Stepper motor driver

[Peter BellParticipant @peterbell11509]

I'm driving a Nema 17 stepper motor driver with pulses from a voltage controlled oscillator. The motor runs fine at high speeds but when I run at slow speeds ie,20rpm the black driver keeps flashing the alarm light and the motor sounds rough, the open frame one sounds better but at slow speeds the motor makes a ticking noise.

Tried various switch settings for pulses/rev and current and adjusting the input frequency to maintain a low speed but no real difference in the noises.

Is this a characteristic of these cheapo drivers, am I expecting too much or are there smoother alternatives?

Peter

[Peter BellParticipant @peterbell11509]

[John HaineParticipant @johnhaine32865]

20 rpm is pretty fast for a stepper! That's 4000 Hz step rate. Or do you mean 20 Hz? Steppers do have resonances where the inertia of the motor and the reluctance torque of the magnetic circuit are close to the drive frequency, and it may be better just to avoid these speeds.

What main supply voltage are you using – and microstep ratio?

[Neil WyattModerator @neilwyatt]

20 rpm is 66Hz without microstepping.

A typical NEMA 17 will run up to about 1800rpm at about 1/5 full torque.

Microstepping will be smoother BUT try changing the decay mode, at low speeds they become 'lumpy' as the inductance causes the small signals to current limit before the motor actually moves so many of the microsteps don't register. Changing decay mode should help this.

Neil

[John HaineParticipant @johnhaine32865]

Oops! Minutes, not seconds. I should engage brain…

[Peter BellParticipant @peterbell11509]

I'm on 1/8th step with 25v supply.

Thanks for the explanation about smoothing, never heard of that. Its on 100% smoothing, if I reduce it in stages the movement becomes progressivly more jerky.

It cycling between 73 and 2350hz under the control of a picaxe to the vco (4046) but that might have changed as I've twiddled!

When I set it for 1/16th steps the speed increases so still wonder if the driver is faulty?

As its a very cheap driver do they become smoother with price, notice some have 4 dip switches to set the step rate?

[Neil WyattModerator @neilwyatt]

Confused – you have two different drivers in the photo, which ones are you using and can you show a photo of the dip switch settings?

A Pololu will work a Nema 17 and costs pence, but you will need to short a resistor on the board for smooth low speed operation.

Neil

[Peter BellParticipant @peterbell11509]

Sorry–the black one is the very odd noisy one and the open frame (2nd pic) ticks and seems to increase the step rate to 1/16.

Just glanced and see plenty of Pololu—like to give one a try–any particular reccomendations?

Peter

[SillyOldDufferModerator @sillyoldduffer]

Odd noises from stepper motors aren't unusual, but the alarm light is worrying.

Some possibilities:

• Incorrect current switch setting
• Power Supply isn't man enough, or it's a Regulated PSU that doesn't like pulsing.
• The control pulses are mishaped or mistimed.

That type of driver is well known and I've not noticed many complaints about it on the web. I don't think it's faulty.

There are a lot of variables and it might be a combination. Can you provide more details of your set up? Circuit diagram, identity of the VCO, PSU and all the switch settings.

Dave

[MuzzerParticipant @muzzer]

Bit of a strange mix of technologies – digital stepper driver and (digital) micro – with an analog VCO in between. Can't you simply generate the pulses directly? It may be that the Picaxe isn't up to the job but there are so many microcontrollers about these days such as 32 bit Arduino compatibles that you shouldn't struggle to manage it.

There's a pretty good roundup of potential options on the Mickmake Youtube channel if you want some ideas.

Murray

[Peter BellParticipant @peterbell11509]

Thanks for the replies. PSU is easily coping with the current which is around 0.2-0.5a as it cycles through its sequence, pulses look good as well on the scope. Tried various current settings without much effect on the sound

Agree its a mix of technologies but a magnetic sensor clocks the shaft movement, (its a belt driven linear actuator), the picaxe counts the sensor pulses so fairly low speed and closes the loop

Its been performing quite nicely with repeatable numbers from the movement for days so dont really want to change too much, was unsure if I had faulty drivers.

Going to follow Neil's suggestion so ordered some Polou drivers to play with–is the low speed mod detailed anywhere?

Thanks

Peter

[Neil WyattModerator @neilwyatt]

Posted by Peter Bell on 16/10/2017 22:00:49:

Thanks for the replies. PSU is easily coping with the current which is around 0.2-0.5a as it cycles through its sequence, pulses look good as well on the scope. Tried various current settings without much effect on the sound

Going to follow Neil's suggestion so ordered some Polou drivers to play with–is the low speed mod detailed anywhere?

0.2-0.5A is quite modest.

0.2A is about right for a Nema 17 on a 3D printer, a rotary table is a bigger load.

OK for the steppers to run warm.

Here:

stargazerslounge.com/topic/268438-pololu-uneven-microstepping-solution/?tab=comments#comment-2938666

<edit> the answer for you should be to set the decay to 0%

Edited By Neil Wyatt on 16/10/2017 22:25:50

[Peter BellParticipant @peterbell11509]

Thanks Neil, thats really good info and things to try I would never have found!

Peter

[Neil WyattModerator @neilwyatt]

To be honest it took me ages to find out the solution, it made a big difference to the smoothness of my scope drive.

Neil

[Peter BellParticipant @peterbell11509]

Ordered some parts and now at the waiting stage.

Its just occurred to me that with my present arrangement the controller I use to drive the stepper motor has opto isolators for clock, dir and enable which I feed with 5v at the appropriate time and also a 5v pulse for clock.

The 3D drivers (DRV8825, TMC2100) I intend trying have no opto isolation. Does anyone have the signal levels that a 3D printer runs these drivers on especially the clock? Has anyone put a scope on these?

I'd like to avoid frying too many before I get something working!

Thanks Peter

[John HaineParticipant @johnhaine32865]

The data sheet here indicates rather indirectly that the TMC2100 chip is intended to run with 5V logic levels.

[Peter BellParticipant @peterbell11509]

Thanks John, Read the data sheet a few times previously but couldnt really find any figures that applied apart from the blandish statement below

"The STEP and DIR inputs provide a simple, standard interface compatible with many existing motion
controllers".

Peter

[John HaineParticipant @johnhaine32865]

Section 14 absolute max ratings: I/O supply voltage max is 5.5 volts; logic input voltage level is V I/O max +0.5 volts.

[Peter BellParticipant @peterbell11509]

Now got my TMC 2100 but I'm not having any success in making it go using the simple pulse generator I used previously. Connected as a standalone driver for one stepper motor as per the pic feeding a nice buffered 5v 274Hz square wave.

From the data sheet info below I naively hoped that all I had to do was hold dir negative, apply a pulse and it would go but dont think it as simple as that! Looks like I've got timing issues at least?

Has anyone used a driver like this as a standalone stepper driver and not on a 3D printer with all the other driver grubbins?

Thanks Peter

[Neil WyattModerator @neilwyatt]

You need to make sure the 'enable' pin is held low/high as appropriate.

[Neil WyattModerator @neilwyatt]

Have you got the enable pin wired up to the right signal level. left unconnected it will probably default to 'not enabled'

[Peter BellParticipant @peterbell11509]

Thanks Neil, I've tried taking enable low, floating and high as per the data sheet but makes no difference. Looks like it uses a pulse into dir as well as step to make it go in the right direction from this below from the data sheet. Looking at the data in 11.1 in previous post had hoped that I only needed to keep dir low to make it go. Wrong again!

Peter

[David JuppParticipant @davidjupp51506]

I don't read the data sheet extract as suggesting pulse needed on DIR. It just says that the state of DIR input decides which direction the STEP pulse is treated as requiring. I can see how you could read it to think it's talking about pulses on DIR.

I've used a similar driver board, no need for pulses on DIR.

[SillyOldDufferModerator @sillyoldduffer]

Hi Peter,

Just had a closer look at this device and the module you have it mounted in.

Firstly, I find you can't read the chip spec and understand from it what's needed to make the motor work. That's because the module contains the control circuits you have access to. It's the module that needs to be understood if you can find a manual for that. Anyway, the crux seems to be that the module board has six solder links that are made or broken to select the chip's operating mode. You need to find out what to do with those solder links; try watching this chap's second video, I think he provides the answer you need. You may need to replay part of it a few times to catch exactly what he says!

Secondly, the module is not 'plug and play' like the Microstepper box you started with. The underlying chip provides two non-standard modes. They're compromise modes potentially useful compared to a conventional driver. In 'Stealthchop' mode , the chip achieves smooth quiet operation by driving the motor constant voltage rather than constant current. As there's no such thing as a free lunch, in quiet mode the stepper motor has low turning power and will miss steps or stall if overloaded. If your application needs to count steps accurately be very careful with this one. Alternatively, the chip can be set to work in 'Spreadcycle Mode'. This uses 'intelligent current' control to provide more grunt at the price of more noise. Be warned, this mode tends to 'fall apart' at higher currents. Both modes have various configuration settings that need to be selected.

So I think the main reason it's not working because you've missed the step where you solder up the right combination of the 6 jumper pads on the module. Once you done them correctly it should work provided the load isn't too heavy and you don't over-accelerate the motor. I agree with Neil about ENABLE and with David about DIR being held at a constant value.

Interesting and useful module, but not exactly user friendly!

Hope that helps,

Dave

[Peter BellParticipant @peterbell11509]

Hi, Thanks for the replies, its really useful having another slant on interpereting the published data. On reading it again it says about sampling dir on each leading edge of the step pulse to determine direction which is different to what I originally thought. I messaged TRINAMIC the makers of the chip and got back

"the 2100 will need a step & direction signal. This means, each step pulse makes the motor go a little bit. If this is, what you need, you can operate the motor as desired without any additional interfacing"

Which sort of agrees with keeping dir to 0v and providing a step pulse. Pretty good video, looks like the jumper between CFG1 and 0v allow selection of spreadCycle with options to tune the exact drive mode with other obscure connections!.

On a table I have it looks like the stepper defaults to stealthChop with all the jumpers out so my conclusion is it should do something wired as is with by providing a step pulse, and with dir and enable held at 0v. I'm now wondering if its faulty and I've ordered replacements.

David J, I would be interested in further info on what connections you use with your board and what are you using to provide the step pulse?

Thanks once again

Peter

[Author]

Posts