Toshiba do not claim to have any on chip protection diodes. They don't need them if the chip is used in accordancve with the data sheet and the chip continues to operate until the currents in the coils of the motor are reduced to zero.
Parasitic diodes just suddenly appeared on page 24 of the linked Toshiba spec
I guess you meant Toshiba does not specify adding external diodes. I overlooked that Toshiba uses synchronous rectification (page 24, Slow mode).
On TA8435 external diodes are mandatory as the chip is not using synchronous rectification.
Considering this the diodes should run much colder with the TB6560. This of cause assumes that they are genuine Toshiba chips. Possible that there are chinese clones without synchronous rectification.
Originally Posted by wildwestpat
During the current limit and shape control the parasitic protection diodes play no part and can be taken out. (However the diodes are needed if or when both sides of the output circuit are non conducting. At that instant the diodes catch the back emf caused by the colapsing field and any mechanical energy due to the shaft of the motor continuing to spin.)
What about the 300ns dead time mentioned on page 25 ?
This is were the parasitic diodes become handy.
Assuming the same poor diode selection on the TA8435 using a 500ns speed diode is not helpfull here.
Actually on synchronous rectification design I would have suggested using a underrated diode but have it as fast as possible. Underrate is here possible as in a synchronous recticication design the diodes are only conducting for the small dead time. Hence the diode current is just a short pulse and we need no diode that can carry the full coil current permanently. E.g. the FR307 would be able to handle 200Amp peak current.
Originally Posted by wildwestpat
On reflection I am questioning the drive circuit for the Toshiba output chip. To get large amounts of power disipated in the protection diodes suggests that the chip is being turned off on a regular basis. This is not the purpose of the chip and is the most probable cause of the heating problem assuming the motors are just not way too large for the drivers. By too large I mean overly high inductance.
Just to sort things, I don't have the TB6560 board. I'm using a TA8435 board that need these diodes.
We need to have this verified from someone with the TB6560 board from eBay. If they also run cold it might be a simple suggestion to upgrade the board for the full 3Amps drive current by changing the reference resistors.
My warning was just an interpolation of the hot diodes on TA8435 boards of the same make with just 1,5Amps drive current and the same diodes seemingly being used on the TB6560 boards too.
I received my board a couple of days ago. The current set resistor appears to be red,red,silver,gold or .22ohm. The diodes read R307 MIC which is the best I can see with them soldered on the board. I have not yet powered up the board, but I will post feedback once I do.
I have also ordered the other board (red PCB) posted on this blog. It will take 10+ days to get. The red board has no diodes and .16 current set resistors based on the photograph on Ebay.
Do you have any information of the chip used in the driver board you are using?
On the TB6560 the point to note is that the two output stages are acting as a bridge and clamp opposite ends of the motor coil to alternate sides of the motor supply in rotation to cause the motor to nudge step to step. Since the output semiconductors can not change state instantly in response to the drive there an arrange driver dead gap so that the power supply is not shorted out by the power output semiconductors. This is normal practice and is arranged either by relying on the charge storage time being consistent chip to chip - batch to batch and opting for a fixed time gap OR by using logic to feed back a detected state that prevents the other half of the pair conducting until it has switched off to say half of the motor supply voltage.
I have designed very high power bridge circuits for inductive loads using this principle. When the current being switched reaches hundreds of amps the storage time in the switching semiconductors increases rapidly and a bit of firm ware logic sorts the problem with out 'letting go' of the ends of the inductance. There would be no reason for Toshiba to put this in the data sheet as it is integral with the computing part of their chip design. The time delay you are pointing to I believe relates to the change of operation of the chip and can be programmed by manipulating the appropriate pins.
Hope this helps but I agree we are all interested in our own boards and even assuming the parts are genuine the configurations are very different.
I have just looked up the inductance data an the three types of stepper I currently have to hand – it is pouring with rain and the workshop is too far from the house for a quick sprint!
These are the sort of motors used on small bench mills with dove tail slides and bags of friction even when fitted with ballscrews! A very different environment to a router mill where the motion speeds and distances are much higher.
Hope we can all become comfortable with our respective electronics.
I looked around the HYU site (as if I understand Chinese) and saw they make CNC Machines of all sizes. These boards must be from one of those machines.
I Also purchased a TB6560 off Ebay from Savebase. I was surprised it actually came shipped from China and it got here fast for being so far away. I am being careful and reading what everybody suggests before hooking it up. I'm using a dead Cricut as a test platform. The Cricut came with an 18v power supply so it shouldn't fry anything right off.
On my board they look Red, Red, Silver, and Gold. I don't know how to tell the 0.33ohm or 5%
"as Originally Posted by Mad Professor
The current sence resistors are, Orange, Orange, Silver, Gold, 0.33ohm 5%."
There are also some little surface mount resistors in same area and pattern with markings of 330K and on back are with markings of 1K and 104.
My board also appears to have the 0.22 ohm resistors and the R307-MIC diodes.
It's the 5-axis tb6560 "borad" from Savebase. www.hyu68.com is marked on the bottom of the board, so it's almost certainly the TB5DV-M (not to mention the manuals match what's available on the hyu68 site).
The board itself seems pretty chunky and reasonably high quality, at least compared to the Xylotex 4 axis I was using previously. So far, so good...
I finally had some time to test my TB6560 board from ebay today.(blue PCB) I purchased a 36V switching power supply to use with the board, but I didn't use it to test the board. I have a 3 amp variable voltage benchtop supply that I use when testing drivers. It has a digital display to allow me to see the amp draw. I started by testing the board at 24V. It powered up and it has 2 red leds to show power is on. Since they are near the 5V and 12V voltage regulators, I am assuming there is one for each voltage.
I attached an Automation Direct 276in-oz motor to the x axis rated at 2.8 amp. I had the board set at 100% current. With the board powered up, the motor wasn't on. I followed the instructions on the mini cd and configured Mach3. You must configure enable outputs in order to turn on the drivers. After configuring Mach 3, the motor locked up with power on as expected. I ran the motor at 1/8 and 1/16 step resolution and I tried all 4 settings for the decay. There is another led on the board that shows when the board is getting step pulses from the computer. This is no Geckodrive. Do not expect the motors to be quiet and vibration free. The motor ran the smoothest on the slow decay mode. When running the motor using the motor tuning screen, the board running 1 axis draws .5 amps from the power supply. The motor remained cool along with the bottom of the axis driver chip. I did accidentally touch the 12V regulator heat sink, and it was hot. I started testing maximum rpm, which was no where near the performance of the Geckodrive G540, but I was testing at half of the voltage that I run with the Geckodrive.
The next step was to turn up the voltage so that I could determine the top speed at the voltage that I was going to run at. I slowly turned up the voltage and tested the motor as I increased the voltage. Everything ran fine at 30V. I then turned up to 32V and smoked the driver chip. The silk screen in the board reads 12-36V. I let the board sit for a minute and then powered it up again. It maxed out the current on my test power supply. I am done testing.
As said back in post #5 I was running the board at 36vdc, the board did power up fine without smoking unto I turned it off and then back on again, So I am thinking it was more a voltage spike more then the set voltage at fault.
I removed the damaged chip and powered the board back up again with my bench power supply set to 13.8volts, and all the power leds came on, and the other axis still work.
Sorry for the misfortune you had. They need to rerate the board to a lower voltage. Maybe you can run 36v but with only a certain size or design. Can a motor design or size affect the amps/volt combo or is it my lack of understanding electric properties. It sounds like they did work fine at lower voltage. Thanks for the good advice.
It sounds like maybe my nema17's at 18v aren't as big a joke as I thought. I was looking for a 4axis at a good price. Even hobby circut were more. Hopefully it will work fine for me. I planed on using four of them for hot wire machine. I wonder what motor/volt work best with these. I thought I could upgrade to a bigger motors later. Maybe not. At least not too much bigger
The board that Tony promotes has a different looking design. Anyone using his? Anyone using his. it looks different than this hyu board. They make commercial cnc machines so it would be nice to find out what they rate the machine they go into.
My chip remained in tact. I only heard a small pop and the distinct odor of burning electronics. After powering down the board, further inspection by smell pinpointed that it indeed was the driver chip that failed.
At the time of failure, the motor was just holding, not turning. I had just finished testing at 30VDC. My plan was to slowly turn it up to 36VDC, since I had already purchased a 36V power supply for the board. I never made it past 32VDC.
I have a board on order as shown in the pictures posted by Tony, except that I had ordered it through Ebay before Tony added to this blog. I e-mailed Tony and he confirmed that his board can be run at 36VDC.
The motor ran the smoothest on the slow decay mode.
I hadn't fiddled with this setting until you mentioned this Ron thanks! I tried the setting and found with my smaller motors that the fast decay setting was smoothest.
Also I have found that with an increase in voltage there doesn't seem to be much to gain and the motors are drawing less current. With my 249oz-in stepper motors at 16v with no load they are drawing an extra 190ma, my system ready but sitting idle draws 440ma for a total of 630ma. But at 27.5v the motors draw only 110ma extra with a system idle draw of 340ma with only very slight improvements in speed and torque.
From what I have seen there seems very little benefit to pushing the limits with this board. One thing is for sure they are definitely not a 36v driver! This info may be of help to those considering purchasing the board and PSU.