[kt88]

Hello,

First post here

I'm building my first CNC machine. Its nothing fancy, just need something to engrave simple texts onto front panels for the stuff I build.

I will probably be purchasing the cheap 3-axis TB6560 board from eBay. I know it has many issues, but unfortunately the TB6560 based boards are all I can afford.

One of the motors I'll be using is 1.4A rated (which I would like to run at about 85-90% of that), so these board that are rated for 1.5-3A on eBay won't do. I've contacted every single eBay seller I could find, and non of them have the 0.5-2.5A version.

My questions are:
1- What value of resistors do these boards use for the current sensing? Are they actually ~3A equivalent (I assume 0.15ohm), or are they higher value resistors (like 0.22ohm) and the add is wrong? I've asked the sellers about it and non of them had any idea what's the color of the resistors on the board. I don't want to trust the images in the eBay add as these could be old or of a different version of the board.
2- Is it ok to simply swap the resistors for something else (like 0.22ohm) or are there any other changes required for that? I assume no other changes are needed, but I would like to make sure.
3- Do you know where I can get the 0.5-2.5A version of that board (0.22ohm resistors)?
4- I thought about perhaps using 3 of these 1-axis boards and a simple break-out board. These boards offers many other current limiting values, and if one of them dies it'll be cheaper to replace. 3 of these and a basic break-out board should set me back around 60$ which I can live with. Does anyone here have experience with these boards? Are they any better (or at the very least of the same level - I dare not say quality ) as these cheap 3-axis boards?

[doorknob]

The TB6560 datasheet gives a formula for determining the value of the maximum current setting resistor, however IIRC there was some ambiguity in the written description.

I bought one of the 3-axis boards (from a Chinese supplier via eBay) that supposedly was rated for 1 amp maximum, and it uses 0.22 ohm resistors to set the max current (red, red, silver, gold), which differs from the datasheet formula.

I have another 3-axis board that I got from someone who gave up on the TB6560 - that was supposedly a 3 amp maximum board, and if I have interpreted the 5-color bands correctly (based on the calculator at http://www.hobby-hour.com/electronic...calculator.php), that resistor value is 0.25 ohms (black, red, green, silver, red) - see below, however.

I believe that I am reading the color bands from the correct direction, because reading them in the opposite direction gives an invalid value in the calculator. The red bands in the 0.25 ohm resistors are a muddy-looking color, but I think that they are actually supposed to be red rather than brown. However if I do interpret the muddy color as brown, then it does calculate out as 0.15 ohms (black, brown, green, silver, brown). So maybe that's the proper interpretation.

At present, I do not have either of the boards actually driving motors, so I can't provide any actual operating experience. But AFAIK, you should be able to change the resistor value to set the max current rating of the driver chip.

[doorknob]

Upon further reflection, I believe that the muddy color bars are actually supposed to be brown, which would make the resistor value 0.15 ohms for the 3 amp max rating.

I'm too lazy to get the soldering iron out to lift one end of one of the resistors and actually measure it, and I'd be hesitant to measure it in-circuit.

[kt88]

Thank you.

How about these 1-axis boards. Anyone has any experience with these?

[kt88]

I've noticed these boards have a "+" and a "-" input for each signal. Can the "-" be connected to ground or should it come from an inverter?

[doorknob]

I've noticed these boards have a "+" and a "-" input for each signal. Can the "-" be connected to ground or should it come from an inverter?

The + and - inputs on that board represent the input terminals of optoisolators.

So, if you are driving the board from a breakout board output pin that is capable of sourcing sufficient current to light up the LED inside the optoisolator, then you can connect that output pin to the + terminal and connect the - terminal to ground.

If, however, your breakout board pin has output pins that can sink current but not source it, then you would connect the output pins to the corresponding - terminal, and connect the + pins to a positive voltage source such as +5 volts (possibly through a resistor).

[kt88]

Thank you very much for that info. Is there any way to know what my PC parallel port can do without testing it myself? I don't think I can find it in the manual of the motherboard.

Isn't there some sort of standard all parallel ports must perform according to?

[doorknob]

According to Parallel port output:

The parallel port data pins are TTL outputs, that can both sink and source current. In ordinary parallel port implementations the data outputs are 74LS374 IC totem-pole TTL outputs which can source 2.6 mA and sink 24 mA.

It is possible that some parallel port implementations may differ. I have not recently looked at the spec sheet for the optoisolators used on the TB6560 single-axis driver boards, but it would be safest to assume that the parallel port is not capable of sourcing enough current to properly drive those optoisolators. So if you want to use the hookup that requires sourcing current, you should probably use a buffer between the parallel port and the stepper driver (such as a typical breakout board). Or you could just treat your parallel port outputs as if they meet TTL specs, and use them to sink rather than source current.

If I were going to use those drivers, I would prefer to have some sort of buffer between my parallel port and the driver inputs no matter which hookup I used. Since the driver board has its own optoisolators, I would avoid using a breakout board that also had onboard optoisolators, because using two optoisolators inline between the PC and the TB6560 chip might lead to unwanted softening of the control signals.

CNC4PC makes a buffer board without optoisolators (although I have not actually used that board). See http://www.cnc4pc.com/Store/osc/prod...roducts_id=203

[kt88]

Once again, thank you for the very helpful response.

I've looked at the spec, these opto's need minimum 5mA of current (15mA max). I think it would be best to use a simple proto-board with some CMOS buffers to drive the lines coming out of the PC.
I will add a simple USB connector to source 5V from the PC as well.

So I'll order a DB-25 connector, a couple of octal CMOS buffers, and a USB jack. Overall it should cost only a few $'s in parts. If the machine will end up working I can even use it to make a PCB for this board

I think I will first order 1 TB6560 board and see if it works well. I only need it to operate at 24V 1.2-1.4A for this particular stepper, so that's well within the specs of the IC. If it'll work well I'll order 2 more of these for the other axis as well.

[boldford]

. . . .I would avoid using a breakout board that also had onboard optoisolators, because using two optoisolators inline between the PC and the TB6560 chip might lead to unwanted softening of the control signals. . . . . .

I've encountered a similar issue in the set-up I'm integrating. In my case I'm probably going to retain the optos on the BoB and eliminate the optos on the single TB6560 board. This allows me to retain complete isolation between the PC parallel port and the stepper driver gubbins.

[dmc.tech.perth]

I have 1 of the 3 axis boards in question. I had a lot of trouble with it.
The pins in the TB6560 were broken at the point of the IC case.
The 12V reg ran so hot because of no heat sink.
But most of all the ENABLE line was tied to all the 3 chips, thus when the X or Y axis was moving the Z motor was locked, drawing lots of current, and getting hot.
Three seperate driver boards allow you to custom set the enable lines eg. Tie the X and Y enable lines together and Z on another line.
It is easy to set up the limit and E-stop switches as well.

With a seperate interface board I could set Mach3 up and test the outputs with a volt meter before connecting up the stepper driver boards.
Also I wired up 2 12V tube globes to the stepper outputs, thus making a dummy load so as to see the action of the output before I connected up the motors.
That's 2 globes in series per output, 4 in total. Some ICs don't like working without some sort of load on the output.
This made it a lot clearer to see where the problem was, Mach3, PC port, Signal inverted, Enable off, wrong pin, no clock, 2 functions set to the same pin in Mach3. ect. ect. Good luck.

[kt88]

I finally have some free time to order these parts. I've ordered 1 of the 1 axis TB6560 drivers, I will post when I have any news if someone is interested.

However, I have one more question. I was going over eBay and noticed these BOB. They are under 12$ shipped, this includes a cable as well. It has USB port for 5VDC as I was planing to use, and it has a 74HC24 buffer. I can't see the writing on the smaller IC's so I can't tell what these are.
Its 5-axis so I will only use 3 of these. It seems to have no optical isolation which is excellent as the TB6560 board already has these.

Any experience with these boards/comments on whether or not I should order one? The parts alone would cost the same when shipping is taken into account.

[lucas]

4- I thought about perhaps using 3 of these 1-axis boards and a simple break-out board. These boards offers many other current limiting values, and if one of them dies it'll be cheaper to replace. 3 of these and a basic break-out board should set me back around 60$ which I can live with. Does anyone here have experience with these boards? Are they any better (or at the very least of the same level - I dare not say quality ) as these cheap 3-axis boards?

There are several things on these type of boards wich I don't like at all.
Most important are those current setting switches, they have to carry part of the motor current and can't be decent quality considering the unit price. Their contact resistance needs to be in the low milli-ohm range. I don't trust this at all and would personally remove them and solder wire bridges to obtain the desired current selection.
Second one is the bending of the pins as already mentioned by dmc.tech.perth.
I also have my doubts on the quality of the "big?" electrolytic capacitor, it must be a low ESR type or it will fail rapidly.

I use my own design of the TB6560 and they still work fine.
Wich type of power supply are you going to use: switched or linear?

Edit: Apparently you ordered while I was typing.... Yes please post your findings with that drive.

The BOB wich you are refering to is indeed very cheap, you can't loose much, but I hope there's a decent instruction sheet included as there are so much options and connectors wich you don't need. It looks like they use some protocol to handle all that, the PC port hasn't enough pins to do all that straightforward.
They don't mention compatibilty with Mach, EMC or whatever. Ask them about this before you buy

[kt88]

At the moment I'm going to use a linear PS. I've picked one up for 25$ with 2x30V 5A max (its a lab PS, so I can set the voltages lower if I'd like).

Eventually I might replace it with a switcher. However, I will be adding significant capacitance at the output of the PS - about 15mF, and if there's a need I don't mind doubling it (I have lots of 6800uF/8200uF caps which I have no use for ).

[lucas]

You need a decent capacitor on the PS output, there's no need for an extremely huge one.
The TB6560 has a bad reputation due to the poor quality of the components used, slow opto's, design errors etc... on these cheap Chinese drives. And then there's the additional issue with some blowing for no apparent reason.
I have my own little theory on that "failing" of these chips and that relates to using cheap switched power supplies, these violate the powerup sequence much more than a decent linear supply.

So could you test that drive intensively with your lab supply at 24V (not higher) and then with a very cheap 24V switched one? This might confirm my theory if it blows with the latter one.....

[boldford]

I have one of the single axis boards in front of me now. There seem to be a few variants and I'd like to make a few comments on the particular example I have.

The one I have only has provision to drive the optos from the low end. I.e. in its current form you need to supply +5v to the board.
As supplied the board does have full opto-isolation between input and output although I have yet to determine if the opto devices are optimally biased. I am disappointed to note the issue of power-on sequence to the TB6560 is incorrect in that the ICs logic 5v is derived from the (nom) 24v. The TB6560 data sheet specifically calls for the 5v to be applied and settled before the motor voltage is applied. These needs attention before the board is put to work.
The board motor voltage is silk-screened 24v. I'm pleased the manufacturer has not perpetuated the error of some expecting these ICs to tolerate 36v.
When I unscrewed the IC and board from the, reasonably sized, heat-sink I found very little thermal compound. I would advise every one check and, if necessary, correct this issue.
I concur with most other posters' comments but would add that if treated as a part built kit readers, with a little patience and decent electronic skills, can modify the single axis boards into a useful driver so long as they do not expect this IC to reliably deliver 3.5A.

[kt88]

Hi lucas, I would rather not blow it

I'm going to use it at relatively moderate currents, 2A at most (actually around 1.4A for the motor I currently have, but I might make some changes).

I will add a couple of capacitors in parallel, that should keep the ESR low enough.

I will also check the thermal compound issue. I will probably replace the HS anyway as I have huge HS's which I'd like to use for all 3 boards eventually.

As for the start up sequence, I'm not sure what is the best way to treat it. Perhaps to supply the 5V externally (remove the on board 5V regulator), and add a relay at the 24V input? The main question is how important is it? This one is the most work

And finally, how about these breakout boards I linked to earlier? Anyone has any experience with them?

[boldford]

As for the start up sequence, I'm not sure what is the best way to treat it. Perhaps to supply the 5V externally (remove the on board 5V regulator), and add a relay at the 24V input? The main question is how important is it? This one is the most work

And finally, how about these breakout boards I linked to earlier? Anyone has any experience with them?

I would speculate the power-up sequencing is the single biggest cause of premature demise of this driver IC. I suspect the internal upper and lower FETs may both conduct until the logic has full control. Your proposed sequencing method is in line with what I have in mind but ensure the recommended power-down sequence is also respected. I.e. 24v off before the 5v.

[lucas]

I am disappointed to note the issue of power-on sequence to the TB6560 is incorrect in that the ICs logic 5v is derived from the (nom) 24v. The TB6560 data sheet specifically calls for the 5v to be applied and settled before the motor voltage is applied. These needs attention before the board is put to work.

Here's my theory in short on that issue:
A linear supply with a decent size filter cap will slowly build up the output voltage at 100Hz (120Hz in the US). The voltage after the 5V regulator will be able to follow the input one. There will be 5V present when the input is at 6-7V and the logic can initialise, if the reset configuration is not extremely long and the filter cap behind the 5V regulator not too big.
A switched power supply (these have much smaller filter cap's) and they build up the voltage at 20+kHz will reach the desired output voltage in no-time, much much faster than the linear one. When the logic initialises: the motor supply is already too high.

I think that a mishap during initialisation doesn't harm the chip when the supply is at 6-7V, but it blows when it's higher.

This is just my theory, could be wrong and would like to know what others think.

[lucas]

I will add a couple of capacitors in parallel, that should keep the ESR low enough.

The low ESR cap must be the one on the drive to absorb the switching currents and spikes. You will have those through your supply wiring creating problems and maybe smoke if it isn't a good one on the drive itself.
The output cap on the supply can be a normal one.
There's only one decent solution if you can't replace a suspected one on the drive due to space or whatever: choose a good one and connect it together with the wires in the drive's supply connector. It must be as close to the chip as possible.

And finally, how about these breakout boards I linked to earlier? Anyone has any experience with them

I edited a previous post wille you already replied and thus didn't see it, please see previous message.

Luc.