3-Axis eBay stepper controller board - Port Setup

I had a EUREKA moment and had to post it as soon as I was sure I wasn't seeing things.

I recently took delivery of one of those Chinese stepper boards...Based on the Toshiba TB6560.

It thinks its machining a torus right now, motors whirring along nicely. These motors are Vexta 177 0z-in, 6V-1.2 Amps/phase. I'm running all this from a 15 volt, 5 amp linear power supply, with a 15000 Uf, 35 V capacitor for a filter.

I have been involved in electronics manufacturing for many years. I have to say I'm fairly impressed with these Chinese boards. The design is solid, and complies with the specs on the TB6560. Manufacturing had only one thing I didn't like. There was a lot of sticky solder flux on the board. No big deal. A little residue-free brake cleaner did the trick.

The Chinglish docs aren't even close to the correct pinouts. I've seen people's posts where they traced the 4-axis board's copper and came up with the correct pinout. Unfortunately, the 3-Axis and 4-Axis boards are pinned differently. I was inspired to grab a multimeter and ring out the board.

Here's the real pinout for the Chinese 3-Axis stepper driver:

PORT FUNCTION

Pin 1 - X Step
Pin 2 - Y Enable (amp enable)
Pin 3 - Y Dir
Pin 4 - Z Dir
Pin 5 - Z Step
Pin 6 - Z Enable (amp enable)
Pin 7 - X Dir
Pin 8 - Y Step
Pin 9 - Spindle (in this case, Relay)
Pin 14 - X Enable (amp enable)

The rest are inputs. Maybe another thread soon on limit switches?
Anyway, if you just set up your 3-axis stepper driver just like the above and it doesn't work, keep reading!

These boards only use Step (CLK), Direction (CW/CCW), Enable (amp enable) to control the motors. The TB6560's reset line is tied to Vcc, so a hard reset to these boards appears to require power cycling. Thats NOT the same thing as clicking the board power button in EMC2! That does disable the chips but it doesn't reset them. So far, I haven't seen any reason to need it though.

There is only one Amp Enable selection in the Wizard, so its pretty confusing at first. Seems odd, assigning the same function to three different pins... Hope they fix that some day.

So check your whatever.hal file. It should reflect the above port assignments in its own language:

net xstep => parport.0.pin-01-out
net yenable => parport.0.pin-02-out
net ydir => parport.0.pin-03-out
net zdir => parport.0.pin-04-out
net zstep => parport.0.pin-05-out
net zenable => parport.0.pin-06-out
net xdir => parport.0.pin-07-out
net ystep => parport.0.pin-08-out
net spindle-on => parport.0.pin-09-out
net xenable => parport.0.pin-14-out

If it doesn't contain this snippit of information, you can edit it easily enough. For some reason the Wizard just keeps assigning all three of the amp enables to the X axis in the hal file.

Another important thing. Check that your Parallel Port is in ECC mode. (Bi Directional mode probably works too). Traditional or "Normal Mode" is one direction only and doesn't play well with the PC's interrupts.

Those little 4-pin Opto Isolators are Sharp PC813's and are about a dollar in the US. I've blown one already by dropping a stepper motor on the circuit board while it was running. But hey, the PC is fine and that what their supposed to do.

Thank for letting me put in my 2 cents. I hope this helps someone. I could have used this information two days ago!


Would anyone be interested in a pictorial of how to signal trace this board?

Thanks for detailed information.

I also have TB6560 board, and managed to connect it to EMC2. But it's hard to confugure it with stepconf because of amp enable. When I assigned three pins to amp enable in stepconf, only one pin was actually working. Hal file must be edited.

That would be nice to add enable pins for each axis to stepconf.

Good point.
I noticed the chip enable pins all come up at once and stay on... making me wonder why they didn't just connect them together - operated from a single pin. *shrugs*

I must have a buggy install. The wizard made a mess of the hal file and severall times caused an error when I tried to test the port. I would assign X, Y, and Z, only to find the hal file contained different assignments altogether. Even deleting the file "my-mill.hal" didn't matter.

Hey garym

I am out looking for other stuff regarding Gecko G540 and EMC2 when I ran across this thread. I have another need for a 4 axis board, cnc hot wire foamcutter. There's number of choices on 3&4 axis Ebay boards, what color is yours?

Reason for asking, maybe you have seen the mmmm 40 page thread on "how I fixed my Chinese board" on cnczone. It's night mare for sure and not the faint hearted. Virtually all the pics I saw with this problem board has been Blue. There is a Red board also based on the same chips. One seller even goes as far as saying something like,, " unlike some other boards ours works".

Apparently a huge number of these Blue boards were made and they were "unloaded" to many vendors who are selling them of course. A few do work out of the box though???? If you have the Red one, I might gamble on a 4 axis Red board and see how it goes. Or would you mind saying who the seller was or an Ebay link maybe? Then I could see IF they have a 4 axis.

No expert here, enough to be dangerous but that Pin out is , let's say very different from Norm, no? And the pin out without a db25 pigtail crossing a few pins might be an issue. The software in mind to run it is limited to Pin out changes.

Regards
Dan
aka off to solve my Gecko/EMC2 issues,,,, a noob to both.

I couldn't tell you what the pinout norm is.. if there is one. I'm a newbie to CNC and an old timer with electronics. I can tell you there seem to be an awful lot of "standard" pinouts.

I have the blue board. The more I fool with it, the better I appreciate it.

Its a simple opto-isolator board with a little glue logic so the LPT port can talk to the TB6560. In electronics, simple = reliable. And its surprisingly robust. To date I have dropped a running servo on the energized board, and after many hours of no sleep, must have hot-swapped motors five or six times. So far no problems except a blown opto. Easy fix, since there are spares (unused) ones on the board...

Apparently a guy or company called www.hyu68.com makes these things. That domain is etched into the board. The website is in Chinese so I can't tell you what it says *S*. I don't know who makes the red board but I would guess its either a knock-off of the blue one or they could simply be a run of the same boards, with a red solder mask instead of blue. (anyone have further/better info?) When I start guessing I know its time for a nap.

You don't want to use my vender because I paid too much. He just resells them but I have a history with this guy so it was kind of a courtesy buy. Just use ANY seller with a 99.5 to 100% sales rating. And get tough on the shipping. It cost, I think, 21 yen to ship my board to me. Pennies via Chinese post.

I like this little board. $60 bucks isn't bad for 3 axis. One eBay vender sells a 4-axis board, with DRO board and manual control pendant for around 90 bucks. Do a search for "TB6560 + 4 axis".

Good luck!

Thanks Gary

It would be interesting to see IF you have the same board. Link to the "How I fixed my Chinese Board" thread with pics. This fix is quite involved. I now wonder if the Real Pinout you discovered has anything to do with Blue not working correctly. I don't recall any discussion on Pinout. I think this one is a lot cheaper than $69 or maybe with freight??

www.cnczone.com/forums/general_electroni...ixed_my_chinese.html

I would have to find the Red one, but I remember it looks quite a bit different.

I will take a look at your link.

Dan

WOW... The work that guy put into the reverse engineering isn't trivial. And he explained something I'd noticed when I was ringing out my board...The enable line's opto is wired wrong. (item 2), It works, only because of the pullup resistor, but it should be driven harder. I knew I saw something wrong, but I was more interested in pin function at the port. I'm still wondering if I can just remove all of the enable optos and wire each one high. Whats the diff? They turn on and stay on for the whole program run anyway.

I also agree the Step pulses should trigger on the leading edge and not the falling edge. Rather than cobble another board to the existing one, I'll just live without opto-isolation and wire across the opto's socket... Making it hc14 driven and inverting the signal. (I haven't proved this yet).