How I fixed my Chinese TB6560 controller (updated)

i have a few questions for the guys running these drivers.
i purchased(traded) for a 3 axis and 4 axis board. the dip switches i was told were set to properly run my motors(1-4 on, 5-6 off) and i was told not to mess with them. i am running a spectralight/sherline machine with the original round motors(see pic). i plan on running 24v 15a psu. but i believe the driver i got with the machines was only 12v(burned up driver when recieved) is this going to damage these motors? i was told no but i am just checking my source. i had to replace two can type capacitors on the 3 axis, one was leaking and another was installed backwards and bulged. i have not done any modifications apart from adding on leads to the input plug. what should i do to get this running correctly and reliably? i am currently getting ready to run with the 3 axis board.
pics of both boards when i received them and a pic of one of my mills is attached.
thanks.


Kevin.
Nsr customs.

After reading this thread and others across the web...

I have the 5 axis board. My mission was to test a 4 channel cnc with a 5th axis for rotary. This seemed to be dream board tb6560.

This is the series of challenges so far with this thing.
Got one board and hooked it up to a 34 volt 10 amp power supply.
It ran great on X A and Z. The Y channel slaved with A did not turn. A would turn but was jumpy. And the channels would leak. Running the Z would leak to the X.
Tried a different computer and same thing. Went back to the original computer and powered up. Jogged the x and it turned then suddenly BAM!
The B channel blew out a chunk of meat on the bottom. B channel was not activated in mach. Looked at the board and sure enough there was a very tiny hair line solder bridge along with gobs of solder flux all over.

Contacted Mega supply from ebay and they gave me a refund on returning the board while they sent me a new one. I am happy.

The 2nd board. Hooked it up with a 24V 15 A power supply and tried it on the original computer.

Tried an old P4 XP machine with a configurable parallel port and X Y Z and A channels work. Not test on the B channel. However it is not smooth running. Also the mach has to have the spindle activated to get the motors to lock. When using the stop button in mach it also deactivates to spindle and unlocks the motors. I have no desire to research Mach to find some spindle setting to bypass this.

I have a Gecko g540 that runs one cnc mill just fine. I also have a hobby cnc board running the other mill. So I have all I need in boards. If someone does not have the money to get a gecko G540 or other like system I would suggest getting the Hobbycnc board. It is low priced yet works well. It has run my mill for about 2 years without a single complaint.

So far the second board sent to me by Mega supply runs. But it is not something I would use. Likewise I do not have the time or motivation to make any mods. I am sure some one could obtain this board from me handle the mods and and be happy experimenting with it. If you want it pls PM me with a reasonable offer with shipping and it is yours as is. I will send the hook up tabs, the parallel cable and the cd with the basic settings manual.

Hi guys,

after messing with the blue version of the board, and getting it to work pretty well after adding the buffer, etc., I found this guy who hacked the newest improved version of the board, the red board :

Contrôleur pas à pas Toshiba TB6560 et pilotage micropas: Pertes de pas et problèmes de fréquence d'horloge - Civade.com

(use google translate if you can not read french, or post your question and I will answer as best as I can)
It does work quite well after the capacitors are replaced as he describes. A much easier fix to do, and a great bargain, I think. If you are looking to buy that board, it is the red board version "V".

I am sorry I don't follow this site closely and didn't see your question. You probably have it figured out by now but you are on the right track. I also removed the 74HC14's and I jumpered from the input of the 74HC14's to the output of the opto isolators. The are three isolators wired backwards that are detailed in other posts that allso need to be corrected. Mine has not missed a lick since I fixed it.

My experience with the slow optos is, you can't get shorter pulses than about 70 µs (about 60 µs longer than the step pulse length at the parallel port).
With respect to the rule in the data sheet of the TB6560 that the duty cycle should be about 50%. You get a period of 140 µs which equals to a step rate of 7 kHz.
This is possible after fixing the wiring of the optos in the step signal lines and adding another 74HC14 between output of optos and TB6065.

If you want to achieve higher step rates, the optos of the step lines have to be removed and bridged.
The max step rate of the TB6065 is 15 kHz according to the data sheet.

A max step rate 15 of kHz gives a minimum step pulse width (50% duty cycle) of 33 µs (1/15kHz/2)
The data sheet states, the period time of the chopper frequency should be smaller than 1/4 of min pulse width, giving a minimum chopper frequency of 120 kHz which can be achieved with a 330 pF cap.
Much smaller caps IMO don't make much sense, because with higher chopping frequencies the efficiency of the stepper motor is reduced.

So I fixed my controller as follows:
1. replaced the 1000 pF cap with a 220 pF cap (330 pF would also do)
2. removed the driver current manipulation circuit by removing the 10 k Ohm resistor (marked 4.7 k Ohm on the silkscreen); the resistor can be reused for the next step
3. added an 74HC14 to the step lines to create the right polarity of the step signal for the TB6560
4. removed the optos from the step lines and bridged from pin 2 to pin 4
5. fixed the issue with the wrong wired opto on the enable line (was already fixed on the layout of my board, so nothing to do, but you should check)
6. grounded the heat sink

Plus
This gives me nicely rotating stepper motors, even with 1/2, 1/8 and 1/16 steps.
I can achieve the full step rate the TB6560 allows (currently my motors limiting the step rate to about 12 kHz in 1/2 mode).
The axis activity indicator led is working properly.

Minus:
No optical isolation.

I will run the CNC from an
- Raspberry Pi
- running Universal-G-Code-Sender
- streaming the G-Code to an Arduino Mini which is connected to the parallel port of the china-driver
- and running Grbl as firmware on the Arduino

So missing optical isolation is not that problem, because damage should be limited to the Arduino Mini

Hi all. I have a Blue board.
I have fixed the heat sink problem by isolation with non conducting heat sink cloth.
Fixed the pin over hang, too close to PCB track.
Removed the 3 diodes to the Ma Mb lines and replaced them with fuses
plus a switch. The Toshiba spec is very specific. Apply Vdd (+5) before MaMb Supply or
the chip may expire (smoke!!!)

Apart from the clock chip, are there any other mods I should do to the board?

Thanks
Phil
Attachment 180872Attachment 180873Attachment 180874

Quote:

---

Originally Posted by ding512

Apart from the clock chip, are there any other mods I should do to the board?

---

Bypass the optos

Hi, I recently bought one of these 3 axis boards. At first it seemed to be working. I tried all axis separately with a small stepper motor while I as waiting for my new ones to arrive. Once I got my new steppers I went to try it out and now I'm having a bunch of problems. I am using kcam4 on xp.

So now the x axis Is not working because it is receiving constant voltage across a+ and a- with both the power supply and the db25 connected.

The y and z axis were working but now the y is not working at all and the z axis is only one step when I press the jog key.

Any help will be appreciated. Thanks, Nathan.

I would like to maintain the isolation, should I look for better opto's ?
Thanks Phil

The way these boards are laid out, the optos don't provide any isolation.

The 5v supply and ground of the board and the PC are connected, so the optos provide only the slowing of the input rise time.

Quote:

---

Originally Posted by nlh205

So now the x axis Is not working because it is receiving constant voltage across a+ and a- with both the power supply and the db25 connected.

---

Most of the 6560 boards are incompetently designed. They will self destruct under normal use. They will also do so if you disconnect a motor with the power applied.

Unless you have electronic skills, don't waste time with it. Replace it with a tb6600 board. Some are equally poorly designed, but the chip is better and won't self destruct as easily.

I am fairly good with electronics. But to be clear what your saying is that most likely it is a bad tb6560 chip? If so is there any way i could possibly test the chip itself. Thanks for the reply.

Yes. If it's dumping the full power supply current into your motor, then it's bad for sure. It will burn the motor and power if you leave it on too long.

The 6560 chip is quite cheap on Ebay. Swap in another and see if it work. Also add in a power supply sequencer to prevent it from dying again.

Quote:

---

Originally Posted by nlh205

I am fairly good with electronics. But to be clear what your saying is that most likely it is a bad tb6560 chip? If so is there any way i could possibly test the chip itself. Thanks for the reply.

---

I have 3 axis driver TB6560 from the link:
One 3 Axis TB6560 CNC Stepper Motor Driver Controller Board High Quality | eBay

Will I burn it if I connect it to power supply - DC 12V 53A (514W)?

Quote:

---

Originally Posted by modric

I have 3 axis driver TB6560 from the link:
One 3 Axis TB6560 CNC Stepper Motor Driver Controller Board High Quality | eBay

Will I burn it if I connect it to power supply - DC 12V 53A (514W)?

---

12V is not enough. You need 24V. not so many amps. maybe 15...
Using 12V will be OK for testing. it will work but not be as fast as you want. so many fixes reqd for this board but main ones are detailed in my post (pg 55). Also testing procedure for your PC port.

Quote:

---

Originally Posted by modric

I have 3 axis driver TB6560 from the link:
One 3 Axis TB6560 CNC Stepper Motor Driver Controller Board High Quality | eBay

Will I burn it if I connect it to power supply - DC 12V 53A (514W)?

---

Are you concerned about the 53A rating of the power supply?

The board should not draw that much current - it has onboard regulation that will (or should) only draw the amount of current that it is set to handle. The 53A rating is a maximum "capability". It is not the amount of current that it will force through a connected circuit (unless that circuit has a very low effective resistance).

I know that speed will be affected but that is power supply what I have in this moment. Using M8 as drive axis and 12V it will be slow I know.

Quote:

---

Originally Posted by doorknob

Are you concerned about the 53A rating of the power supply?

The board should not draw that much current - it has onboard regulation that will (or should) only draw the amount of current that it is set to handle. The 53A rating is a maximum "capability". It is not the amount of current that it will force through a connected circuit (unless that circuit has a very low effective resistance).

---

This "should" in your first sentence stands. :) I will have 2x6A (12) active coolers (I sow pepople writing overheating is problem), 2x 4cmx4cm small coolers on heatsink, I will also have some leds etc. I am thinking to put 10 A fuse before driver because of this "regulation".

Quote:

---

Originally Posted by modric

This "should" in your first sentence stands. :) I will have 2x6A (12) active coolers (I sow pepople writing overheating is problem), 2x 4cmx4cm small coolers on heatsink, I will also have some leds etc. I am thinking to put 10 A fuse before driver because of this "regulation".

---

The typical overheating that people complain about with that board is overheating of the 3-terminal voltage regulators with the tiny heat sinks that are used to drop the motor voltage down to 12 volts for the fan and 5 volts for the logic. If you're powering the board with 12 volts, then the 12 volt regulator will not pose a heating problem, while the 5-volt regulator heating will be largely unchanged. So I suspect that incorporating active cooling on the main heat sink is not going to give you much of an advantage, if any.

Thanks for reply. I didn't know that. I will also power active heat sink on chips directly fom power supply. I sow people wrote it is because bad gnd. I have blue tb6560AHQ driver from china. Do you know is this driver also affected with this things from this thread?