[ImagineRobots]

James,

Based on your testing would you conclude that this chip doesn't suffer from the special power up sequencing requirement of the TB6560 chip? The requirement that logic power must be applied before motor power on power up and vice versa on power down.

thx

I just looked at the datasheet and it appears the chip is protected from the logic and motor supply sequencing design flaw of the infamous TB6560 chip. They recommend a power sequence but say the "ic can not break" if you don't follow it. This is good news if true. See data sheet excerpt below:

"We would like to recommend a way to turn on the power as shown below. However, if you do not do what we mentioned, IC can not break.
Turn on VDD. When the voltage has stabilized, turn on VMA/B.
In addition, set the Control Input pins to Low when inputting the power.
(All the Control Input pins are pulled down internally.)
Once the power is on, the CLK signal is received and excitation advances when RESET goes high and excitation is output when ENABLE goes high. If only RESET goes high, excitation won't be output and only the internal counter will advance. Likewise, if only ENABLE goes high, excitation won't advance even if the CLK signal is input and it will remain in the initial state.
"

[James Newton]

James,

Based on your testing would you conclude that this chip doesn't suffer from the special power up sequencing requirement of the TB6560 chip? The requirement that logic power must be applied before motor power on power up and vice versa on power down.

thx

Yeah, I couldn't kill it. If you must have a bipolar driver in that power range and you can't afford a GeckoDrive, this is the chip I would try. I'm considering making a kit available, but haven't see a lot of interest yet... everyone seems to be quite happy flushing money down the toilet on those cheap TB 6 5 6 0 drives from ebay, and you can't fight stupid.

[ImagineRobots]

Yeah, I couldn't kill it. If you must have a bipolar driver in that power range and you can't afford a GeckoDrive, this is the chip I would try. I'm considering making a kit available, but haven't see a lot of interest yet... everyone seems to be quite happy flushing money down the toilet on those cheap TB 6 5 6 0 drives from ebay, and you can't fight stupid.

I got burnt twice by buying TB6560 boards from Ebay. Just bought two of the THB6064AH test boards and will test the hell out of them when they arrive. I have a mobile 28v robot application where killing power to the chip and back driving of the stepper motors is common. Should be a good test to see how robust the chip is or isn't.

[James Newton]

I got burnt twice by buying TB6560 boards from Ebay. Just bought two of the THB6064AH test boards and will test the hell out of them when they arrive. I have a mobile 28v robot application where killing power to the chip and back driving of the stepper motors is common. Should be a good test to see how robust the chip is or isn't.

Let us know how it goes? I may be making a PCB for this driver chip in the future, and I'm more likely to do it as more people try it out, assuming they have a positive experience.

[dubstar_04]

James,

I have a hpgl based engraver mill that I would like to upgrade.

I have been contemplating buying a tb6560 driver board for months now but really put off by all the problems.

If you were to produce a kit based on this chip I would definately be interested and almost certainly purchase, depending on the price point.

Thanks,

Dan

[James Newton]

It's on my todo list... before the dragon even.

[mmomike]

I got burnt twice by buying TB6560 boards from Ebay. Just bought two of the THB6064AH test boards and will test the hell out of them when they arrive. I have a mobile 28v robot application where killing power to the chip and back driving of the stepper motors is common. Should be a good test to see how robust the chip is or isn't.

When you ordered the parts, did they say how long the shipping would take?
I am interested in working on this chip as well.

[lihaijiang]

When you ordered the parts, did they say how long the shipping would take?
I am interested in working on this chip as well.

Hi,my freind.

We always arrange delivery by China Post express mail. It will take 5-7days to USA.It's very safe and fast.

Last time,ImagineRobots has got his parcel in 4 days.

Regards,

[ImagineRobots]

Hi,my freind.

We always arrange delivery by China Post express mail. It will take 5-7days to USA.It's very safe and fast.

Last time,ImagineRobots have got his parcel in 4 days.

Regards,

I got the boards fast. Sitting on my table right now. Will test soon. If it tests good I will be laying out a board for my robot product and perhaps provide it for CNC use too.

[lihaijiang]

I got the boards fast. Sitting on my table right now. Will test soon. If it tests good I will be laying out a board for my robot product and perhaps provide it for CNC use too.

If you need any help,please contact me at any time.

[mmomike]

James and ImageRobots,

I am interested in designing a driver with the THB6064AH and the Adruino. Analog electronics is not my strong point. C and assembly programming and digital electronics are my strengths.

If either of you are interested in using the Adruino (or similar) and would like some help, let me know.

[ImagineRobots]

Let us know how it goes? I may be making a PCB for this driver chip in the future, and I'm more likely to do it as more people try it out, assuming they have a positive experience.

This chip appears to be rock solid so far based on my testing. My application is for a mobile robot that weighs over 100 pounds.

I'm running the chip off of two 12V lead acid batteries wired in series. I've run tests with my stepper motors wired for bipolar parallel and bipolar serial configuration. I definitely get more torque out of my motors compared to the Sanken SLA7078MPR unipolar chips in my original design (too bad I bought 250 of these chips already for a prototype run!)

I'm currently running the demo boards in 1/64 step configuration and 20% fast decay mode. This appears to produce the least amount of vibration at low rpms.

I'm going to make a few circuit modifications to the demo board and see if back driving of the steppers does anything bad to the IC. If results are good then the next step will be to layout a custom board for this IC. Based on my bad experience with the TB6560AHQ chips (the infamous "blue" stepper boards from Ebay) this THB6064AH chip appears to have fixed the design flaws.

[James Newton]

I'm starting work on a board design for the THB6064 which I will have prototyped (small / high cost run) and try out with the sample chips I received. I'll probably make one or two of those available for sale to anyone who wants to try them, and I'll refund the purchase price if they actually post results.

The PCB will be as small as possible to reduce production cost, and pretty simple. No opto-isolation, (unnecessary given a bit of care, although I may make a separate board at the same time with the isolators on it), no power regulators (that is just too easy to get from another source). It will have the PMinMO standard connector on it for step, direction, and logic power and micro screw terminals for the motor and power connections.

It will be double sided, PTH, soldermask, silkscreen and quite compact with tight traces / spaces and therefore cheap with a large board run, but not ideal for hobby / home production. I'll make the schematic available so anyone can easily layout a single sided version for people who want to etch their own... this isn't that. I'll host those layouts on the site if anyone contributes them.

If the initial testing goes well, and there is interest, I'll do a large board run and sell the PCB at a minimal cost. Then I will order a batch of chips from lihaijiang and spread the shipping cost over the units. The shipping on the free samples they sent was $40 so you can see that making a large order of chips is very necessary in order to get the per unit cost down. At $6.60, if you buy 3, they would cost about $20 each, but if you buy 40, they will cost $7.60 each.

Maybe some people will want to go in on that order to help finance it? Or I might do a kickstarter. The end result will be (I hope) a very reasonably priced kit that anyone can solder up for about the price of one of those crap TB6560 drivers... and which will then /actually/ perform the way they advertise that the TB6560 will perform, /without/ dying randomly on power up.

Does that sound ok? Anyone else want to do it instead?

[ImagineRobots]

I'm starting work on a board design for the THB6064 which I will have prototyped (small / high cost run) and try out with the sample chips I received. I'll probably make one or two of those available for sale to anyone who wants to try them, and I'll refund the purchase price if they actually post results.

....... Maybe some people will want to go in on that order to help finance it? Or I might do a kickstarter. The end result will be (I hope) a very reasonably priced kit that anyone can solder up for about the price of one of those crap TB6560 drivers... and which will then /actually/ perform the way they advertise that the TB6560 will perform, /without/ dying randomly on power up.

Does that sound ok? Anyone else want to do it instead?

James - I think you could start a sucessful Kickstarter using this chip. This chip is essentially the bipolar alternative to the rock solid unipolar SLA7078MPR driver but with the performance gains one gets from a bipolar drive. It's a gem that few know about. Too bad I have 251 SLA7078MPR chips I need to get rid of now that I have confidence in the THB6064 design. Will probably sell them cheap on Ebay.

The board I plan on designing will have circuitry that is tuned for mobile robotic applications so it won't be as simple as the board you are designing.

Also, gonna send you a PM about some other info.

[James Newton]

James - I think you could start a sucessful Kickstarter using this chip. This chip is essentially the bipolar alternative to the rock solid unipolar SLA7078MPR driver but with the performance gains one gets from a bipolar drive. It's a gem that few know about. Too bad I have 251 SLA7078MPR chips I need to get rid of now that I have confidence in the THB6064 design. Will probably sell them cheap on Ebay.

Are those anything like the SLA7062M we use in the SLAm Stepper Driver

The board I plan on designing will have circuitry that is tuned for mobile robotic applications so it won't be as simple as the board you are designing.

What sort of things do you need? Any chance I can accommodate all or part of your requirements? Having a known buyer for some of the boards would be a solid motivation. Perhaps we can cooperate to make a board set; so that your board plugs into my board and extends it to mobile robotics? I'd be happy to sell at cost if you invest design time up front.

[ImagineRobots]

Are those anything like the SLA7062M we use in the SLAm Stepper Driver

The SLA7078MPR has two extra pins, built in sense resistors and some circuitry to protect against open and shorted motors. This is the chip used in one of the HobbyCNC boards, a Probotix CNC board, and the controller sold by Sherline. It's a solid chip but in my robot application I don't get enough torque for my main drive motors at 24V. Forces me to use a bigger stepper. With the THB6064AH I can get away with a smaller stepper motor driven in bipolar configuration.

What sort of things do you need? Any chance I can accommodate all or part of your requirements? Having a known buyer for some of the boards would be a solid motivation. Perhaps we can cooperate to make a board set; so that your board plugs into my board and extends it to mobile robotics? I'd be happy to sell at cost if you invest design time up front.

I'm still coming up with requirements for the board but here are some givens:

Based on my tests the board needs to have a separate relay to detach motor power and another relay to detach +5V from the board in order to allow coasting. You can't have just one main relay to kill system power (call it Vpp) to all of the stepper drivers because a regen due to coasting just one motor could generate sufficient Vpp to power the entire system electronics (even when disconnected from the battery!) Coasting of the steppers is likely in a heavy mobile robot because people will push the robot to move it when unpowered.

I also need to make the microstepping selection controllable from an Arduino.

Up in the air is whether or not I add some type of digital pot to the board to make the coil current setting programmable.

Finally, if you use screw terminals add two extra screw terminals that go to nothing. This makes it real easy to connect an 8 wire stepper in either the bipolar parallel or bipolar serial configuration. In bipolar serial, each of the free screw terminals is used to connect a set of drive coils in series by shorting two wires together.

I'm gonna spend a few more weeks thinking and finalizing the requirements based on some additional testing. As you can see, this may be overkill for CNC requirements. BTW - I have 4 CNC machines: a ShopBot, converted X3 mill, crappy imported desktop router and a MakerBot Thing-O-Matic. I also bought a Magtrol electronic brake a few months ago to make a dynamometer for stepper motor testing (not built yet). In addition, I have a scope and current probe that can be used to examine the driver waveforms if you need any help with that.

For comparison testing I own just about one of every stepper driver board out there - Kelings, Gecko Systems, Hobby CNC, Probotix, Makerbot, couple of crappy TB6560 boards, LiniStepper, Pololu, etc. So far the THB6064AH is the best fit for my application.

[JohnnyVegas]

I may have 200 of these coming in soon. Was asked if I want them. Is there enough people wanting to purchase these? If so I would be happy to get them in and sell them very affordable.

[Hommersimpson]

i will be soon.. as I just got burnt by the TB6560 ebay board..

[James Newton]

Ok, I have a first version the the schematic done for a basic THB6064AH driver PCB / kit. I'm posting it here for review and comments / criticism, but before posting anything, please read this post so you understand my goals and why I did things a certain way? Thanks!

The design goals were:

1. Don't put it on the board unless it's absolutely necessary. I want to make a driver board that /just/ drives a stepper motor. Isolation, +5 power regulation, etc... are all things best done on secondary, optional boards. This block diagram: techref.massmind.org/techref/io/stepper/blockdiagram.htm lays out the complete system as I see it, with the THB6064AH driver as another option to replace the "Linstepper V2 / SLAm / PMinMO compatible" controller.


2. Reduce the number of components to simplify construction. One of the really nice things about the 6064 is that it's through hole, so it's easy to solder, but that isn't much good if you have to solder in a large quantity of other components.

- a. The pull up resistors for all the mode switches, EN, and the LEDs were set to the same value and consolidated with the current adjust pull up resistor. They can all be 2k7 without much trouble. That reduced 8 resistors into one network and saved some time and a couple cents. I'll include a bunch of holes in a Vcc bus so you can still use individual resistors if you like.

- b. The RESET circuit pull up resistor was about double the value of the OSC resistor, and 3 of 1 resistor value costs less than 2 of different values. That also allows for isolated network resistor which is quicker to solder in. However... I'm noticing that the only versions I can find in that value are SMT... 1206 SMT though... so not horrible... I'm going to try to set it up so that you can use a 51k and a 91K or 3 51K's or a thru hole OR SMT network.

- c. There are also 3 470 ohm resistors. Same as above.


2. Reduce the number of different parts so that I can bulk purchase the parts and still make enough of a profit on selling a kit of all parts to make it worth doing. If this doesn't work out, I'll still have the board... or rather, you will!

- a. Besides the resistors, there are many capacitors, so I'm hoping I can consolidate the values and stick with what I have in stock already: 0.22uF, 10uF, 100uF, 1000uF. See questions below.


3. Stick with open source standards.

- a. The connection for logic power, step, direction, and enable are on the PMinMO standard header. See DIY Interface Technical Information for more.

- b. All components easy to find at local electronics stores, or at least common in distribution. No weird stuff. If there are components which may only be available in a larger version then I provide holes for that and ignore the unused parts. Specifically, I placed an 8 switch DIP, although we only use 6. The space for the POT on the board will have holes for most common standard layouts of trim pots.

- c. And before someone asks, I'm releasing the schematic under creative commons, but the PCB layout will be a VERY tight multi-layer PCB and not really conducive to being made at home, so I probably won't publish it. I'll be happy to host other layouts on web site as I have for the Linistepper, etc..


NOTE: On the schematic, values put multiplier symbol in place of decimal point. e.g. u1 = 0.1uF = 100nF, u47 = 470nF, 47u = 47uF,

Non-critical resistor values have been changed to consolidate LED, switch and current set pot pull up resistors into one network thereby simplifying construction, reducing price and parts count.

If paths / space allows, will also consolidate common values in SMT 1206 iso networks inside thru hole pads to speed assembly
51K resistors with e.g. 744C083513JP (note: reset pullup resistor will be 92k)
470 resistors with e.g. 744C083471JP

Which brings me to the QUESTIONS I have and would love to hear back from lihaijiang on:

Questions:
- Is ok for C4,C5 and C11,C12 to be 10u / 100u vs 220u / 470u?

- Are C8,C9 really needed or were they included for the optos?

- Since +5 will be provided offboard, do we need C11,C12?

- Can C7 be a 0.1uF and R2 be eliminated? E.G. reset circuit is just a 51K resistor and a 0.1uF cap.

- What wattage and precision for the current sense resistors?

- 10K pull down resistors are included on each output line in the sample boards, but these are not mentioned in the datasheet. Are they necessary?

- Anything else on the schematic that looks bad?



...

ok, I guess that does it. Again, please don't comment without reading what I wrote here? But if you have read it, and something doesn't make sense, speak up. I don't mind answering questions, and I may have been stupid. It happens. Better to find out now.

[lihaijiang]

Hello James,

Thank you very much for post these here. Hope you can finish your board ASAP.As your request,answers as following:

1.One of C4,C5 and C11,C12 is 47u;the other depends on the current of the motor, it is about 100u to 470u,or more high.
2.C8,C9 can filter out external interference sources which is from input lead . It is about 200pf to 1000pf
3.It is necessary to leave a 47u cap. It means that you just can delete one from C11,C12.
4.Reset can be connected a 51K. And C7 can be set 0.1u or deleted. In another words,C7 can be a 0.1uF and R2 can be eliminated.
5.The sample resistor is related to the current of motor. The precision should be 1%.The current is below 2A,the resistor should be 1w0.22Ω;if it is over 2A(contain 2A), the resistor should be 2w0.22Ω.
6.In order to reduce induced electromotive force burr, we use 10k pull down resistors. I think it is a very simple. Of course, if you have another way to reduce the effect of chip from the induced electromotive force, you can use your method instead.

Look forward to your good news.If you still need my help,just post here.I'll try my best to help you.