PDA

View Full Version : Opinions wanted on stepper driver



wierdscience
08-04-2008, 11:58 PM
Collecting up parts for an over-winter CNC plasma table build and I was wondering what's wrong with these drivers?

http://www.probotix.com/manuals/SideStep_manual.htm

$35 ea.,are Gecko and Larken drivers really that much better? These just look like to good of a deal to be true.Anybody have experience with these?

MTNGUN
08-05-2008, 12:45 AM
Only 2.5 amps and 32 volts ? A gecko is 7 amps and I don't remember the voltage rating but a lot more than 32 volts.

I'm not qualified to critique the design, but 2.5 Amps/32 volts is not a lot. Perhaps enough for your plasma table, I dunno.

The drivers, power supply, and motors need to be compatible. Select the motors first, then get a power supply and drivers that meet the motor's specs.

John Stevenson
08-05-2008, 03:36 AM
Wierd, There are plenty of these drives around based on this Allego chip but they are low power drives.
At 2.5 amp and 24 volts max which is a good supply limit they are on full power and to be honest they are stretching it.

What you get with the more expensive Gecko and Keling drives is piece of mind and some protection from shorts and overloads.

Gecko is supposed to be bringing a new drive out called a G250 which is a small cheap stand alone going up to 3.5 amp and 50 volts.
Can't find any release details on it but their G540 which is 4 of the 250's mounted on a breakout board is listed here.

http://www.geckodrive.com/product.aspx?c=3&i=14469

May not fit the bill at $299 but it does get you 4 drives and the breakout board plus simplified wiring.

I did hear that the separate G250's were going to be about $35 each but don't quote me on this.

.

Evan
08-05-2008, 05:13 AM
I'm using the Xylotex drives which use the Allegro chip. I removed the heat sinks and the crappy thermal tape and clamped on BIG heat sinks using Arctic Silver thermal paste. I then added a good fan to cool the lot. I am running them at 29 volts from a !very! stiff supply and VERY fat wiring and have the current control on the boards maxed out. Not one bit of trouble from the drives and the performance is excellent but if the fan is turned off they go into thermal shutdown in 30 seconds. Before anyone starts to comment that this is pushing it to the ragged edge and not really a good idea consider that this is how your computer's cpu is run as a matter of standard practice.

motorworks
08-05-2008, 05:50 AM
I have been running Larken drivers for over 5 years on my mill
No problems (knock wood!)
eddie

Just Bob Again
08-05-2008, 06:06 AM
Depends on what you got for motors. A reasonably modern stepper is a lot more efficient than the old ones. The ones I use are about 650 Oz-in and only need 2A at 6v. You don't need a million microsteps. They make things smoother but the motors are not very linear position-wise in microstep mode. The drivers you mentioned are small and reasonably priced. Not industrial-duty but should be fine if you don't make any mistakes. They will fry in an instant if you make a wiring error or disconnect a motor with the power on.

wierdscience
08-05-2008, 09:30 AM
Ya,I noticed the diffrence in power output,but I'm not building a router or mill for that matter,still more is better might end better in the long run.

I would never make it as a woman,I hate shopping:D

John,yes I am also going to look at Ke-Ling:) their KL-6050 looks promising at $60/pop,also weighing build a power supply verses buying one.

What Evan says is true also,I noticed computer towers are happy doubling as toaster ovens so long as there is ample breeze blowing.

Okay,so mixed reviews it is.I was thinking of buying one their kits,no fuss no muss right?Well maybe not,anyway guy down the road built a decent 4x4' router using this one-

http://www.probotix.com/3_axis_stepper_motor_driver_kits/3_axis_monster_stepper_motor_driver_kit/

After seeing it make some fairly hefty cuts in wood and noticing how much drag his gantry has built into it I was thinking the same should work well on a better gantry design moving at a slower pace with much lower inertia to overcome and nowhere near the forward resistance.

This guys gantry works,but it's no beauty queen.Plain b-7 althread for the screws,plywood for the frame and the linear slides are flat bar(aluminum) and grooved out skateboard wheels,yet he gets reasonable accuracy out of it.His gantry weighs a ton and the whole mess sits on furniture levelers.

lazlo
08-05-2008, 09:44 AM
Wierd, There are plenty of these drives around based on this Allego chip but they are low power drives.
At 2.5 amp and 24 volts max which is a good supply limit they are on full power and to be honest they are stretching it.

Those are all single-chip drivers based on the Allegro 3977, which has a max rated output current of 2.5A at 35V. But the datasheet has a stern warning no to exceed a Tj of 150C, which I'd figure is a package temperature of around 125C.

http://www.allegromicro.com/en/Products/Part_Numbers/3977/

The bigger stepper drivers like the Gecko's have external power FETs (i.e., they're not a single-chip solution), so they can drive much larger motors. Hence the higher cost.

Come on Marriss -- step up to the plate :D

Evan
08-05-2008, 09:58 AM
The data sheet for the Allegro chip also has some waffle wording about the absolute maximums, especially the 2.5 amp figure. Basically it says as long as the maximum junction temp isn't exceeded it can be run higher than 2.5 amps. That is apparently the case by my experience.


ABSOLUTE MAXIMUM RATINGS
at TA = +25C
Load Supply Voltage, VBB ............. 35 V
Output Current, IOUT .................. 2.5 A*
Logic Supply Voltage, VDD ........... 7.0 V
Logic Input Voltage Range, VIN
(tw >30 ns) ..... -0.3 V to VDD + 0.3 V
(tw <30 ns) ........... -1 V to VDD + 1 V
Sense Voltage, VSENSE ................. 0.5 V
Reference Voltage, VREF ................ VDD
Package Power Dissipation,
PD................................. See page 3
Operating Temperature Range, TA
(A3977Kx) ............ -40C to +125C
(A3977Sx) .............. -20C to +85C
Junction Temperature, TJ ......... +150C
Storage Temperature Range,
TS ......................... -55C to +150C
* Output current rating may be limited by
duty cycle, ambient temperature, and heat
sinking. Under any set of conditions, do not
exceed the specified current rating or a
junction temperature of 150C.

BTW, the Allegro chip can handle significantly larger loads if synchronous rectification is disabled and external flywheel diodes are installed. Such diodes only cost a few cents each and take the load off the internal H bridge.


Disabled Mode. When the SR input is logic high,
synchronous rectification is disabled. This mode is
typically used when external diodes are required to
transfer power dissipation from the A3977 package to the
external diodes.

Ryobiguy
08-05-2008, 12:28 PM
How quiet are the steppers with the Allegro driver chip? Allegro says it's nice and quiet (of course,) but I ran across this thread at CNC Zone where Mariss openly talks about the development of the mini-Gecko:
http://www.cnczone.com/forums/showthread.php?t=51083

Quite a long thread, I spent a whole week reading through the 200+ pages(!!) Very interesting to see a new piece of stepper technology being created. BTW, the ~$30 "target price" listed on post #1 jumped to something like $50 or $60 for single volume pricing, which is a bit premium for 3.5 amps, but then so is the product.

On page 28 in that thread, Mariss says this:
A chopper (L297, Allegro, et all) has two free-running 20 - 25kHz oscillators. These oscillators phase-lock and break phase-lock hundreds to few thousand times a second. This "make, break" phenomena produces beat frequencies that fall smack-dab into the audible range producing the grunting, hissing, squealing and whistling everyone is so fond of.:-)

So how noisy is that Allegro drive? I'm guessing it's probably quieter than the previous generation of stepper drivers, but probably not as good as a Gecko drive.

-Matt

Forrest Addy
08-05-2008, 12:44 PM
How about this one.

P/N: CW250
INPUT: 20-60VDC
OUTPUT: Selectable up to 5A max.
STEP ANGLE: 1.8 deg:200steps/Rev.
Microstep: 1/8: 1600 steps/Rev.
Full bridge driver for hybrid stepmotors.
Switch Selectable output current.
Over temp & Output miswire protected.
Opto isolated Step pulse input (Clock input +5V Level).
Opto Isolated Direction input ( +5V Level)
Terminal strips in/out.
L: 5-3/4" W: 3-3/4" H: 1-3/4" WT: 1.5

No endorsment or reccommendation. Just something I stumbled across.
https://www.mpja.com/prodinfo.asp?number=17452+MS

lazlo
08-05-2008, 12:45 PM
The data sheet for the Allegro chip also has some waffle wording about the absolute maximums, especially the 2.5 amp figure. Basically it says as long as the maximum junction temp isn't exceeded it can be run higher than 2.5 amps.

That's the 150C Tj I mentioned in my post above. That's not waffling -- that's how most power devices are spec'ed: how high the junction temperature can get before the power transistor fries. There's also a reliability issue: the lifespan of the transistor is inversely proportional to the Tj.

Microprocessors, for example, are usually designed for a 7-year lifespan, so they're usually spec'ed with a much lower Tj: 90 - 95C. If you overclock the processor, the Tj goes way up, and the lifespan of the chip decreases proportionally.

Just remember that the internal junction temperature is a lot hotter than you can measure on the surface of the package. That's why modern microprocessors have on-die thermal diodes...

lazlo
08-05-2008, 01:01 PM
How about this one.

https://www.mpja.com/prodinfo.asp?number=17452+MS

That's the Chinese CW250. The datasheet is amusing -- it's a "Micorstepping Driver".

BobWarfield
08-05-2008, 01:13 PM
your plasma gantry should be light and require minimal torque. OTOH it needs to move the torch fast for a smooth cut so you can't gear down the motors much. Be sure you've put a pencil to that aspect.

Best,

BW

Evan
08-05-2008, 01:52 PM
That's the 150C Tj I mentioned in my post above. That's not waffling -- that's how most power devices are spec'ed: how high the junction temperature can get before the power transistor fries. There's also a reliability issue: the lifespan of the transistor is inversely proportional to the Tj.


That isn't what I mean. They are waffling on the current limit, not the temperature limit. First they say 2.5 amps and then they basically tell you how to circumvent that limit and say it's the temperature that really matters instead, which it is. 150 C sounds rather high to me.

wierdscience
08-05-2008, 01:53 PM
Yep,I have been looking at that.Most cutting occurs at or around 120-150IPM which works out to 300-350rpm motor speed.Most of the HSM/light industrial machines I have looked at top out at 500IPM in the range of thicknesses I will be cutting.I also want oxy/fuel capability which on thick plate can be down as low as 3-4 IPM.

Evan
08-05-2008, 01:59 PM
So how noisy is that Allegro drive? I'm guessing it's probably quieter than the previous generation of stepper drivers, but probably not as good as a Gecko drive.

It plays music nicely. http://www3.telus.net/metalshopborealis/76tbones.wmv

But, to make it that loud I had to clamp a pie plate to the carriage to act as a resonator and the spindle was off. You can't hear the steppers over the spindle drive.

lazlo
08-05-2008, 02:04 PM
That isn't what I mean. They are waffling on the current limit, not the temperature limit. First they say 2.5 amps and then they basically tell you how to circumvent that limit and say it's the temperature that really matters instead, which it is.

The Tj is proportional to the current flowing through the transistor. So if you liquid cool the chip, you could get more than 2.5A out of it without the Tj rising above 150C.


150 C sounds rather high to me.

It would be, for a microprocessor (digital) process, but the Allegro is an old chip (circa 2000) and those are giant feature sizes in old DMOS (analog) process technology.

Evan
08-05-2008, 02:07 PM
Darin,

It sound like this will do what you need. 4 amps, 50 volts and $20.

http://www.futurlec.com/StepperMotorController.shtml

Evan
08-05-2008, 02:19 PM
The Tj is proportional to the current flowing through the transistor. So if you liquid cool the chip, you could get more than 2.5A out of it without the Tj rising above 150C.

It's also proportional to the thermal resistance to the ambient without liquid cooling. That's why I replaced the teeny heat sinks and thermal insulator tape with real heat sinks and Arctic Silver.

http://vts.bc.ca/pics4/drivers1.jpg

lazlo
08-05-2008, 02:59 PM
It's also proportional to the thermal resistance to the ambient without liquid cooling.

I was just using liquid cooling as an example -- I should have said "liquid nitrogen" :)

So the Tj is going to increase the more current that's flowing, and Allegro tells you that it's going to die above 150C, so it's up to you remove the heat somehow.

If you immersed it in liquid nitrogen you could probably double the current output, but it'd be cheaper to buy a Gecko :D

studentjim
08-05-2008, 07:34 PM
Wierdscience, have a look at this site, that I came accross recently looks interesting.http://www.dtllc.com/products.html

wierdscience
08-05-2008, 11:02 PM
Wierdscience, have a look at this site, that I came accross recently looks interesting.http://www.dtllc.com/products.html

Thanks for the link,looks interesting,sooo many choices,sooo many decisions,sooo few dollars:o

lazlo
08-05-2008, 11:08 PM
Thanks for the link,looks interesting,sooo many choices,sooo many decisions,sooo few dollars:o

Those are Unipolar drivers Wierd. What kind of steppers do you have?

wierdscience
08-05-2008, 11:26 PM
Those are Unipolar drivers Wierd. What kind of steppers do you have?

That's the beauty of it,right now I don't have any,I'm still shopping.This would all be easier if I had $50,000 to just VISA my way into a machine:D

J Tiers
08-05-2008, 11:35 PM
Weesh.....

If you think THAT is "waffling", you don't do power circuit design.............

Just about ANY power semiconductor has at least two basic current limits.

1 is a long term current limit that is really a power limit based on a given case temperature, or possibly junction temp directly, and certain conditions of "on-ness" like gate voltage, etc. It is often an I^2 * R limit based on the Rds(on) for a mosfet, for instance..

2 is a "bonding wire limit", basically, over this current bad things happen because wire bonds to the chip cause trouble.


The result is "time-power" limits...... So long as you don't exceed the Tj, you can give the device currents etc of anything up to the bonding wire limit.

usually there are short-term limits, line single-cycle current limits, which for a 1A rated 1N4000 series diode may be as high as 30A.

For a Mosfet or IGBT, there may be current limits of 80A or more for a 20A nominal device. If you followed the jeep winch driver thread a while back, those limits were featured on the device datasheet, with spectacularly high pulse currents allowed.

They give you a curve of the time-thermal response, and YOU get to calculate your exact duty cycle vs this,, to decide on device suitability, heatsinking, etc.

Just another reason engineering is such an interesting field. Very little is fixed in stone, unless it is a safety agency requirement. Even then, for CE, they will change the rules regularly as a form of economic protection against imports......

studentjim
08-06-2008, 06:37 PM
Those are Unipolar drivers Wierd. What kind of steppers do you have?
Lazlo, look again there is a bipolar driver also
http://www.dtllc.com/1xcnc15bb.html