the number of counters (magnets) installed on the advance wheel.
Throw in a selection for material type and this is a viable tool to use on a manual lathe. Will save a few cutting tools and give some more life to them.
A roller sprung against the spinning part can have it’s RPM counted and will give a relevant surface speed.
Better yet calibrate it to actual speed. A good use for a ‘duino. Use a mouse wheel opto.
Experience with motors and other rotating things has the visual vibe good enough for me.
Unfortunately that doesn’t account for lashings of cutting oil and/or square parts. Some kind of visual measurement of the diameter would probably be best
If the spindle speed is not what you set it to be something is very wrong with your lathe (and you would know it too).
I don’t see this as being useful in any way…
Was wondering about that, too. Are there lathes that you just turn on and pray? Maybe he’s got one that’s lost all of its documentation and signage.
His is not. The forum post of his project (one of the links) has pictures of the lathe. It’s a geared unit that has a big metal RPM chart below the gear selector. This seems to be a popular mod for users of that forum.
Umm, use the super simple formula:
Rpm=(cutting speed x 4)/dia of part at cutter contact
Rearrange formula as req’d..
Hss cutting tool on mild steel: 100fps cutting speed
samr on al: 400fps
Double to quadruple cutting speeds for carbide cutters…
Haven’t been in a shop for 20 years, but this is part of my brain.
Except for Paul, there’s not a machinist among you
…and the crux of what he’s saying is it really doesn’t matter much. If it did, you wouldn’t casually toss out “double to quadruple for carbide” – that’s a big range y’know.
Watch the old guys facing off something big. Surface speed will be high at the edge, and zero in the middle, so as you move towards the center you should up the RPMs to compensate, right? No one does.
You do a few rough cuts at low speed (a few hundred RPM), then you crank the speed way up (say 1500) for the light finishing cut.
Even CNC users do that as many don’t have a variable speed spindle, especially in the hobby or low end production stuff.
My lathe has a digital RPM readout. It broke ages ago (loose wire probably) and it’s not worth fixing it.
I’ve worked in a CNC machine shop (I was QC) and speed isn’t just kinda important. In Aircraft parts manufacturing they get very picky on surface finish, which is the crux of the article, but agree that most CAD/CNC program suites will do the calculations for desired finish.
However, depending on what ancient equipment you’re working on, even if the gearbox legend says it, mistakes can be made and belts can slip. A tach is cheap insurance. Some of the older equipment I’ve seen still used in production used CVT-type belt transmissions, and most of them had tachs because you couldn’t always trust the scale on the control handle.
I came here because I thought there was going to be a stepper attached to the leadscrew. So I don’t have to adjust those pesky change gears to cut different threads.
Caveat: Have not used a metal lathe for a while. Can’ t remember the brands, but these lathes were dark green and blue in color]
Last time I had access to metal lathes, the shop had one monster [faceplate could hold a 26″ rim with lots of room to spare.] the size of a truck that the seniors got to play with. Us “unders” were on bench lathes, every one of them had big dial tachs mounted on the headstocks. Not for tracking the surface speed – which was pretty much set and forget [exactly as Paul describes] on the cheesy gear-change lathes we had at school – but for making sure you didn’t do anything stupid. And trust me, especially in an educational setting, people with lathes can do a lot of stupid. Flying chucks, tool posts, and parts.
I read a lengthy article once about using a neural net on a milling machine that was fitted with several sensors including a couple microphones. The aim was to optimize for speed and minimize tool wear. IIRC the results were quite a bit less than what a trained machinist could do “by ear”. Negative results are very useful too. It’s too bad they don’t get the glory that positive results do.
ITT we assume our experience is universal to all industries, hobbies, and specializations. If speed does/doesn’t matter for your purposes, then the same must be true for everyone.