Wimminz – celebrating skank ho's everywhere

September 2, 2017

Goose stepping

Filed under: Wimminz — wimminz @ 2:07 pm

The image on the right is a cheapo stepper controller, bough from amazon with prime so bought and delivered for just under ten quid.

So, with no more ado.

From top to bottom

  • ENA –
  • ENA +
  • DIR –
  • DIR +
  • PUL –
  • PUL +
  • B –
  • B +
  • A –
  • A +
  • GND
  • VCC

I’m going to assume you have some sort of breakout board or arduino or something that is going to generate your direction of rotation and movement pulses, I’m going to assume you have a suitable 12-24 VDC PSU, and I’m going to assume you have one stepper motor, up to around NEMA 23 and up to around 1 NM torque and 2 phase (that means 4 wire) though you can run a 4 phase if you bond the phases together in parallel.

There are some dip switches at the top that set current limiting and microstepping, if in doubt start at a low current like 1 amp and start at 16 microstepping.

Steppers motors can get warm or even very warm, but never hot.

Stepper motors should sound smooth, no vibration, if there is vibration or harmonics you have a setting wrong.

Connect one phase of the stepper across the A terminals and the other phase across the B terminals, just make sure you get the colour coding right, see image on right.

Now one wire from your “pulse” output to PUL +, and one wire from your “direction” output to DIR +, and a common ground for PUL – and DIR – to your ground (GND should do) connect PSU -ve to GND, PSU +ve to VCC, and you’re good to go.

Pretty much all steppers are 1.8 degree which means 200 steps per revolution, at 16 microstepping that means 3,200 microsteps per revolution, if you run a 5 mm pitch ballscrew directly coupled to the stepper then 3,200 / 5 = 640 microsteps per mm of X travel along the ballscrew.

At 640 steps per mm I can get 1,000 mm/min out of a real cheap NEMA 23 4 wire stepper, that’s 16.7 mm/sec

1,000 mm/min x 640 microsteps per mm = 640,000 microsteps per minute.

640,000 microsteps per minute / 3,200 microsteps per revolution gives us 200 rpm at the stepper to get that max 1,000 mm/min out of a 5 mm pitch ballscrew.

A 10 mm pitch ballscrew would give us 2,000 mm/min at the same RPM, but half the precision / accuracy per step.

200 RPM is, being practical, all you are going to get out of a cheapo 4 wire stepper connected to anything, sure, you may get 600 RPM running it free, but when connected you do NOT want to miss steps, or put up with noisy / harmonic motions, I’m talking “sweet spot” and the sweet spot is going to be around 200 RPM for this stuff.

“This stuff” assumes things like 100 kHz signal between the “computer” and the controller above, eg cheap, sure you can spend 900 quid on a PCI-E driver card that will talk to controllers in mHz speeds, but it’s not relevant here.

Get it all right and it sings sweetly, runs smoothly, and steppers never get more than warm to the touch.

I’m using one of these and a 12 VDC PSU and a cheap NEMA 23 stepper on a linear slide with a 5 mm pitch C3 ballscrew, in theory the ballscrew is ground to 30 micron (0.030 mm) accuracy, and in theory my 640 microsteps per mm gives me 1.5 micron accuracy, so in *theory* I can get 32 micron repeatable accuracy out of this and rapid speeds of up to 1,000 mm / min, and I have an 800 mm travel on this so 47 seconds from one end of the slide to the other, for a 400 quid linear rail and a 10 quid controller and a 15 quid PSU, and a PC running Mach3.

32 micron we are talking about the focus dot size of a metal marking fibre galvo laser (which in reality ranges between 15 and 50 micron depending on the final f-theta lens)

So call what I have 450 quid, theoretical 32 micron accuracy with a travel of 800 mm and a full excursion time of 45 seconds or so.

Want 5 micron accuracy over the same distance with the same speeds?

Expect to pay 5x as much.

Wants 5 micron accuracy and 60x those speeds?

Expect to pay 50x as much.

The fact is that today in 2017, and this is the point of this article, control a reasonably good precision and excellent value 400 quid ground 5 mm pitch ballscrew linear motion table from amazon with a 10 quid stepper controller from amazon, and pretty much out of the box get 50 micron or better accuracy, repeatable accuracy.

50 micron is two thousandths of an inch.

older more mature readers will know that 2 thou is good enough for 99.9% of everything, with some specific exceptions, all mechanical diesel injection pump for example.

Being able to do this at the touch of a button on a screen for less than 450 quid…. that’s fucking awesome. Flash Gordon awesome.

Because it means 20 positions per millimetre, and a 800 mm travel means 16,000 positions, I can say “go to X0” and it will, to within 0.050 mm or better, I can say “go to X550.75” and it will, in 32 seconds, to within 0.050 mm or better, I can say “go to X250.55” and it will, and it will do it all day long, with the stepper controller set to 1 amperes and drawing 12 VDC, so PC aside I’m using 12 watts of power to do this.

The linear rail can carry 80 kg, and even at the low “stepper runs luke warm at best” 1 amp 12 volt supply, a 16 mm diameter 5 mm pitch ballscrew will crush your fingers so badly you have to go to hospital for significant surgery.

This is fucking peanuts, money wise, for what you get and what you can do.

The only downside is you MUST either have or be prepared to learn, some basic CNC skills, such as Mach3, and learn about things like steppers and microstepping and pulse durations and so on, but it’s not actually that hard, it’s just complex.

I have never had or used an arduino or pi, I am always tempted but time etc, but I think it is a reasonable assumption that anyone who isn’t daunted by buying an arduino and a couple of shields and knocking up some gcode could do the hardware side fairly easily, once they read something like the Mach3 manual, (103 page pdf) even if they don’t have Mach3, it’s a good foundation for CNC.

  • A cheap but usable and practical stepper controller is now a tenner a pop, so for a 3 axis machine that is 30 quid for 3 controllers, or 40 for 4 if you do a dual X axis system for rigidity.
  • A cheapo 2 phase NEMA 23 (NEMA 23 = 2.3″ hole centres for mounting) 1 Nm stepper is again around 15 quid.
  • Even a cheap ballscrew is better than any other kind of threaded bar, even precision ground acme, if you’re building a disposable 300 quid 3d printer then fair enough, use threaded bar or ground acme, but 450 quid gets you 50 micron or better repeatable accuracy over 800 mm of travel, and 1,000 mm/min, nothing beats that but 5 micron accuracy and 1,000 mm/sec, but it’s nearer 25k for that…. not 450 quid.
  • DO NOT be swayed by the belief that you need some uber speeds, unless you are in production engineering 17 mm/sec is enough for most things, while 50 micron repeatable accuracy is arguably good enough for *anything* the DIYer will want to do, and a decent ballscrew stage should last many many many years.
  • DO NOT be swayed by people and products that *claim* the same “theoretical” accuracy, go 64 microstepping and use a 1:4 reduction and generic cheap 5 mm pitch threaded bar and I could CLAIM 0.1 micron theoretical accuracy, and shure enough Mach3 or whatever would make the move to X117.4565 mm… put a fucking glass scale DRO on the bastard and you’ll be lucky if you are within 500 micron or 0.500 mm of that.
  • DO NOT be swayed by stuff that is not RIGID, it doesn’t matter what you tell the tool to do, if the framework isn’t rigid the tool will chatter and wander and effective material accuracy will fall off a cliff, even though the stepper / linear side of things is retaining it’s claimed accuracy.
  • DO NOT forget basic engineering and math, a 900 mm long 16 mm dia steel ballscrew (for 800 mm of travel) has a coefficient of thermal expansion, and the difference between an ambient winter 5 degrees C and a summer 20 degrees C is enough to exceed the accuracy of the device itself.
  • For linear motion speeds of up to around 300 mm/sec I’d consider hywin rails over pukka ball linear rails, they are much cheaper and just as good down at that end of the scale.. more resistant to damage too…  twin hywin rails in parallel give great rigidity for less money than a single round rail.

Buy this shit while you still can in 2017, most of it is only so good and so cheap because it is made in China, which means you’re at the mercy of exchange rates and geopolitics, and there is nothing there to suggest that 2018 is going to be a whole lot cheaper than 2017, quite the opposite in fact.

If you’re in the UK and want slightly better quality stuff but still reasonable prices, but want to buy locally, I’ve used Motion Control Products for years, and have steppers and controller and stuff from them that is going on 10 years old and still works like new.

In closing, if you’re not sure, and wavering, consider this…

A “used” acme etc linear motion jobby is worthless even if it only has 2 hours hobby use on it… a “used” C3 ballscrew linear motion jobby with a few hobby hours on it retains 95% of its value, so all you’re looking at is a 10 quid stepper controller and a 15 quid stepper motor, if you already have an arduino.

If you have a PC you *MUST* have some sort of breakout board of box to talk to a stepper controller, MCP (listed above) do a parallel port breakout for 25 quid, if you do not have a parallel port you can get a USB2 to parallel converter specially for mach3 for 80 euros, there are other option such a smoothstepper etc and on upwards in price.

An arduino running a shield and gcode is NOT the same thing as a PC running Mach3, but they will talk to the same sort of hardware to produce motion, so it depends on what you want to do with that motion… if you do not want to input various parameters and cut a keyway or mill a pocket and generate the gcode and then control the machine all in one program then you do not need mach3, an arduino can run a stepper well enough to drive a plotter axis for example.

YMMV but IMHO you won’t learn shit worth knowing by starting out taking apart printers and floppy drives for steppers / servos etc, and arduino or a pc and a shorter linear rail with ballscrew could come in less than 200 quid, and becaue it would have some useful precision and repeatability and rigidity and power, you could learn one hell of a lot from playing with it.

 

 

3 Comments

  1. off topic

    Any thoughts on HCCI gasoline engines?

    Comment by guest — September 2, 2017 @ 2:49 pm

    • HCCi is just diesel with a different fuel, but because it is a different fuel it works at lower compression ratios, so it should produce less “harmful” oxides of nitrogen as they are related to combustion pressures and temperatures.

      The other issue with petrol is the propogation speed of the flame front, diesel is not a problem, try the same thing with petrol and you’ll get “knocking” and blow the bearings out of the thing, and there is only so much you can do, traditionally, to slow down the flame propogation front, traditionally an alloy head absorbed more free electrons and slowed it down that way, compared to an iron head.

      As I know *nothing* about the engine in question I’m guessing they have solved the knock issue by some sort of fancy combustion chamber design and some sort of fancy metered direct injection system, then you have a compression ignition petrol (even if you use a spark or glow plug to initiate combustion) engine which due to the higher compression ratio is closer to a diesel for torque, so if we say for example that a 2 litre diesel pulls as well as a 3 litre petrol, now maybe a 2 litre diesel will pull as well as a 2.2 litre hcci petrol.

      Lower compression means lighter build weight / quality / cost.

      but it is STILL a heat engine, and if it is based on poppet valves and reciprocating pistons, that is about as good as it gets, 60% of the energy in the fuel is going to go down the pipe as heat.

      Perhaps a 1300 cc hcci will do in place of a 1950 cc normal petrol engine, so I can see the economic case for it.

      Dowsides are it is going to be a fucking bitch to work on and maintain compared to an OHV spark ignition 2 valve petrol lump.

      Upsides are a 3 litre hcci celica would haul ass

      Connections with electric vehicles, none, it doesn’t change anything, rolling resistance of a tesla isn’t much different than any other similarly sized and weight car, it can’t be, ditto aero resistance.

      The downsides of the electric car is the IC engine has a fuel tank that costs less than 100 bucks to make, and said tank can be charged at a rate measured in megawatt/hours per minute, while electric is stuck at the *****very***** low end of the kilowatt/hours per minute rate….

      The ONLY case for an electric vehicle is someone operating a fleet in london, they beat the congestion charge and they have a site to park and overnight charge, nota bene, electric VEHICLE, not car.

      hcci is “just” going to be a more efficient way of burning gasoline in an IC engine, not that that is anything to be sniffed at, but we are talking about something on a par with fuel injection or electronic ignition, rather than the invention of the poppet valve.

      Consider the hayabusa, the engine has to be relatively light, it’s 1,300 cc and 175 bhp, so 7.4 cc per bhp.

      The dodge hellcat is 6,200 cc, and 707 bhp, making 8.76 cc per bhp, with a haybusa tune it would be 838 bhp

      Meanwhile back in the real world her indoors drives an 02 1,300 cc kia rio that makes 96 bhp, it costs MORE to road tax every year than the 01 1,950 cc volvo diesel that I drive, yet both make similar (to within 5%) mpg if driven steadily, the difference being the 2 litre volvo doesn’t slow down on steep hills and has overtaking power, whereas the 1.3 litre rio struggles at bit.

      hcci should make a 1300 cc petrol that can keep up with a 2000 cc diesel in those scenarios.. without having to either put the pedal to the metal or drop a cog.

      that’s my 2c anyway

      Comment by wimminz — September 2, 2017 @ 3:59 pm

  2. Cool, fukn saved for later.

    Comment by hans — September 2, 2017 @ 3:51 pm


RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

%d bloggers like this: