[James Newton]

Looks nice, but I think all these mounting options are even easier on the version you did with the chip laying down off the edge of the PCB? No?

[lihaijiang]

@ lihaijiang: the diodes are BYV28 trr max. = 30nsec.
Did you experiment on the effect of the diodes? The results maybe different depending on the drive settings, supply voltage, motor induction... I just don't have time enough to test this all out.

Dear Luc.

Three factors have influenced temperature of chip:reverse recovery time of diodes, decay mode and decay time. 30nsec. is too fast,so you can change the decay time in order to reduce the temperature of chip.

[lucas]

Dear Luc.

Three factors have influenced temperature of chip:reverse recovery time of diodes, decay mode and decay time. 30nsec. is too fast,so you can change the decay time in order to reduce the temperature of chip.

I don't fully inderstand what you mean (30nsec is too fast), wich type of diodes and settings would you recommend to obtain the lowest chip temperature?

Looks nice, but I think all these mounting options are even easier on the version you did with the chip laying down off the edge of the PCB? No?

Yes, that's the big advantage of "the flat" version, I received both styles of PCB's at the same time, didn't assemble the one wich can be seen in the previous messages, just showed some pics.
The flat version is assembled and tested, looks really nice.
Only possible problem is the clearance between PCB and heatsink, most component pins protruding the PCB must be cut as short as possible so that they never can't touch the heatsink.
Pcb size is 50x50 mm (approx. 2"x2"). The heatsink isn't big and I get approx the same temperature results at 3 Amp. Of course a bigger one is dead easy.
And this time I mounted it with the required screw to hold the Pcb in place..

[lihaijiang]

I don't fully inderstand what you mean (30nsec is too fast), wich type of diodes and settings would you recommend to obtain the lowest chip temperature?

When the reverse recovery is 30nsec, it is helpful for reducing heat.If you still think the chip is too hot,you can add the resistor value of PIN23. By this way, the decay time will be longer.

[lucas]

@lihaijiang:

So you are saying that the choice of diodes is correct?
Changing that resistor is not an option, I wanted to see the difference with and without diodes and that should be the only difference between the 2 drives.

It's mentioned in the tips document, I assume it has been tested.
Thus my question to you: wich diodes were used and what were the results (differences)?

It's normal that he chip get's hot with a small heatsink at 3Amp, it has to dissipate approx 8 Watt but had expected a substantial difference with the diodes.

[lucas]

@ everybody else:

Sorry for my rambling on the effect of the diodes but I need a clear answer, it looks like I have been put on the wrong foot and did spend quite a bit of time and money investigating the differences wich seems to be minimal.

Let's continue with the issue on hand: the 2 designs, normal(traditional) and flat one.

A lot has been said and one must read the whole thread to get an idea of the differences between them.
I am working on a website, it's still very much under construction but this page list the 2 versions with their options and differences:

http://users.skynet.be/ldt/CNC%20ele...THB6064AH.html

As you can see there: I will offer several options of kit's, I am working on the pricing and will try to finish that asap.

Luc.

[lihaijiang]

Yes, the choice of diodes is correct.

We usually use Schottky diodes. But I think your diodes is OK.It can help reduce heat by several degrees temperature.Obviously, it don't meet your requirements.
If it is still hot,just try other ways to cool it, as big heasink or big resistor of PIN23.

[lucas]

lihaijiang,

thanks for your reply, this does confirm my results. I expected a bigger difference.
The diodes are usefull if one is looking for 2-3°C temperature drop in a very specific application.

This designs I made are not specific, they should be able to exploit the chip in different configurations and settings, 8 diodes with little result doesn't justify the costs and PCB space.

I was always wondering why it was mentioned in the tips document but they were not installed on the evaluation PCB.

Luc.

[lucas]

Yesterday I assembled my "traditional" version without diodes and a "profi heatsink" from the junk box and tested it today.

See pic's, looks real nice even if I do say it myself :-).

Everything works fine, 40Vdc supply, 2Amps: no heating and at 3 Amp's the heat starts up as expected but a bit less than the others due to the bigger heatsink.
With a small fan blowing at the unit (not mounted on the heatsink) everything stays cool, I then slowly crancked up the current setting to 4.5 Amp!!! for about half an hour, no problems, no bang or smoke just a very,very hot motor.

The single diode in the second pic is there to protect for the reversal of the supply wires, it will then just blow the fuse without further damage.

There are now 2 designs finished and tested, all the results have been posted here, my conlusion is that this chip is a good one but the unit needs some cooling at or above 3Amp, wich is normal when 6-7 Watt's have to be dissipated, even 13 Watt at 4Amp....

This is about as much as I can do, don't see anything more, there are some minor cosmetic changes noted for eventual new revisions.

Side note: the TB6560 works fine here within it's limitations but this chip and the 2 designs are a lot more powerfull and can even compete with commercial drives at a much lower DIY price.

Luc.

[lihaijiang]

lihaijiang,

I was always wondering why it was mentioned in the tips document but they were not installed on the evaluation PCB.

Luc.

Dear Luc,

Due to the volatage and current of the evaluation PCB is not very big,so it is unecessary to use it.

By the way, you really did a very good job.

[James Newton]

I agree... Luc does fantastic work. Wait until you see the version he did for me with the chip laying down off the edge of the PCB. I will be making those available very soon, and I'm looking for some people to try them out as beta testers.

[lucas]

Had an idea last weeked for the mounting of the LM317: it needs a heat sink for higher input voltages and mounting it vertically with a separate heatsink attached looked silly when the entire unit is mounted on a heatsink.
Attached are a few pic's of the alternative solution.

I changed one of my prototypes to the "new mounting style" as shown in the pictures, the tab of the LM317 is connected to +5V so isolation hardware is required.

Used an other motor, I had to readjust the motor current, after running the drive for 20 minutes I noticed that the drive and motor were extremely hot, I used the scope to check waveforms and voltage levels and finally found out that I made a mistake in "head-calculating" the Vref voltage for the current setting and the unit was actually running at 5 Amps @ 40VDC iso of the 4Amps wich I needed. (chair)

This little "accident" once again proves the performance of this chip, it didn't blow or shutdown.

[James Newton]

Had an idea last weeked for the mounting of the LM317: it needs a heat sink for higher input voltages and mounting it vertically with a separate heatsink attached looked silly when the entire unit is mounted on a heatsink.
Attached are a few pic's of the alternative solution.

I changed one of my prototypes to the "new mounting style" as shown in the pictures, the tab of the LM317 is connected to +5V so isolation hardware is required.

Nice! It does require a rather long heatsink. I still think +5 power should be supplied by the BOB or a PC or USB supply.

Used an other motor, I had to readjust the motor current, after running the drive for 20 minutes I noticed that the drive and motor were extremely hot, I used the scope to check waveforms and voltage levels and finally found out that I made a mistake in "head-calculating" the Vref voltage for the current setting and the unit was actually running at 5 Amps @ 40VDC iso of the 4Amps wich I needed. (chair)

This little "accident" once again proves the performance of this chip, it didn't blow or shutdown.

So we can advertise the new kits as supporting 5 amps now? COOL! ,o) Just kidding.

[lucas]

I used a matching heatsink for the prototypes only.

For a real machine there will be 3-4-5.. units on a single alu plate together with the BOB, the plate will be used for mounting and heatsink at the same time, reducing costs and simplify cabling using cable guides etc.. on the same plate.

My PCB machine using the TB6560 was made like this and also had the spindle driver and a small psu on it.

Alu sheet is inexpensive compared to real heatsinks and you need a mounting plate anyway, a little fan creating some airflow in the cabinet and all heat issues are solved.

[James Newton]

Anyone interested in Luc's excellent PCB and a kit of parts may now order from this page:
THB6064AH Stepper Motor Driver Kit. massmind

More details, including schematic, board layout, drill guide, and full assembly instructions are found at THB6064 MassMind Stepper Motor Driver Kit

[lucas]

A little clarification for new readers and for those which didn't follow both threads:
I designed 2 "full option" versions of this drive and tested them extensively, all details are on my website: http://users.skynet.be/ldt/CNC%20ele...THB6064AH.html and can be ordered there, just drop me a mail with what you want.

A third version has also been developed for James as an addition to his line of products with some design differences for compatibility with that product line.
See the link in the previous post for the details of that design.

James and I have been working intensively together during the whole design process to obtain the best quality/economical drive for DIY people.

All 3 versions are available now but where's the initial animo for a great DIY price/performance drive gone?

Is there a need for assembled and tested drives iso DIY kit's? PCB's only?
Everything is possible, just ask.

[lucas]

A few updates:

Made a new layout based on feedback for the flat version a while ago, I'm running out of pcb's and had to reorder. The new layout has mainly cosmetic changes but also an improved layout and some improvements on component placement.
They are on order and expect them in a few weeks.

- Ground planes are much bigger now.
- Reduced the standing resistors to 3 iso 6 and changed the pitch for the 3 remainig one so that a resistor array can be used also.
- Space between the 2 current resistors is widened for easier soldering.
- Component renumbering.

Attached is a silk PDF.

I looked extensively into the heatsink for the traditional version and came up with a real cheap and nice solution this weekend using a CPU heatsink, below is a picture of the result, more details on my website: from extremely simple to almost professional for a DIY version. I'm really pleased with the result. , The connector and mounting holes on the rear side are perfectly accesable.

Please remember that this is a larger layout wich is better than the flat one,
it has multiple power planes, amplifiers after the optocouplers, 4 iso 2 CS resistors etc...

Doesn't it look great?

[Duzzy]

"All 3 versions are available now but where's the initial animo for a great DIY price/performance drive gone?

Is there a need for assembled and tested drives iso DIY kit's? PCB's only?
Everything is possible, just ask."

Hi all,

Because this is still fresh in my mind here is a response to the above.

I only found your controllers after buying a blue board 6560. I'm not smart but I got it working first time. Thought some proper set up info would be good and found the thread about getting this type of board to work.

Oh great.... I've been suckered in again.

Why?

I was looking around the net about my longtime dream of a DIY CNC, thought I'd do an old scanner, dvd burner sort of thing.

Google showed me fantastic DIY machines and I got carried away.

The plug and play appearance of the ebay item looked ideal. (4 axis controller, motors, etc)

Also blue boards seem to have distributors everywhere so I could purchase in the UK knowing I wouldn't get hit with taxes. (I turn off non UK results).

Maybe you should offer more options on ebay. '4 Axis Controller' would have got my attention (I never even searched for drivers), the fact that there are 5 boards would have been fine. Especially if you had some info telling me why using separate boards might be better.

If you did kits for 3/4/5 axis in various combinations you could start to be more visible in the listings. (All I see is blue ones and red ones so they must be ok....)

Could you do a complete newbie 'plug and play' pack (pre assembled with some motors, psu) at a reasonable price?

Just my thoughts, anyway I've found your stuff now so I know where to go when the 6560 packs up.

All the best,

D

[vegipete]

Anyone interested in Luc's excellent PCB and a kit of parts may now order from this page:
THB6064AH Stepper Motor Driver Kit. massmind

More details, including schematic, board layout, drill guide, and full assembly instructions are found at THB6064 MassMind Stepper Motor Driver Kit

I have just finished assembling one of these kits and spun a motor last night for the first time. Very nice! 24 volts could get the little NEMA17 motor spinning around 2000 RPM when half-stepping (though there isn't much torque at that speed.) The driver stayed cold, there was some warming of the motor. The motor wasn't connected to anything and generally only drew 60 - 70 milliamps, although when decelerating the current briefly rose to around half an amp. Overly aggressive accelerations stall the motor - no surprise there.

I'll post more results as I learn more.

[vegipete]

Everyone like pictures. So here are two. The first shows the board nearing completion. I tipped the current-set pot up at an angle so that screwdriver access was easy.

The second picture is the completed board with a small stepper motor connected. Power and PMinMo cable are disconnected.

The heat sink comes from a Pentium something or other. There's a fan that clips onto the fins if temperatures really get out of hand.

Enjoy!