How do you interface the stepper contollers?

Hi Blackkrystal, welcome to the forum.

There are basically two different configurations with stepper motors ‘unipolar’ (5 wires) and ‘bipolar’ (4 wires) and the controllers for each are different and non-interchangeable. (Just to add a bit of confusion, there are also stepper motors with more than 5 wires and these can generally be wiring configured to work with either unipolar or bipolar controllers).


Bipolar stepper motors and controllers are probably the most commonly used with CNC machines and in my opinion the best to use.


Stepper motor controllers are usually interfaced to the computer with a ‘break-out board’ which can have an LPT parallel port connection or a USB connection or an Ethernet connection etc.

I use the parallel port with one PC and a USB with another PC but newer controlling software, such as Mach4, has been designed primarily for Ethernet connection (although a parallel port option will be available).


Hope this goes some way towards answering you questions.

Tweakie.

 

For more detailed technical information on stepper motors in general please see http://www.geckodrive.com/support.html

(also lots of other information on the Gecko site for those that are interested).

Tweakie.

 

No, it's not a joke but there is no backup and the seller is unlikely to help you if you have any problems with connecting or setting it up etc. but it's cheap enough to treat as a 'throw-away' if it goes wrong.

Personally, I would be happy to use this type of board to drive my CNC machines but others may disagree. It uses the LPT parallel port connection so you should ideally be using it with a PC (with onboard or card type parallel port). If your computer only has USB / Ethernet connection then there are probably better choices of controllers.

Hope this helps.

Tweakie.

 

I suppose it all depends on the load / force you are planning on putting on the cutter / router bit but I would think that NEMA 23 would be a more suitable size for a wood router.

Tweakie.

 

@JustinTime,

It’s perhaps going a bit off-topic but Mach4 will require an external motion controller (or pp plugin), most of which (SmoothStepper etc.) have now opted for the Ethernet connection as it is giving much superior performance over USB. I think the reason for dropping the parallel port is because new laptops just don’t have the Cannon 25 way pp connector any more.

My view, at present, is that I have a reliable, stable version of Mach3 that does everything that I want it to do so why would I want to change to Mach4 anyway – OK there are a lot of improvements including the algorithms controlling the motion but I still think I will be staying with Mach3 for the foreseeable future.

Tweakie.

 

Basically, just my opinion...

Many of these type of boards are assembled without proper quality control and never tested before dispatch – most work just fine but there will always be some don’t. They are often dispatched with incorrect setting-up instructions, pirate software and pirate licenses etc. etc.

If you browse the internet very few have any good words to say about the TB6560 boards but to be objective most have little understanding of how to treat and handle electronics in general.

One guy, doing a review, said that he destroyed one driver chip just by connecting his oscilloscope to the output – well had he known what he was doing and isolated his probe earth connection the outcome may have been different.

Another guy only swapped two stepper motors and it damaged the board – again, had he waited for the PSU to discharge before he disconnected a stepper motor the outcome may have been different.

The stories go on but my view is still that they are cheap enough to be ‘throw-away items’ if they fail.

Dependant on available finances, you could opt for something like the Gecko G540 which has excellent quality, support and documentation and would serve you well – perhaps it all boils down to ‘you get what you pay for’.

Tweakie.

 

I have this card. It's called the Sainsmart tb6560. Mine has 4 axes plus a spindle on relay. Everything works but I had to do a lot of research to configure linuxcnc. Maybe 1 week of late nights reading endless google searches. If you increase the current to the motors it will miss some steps because of noise in the board. There are workarounds that require soldering stuff to your board.

The older 3-axis boards have even more problems and require changing capacitors, optocouplers, and then all the stuff the 4 axis boards need.

Anyway, I used twisted noise-cancelling cat3 wires for all my motors and I bought high-end Nema 17's. These are so strong they overpower the routy router so it will be fine for your application. The max current of the TB6560 is 3.5A per motor and my motors are rated at 2.8A. You probably won't be able to run a high end nema 23 from this board but a low end will work. I run my motors at about 25% current(0.875A) and the x-axis has 2 motors so I run it at 50%(1.7A). The z-axis has to be at 50%(1.75A) or else it squeals.

So if you want 4 axes at 3.5A then you need 14A just for the motors. So maybe get a 20A power supply for future upgrades to nema23.

I kinda wish I had gotten something with less problems but hey that's what's available for that price. I really like the upcoming smoothieboard but I haven't quite figured out how to make it run more powerful motors or how to interface it to linuxcnc. Maybe when I get those resolved I'll make one myself .

 

The documentation provided with the tb6560 is for a different pin configuration than the one that comes with the documentation. Imagine trying to figure out where 20 pins go while sorting the 4 motor wires and 8 pin dip switches.

As far as the smoothieboard goes, linuxCNC has a parallel port interface but the smoothieboard doesn't. There's no real-time USB driver and the ethernet one is out but nobody knows how well it works. I'm planning to test the ethernet in the future.

 

Hey Black krystal,
Google hobbycnc ez driver. It's a kit you solder together (very simple, clear direvtions). Everything is all on one board, no breakout board needed.

However this board can only use unipolar motors. 6, 8 wire, which isn't a big deal.

From all of my research the are regarded to be much more reliable than the tb6560. I can vouch for it as well. I have a ez driver from them and it works great! They also have a 4 axis version.

I'll boil it down for you:
Cheapest to best drivers

Tb6560, hobbycnc, Gecko Drives And Smooth steppers.

I don't really regard arduino stepper controllers as a great solution

 

Look what I found: http://www.ecomorder.com/techref/ecomprice.asp?p=416074
I'd like to try to make my own board based off this design. It looks like it would even work with the OX and NEMA23's and maybe cost les than the TB6560 in a DIY rig. Anyone have experience with these? Will it work?

 

Well I was kinda thinking of manufacturing my own PCB with my routy and add it to the builds as a 4-axis controller board. We'll see, still dealing with my POS tb6560 missing steps.

 

Try running the cooling fan of the tb6560 board from an alternative power source and cooling the regulators of the board some how.
Should help a bit, bin there and done that, had to dissect my board and build my own stand alone boards.

 

Yes I agree with justintume that the tb6560 has it's problems, I am working with the tb6600 IC right now. I am attempting an all in one 3-4 axis
GRBL controller with SD for my OX AI.

 

I still haven't decided on a chip but I do recognize the enormous price and power savings these ICs give. My first build was an h-bridge with transistors so it took a lot of them and at a premium for money and power. I saw that the TB chips will go up to 3A on digikey but some others were claiming 4A(continuous). I'm curious what kind of current robert is pulling on his motors. What does the 6600 put out? I like the through-hole and one board = one axis design. If we built it like that then it could serve any break-out-board(parallel, ethernet,etc) and any number of axes. I know I'll probably need 5 axes when all is said and done plus a transition from parallel to ethernet. I'd like to be able to drive those heavy duty NEMA23 motors to lug around a 2.5HP router.... one day

 

I am running the tb6560 at 24vdc and 3 amps, anymore and your pushing it.
I had to also build a dedicated power distribution board in order to power up the IC in the correct spec sheet manor.
The TB6600 is the fixed/updated IC and can handle 4 amps.
I have tested the TB6600 with an arduino nano flashed with GRBL and they play nice
Here is the TB6560 setup I use on my tabletop mill

 

I'm having trouble finding the 6600 on digikey. They're like $8 on ebay. So I looked around and found a 16 pin DIP chip that does 4.5A
http://www.digikey.com/product-detail/en/TB6549PG(O)/TB6549PGO-ND/1730048
Would that work too? It's only $5.54

On the other hand, TI has a dual-motor controller for $8.94 that goes up to 7A
http://www.digikey.com/product-detail/en/DRV8412DDWR/296-25561-1-ND/2197040

 

You have to factor in the costs of the following:

1. Reverse-engineering the board
2. Replacing the fake Chinese components with OEM parts
   
3. Redesigning the board to correct the faulty design made by the Chinese EE with a fake degree
4. Fried parts as a result of 1-3
5. The time of everyone who helped you do 1-3
6. A soldering station, solder, flux, multimeter, etc

I can probably make the single-axis board myself for less than $20 in components and less time than troubleshooting+fixing a cheap Chinese board. This isn't my preferred activity but has become necessary due to the high price-points set by others on the one hand, and shoddy workmanship by the others for a low price on the other hand. I already tried downloading somebody's instructable DIY controller and wasted a lot of time and money on it only to discover that it won't work with my motors and needs lots of improvement. I've already lost a lot of time and money on this and am to the point where I realize the only way I'm going to get where I need to be is to do it myself. I can't keep spending money to try things and then find they have problems. I can't afford $300 for $50 in electronics either. I can't afford to be locked into a product that will go away in 6 months or has obsolescence designed into it. lol

 

Anyway, I think Robert's single-axis board would be most useful to me because I already have the TB6560 chips which are awesome if setup right. I have 4 of them and then in the future I could make a fifth with the 6600 or something even better. I can recycle my parallel port connector off my board and the optocouplers to make a breakout board that is properly isolated and can be later upgraded to ethernet. C'mon, who's with me? haha

 

The 6549 IC only controls 1 phase of the motor so you would need two IC's per motor putting cost above the 6600.
As for the second IC I have no idea but I'm not much of a smd guy "yet"

 

I have the single sided PCB layout in PDF I just need to find it.
Most successful way of making the board is positive photo resist.
The most important thing with the 6560 chip is power up, 5vdc first then motor power.
If you want to try salvaging your board let me know and I will send ya the PDF or you can try designing your own following this drawing
http://www.cuteminds.com/index.php/en/tb6560ahq

 

Yes please do send it. I had read about the power up sequence and it would save me a great deal of time to start with a board that "just works". Right now I'm rolling the dice every time I start my machine