[HomemadeCody]

Hello everyone!

This is my first post! I'm pretty new to hobby cnc. I've programmed and operated them for years but this will be my first build attempt. I'm starting with the electronics portion of the build first just to make sure i can accomplish it.

I have three motors that i got for pretty cheap. The are Applied Motion Product 4023-828, Nema 23, 1.7v 3.3a. Here's the link 4023-828 | Stepper Motors | Applied Motion

My problem is finding a cheap driver that will work for these motors. Most of the cheap ones i find on ebay are only rated to 2.5a or 3a with 3.5a max Here's a link to one i've been looking at its a Toshiba TB6560AHQ chip. CNC 3 Axis Stepper Motor Driver Controller Board TB6560 - eBay (item 280590409851 end time Dec-16-10 07:38:16 PST)

Would this driver work or would my 3.3a motors fry it? And will a 24v 15a power supply work? Please Help!

Thanks!

[James Newton]

Since the driver powers the motor, you can always use a smaller driver with a larger motor.

I /strongly/ recommend against buying that cheap TB6560AHQ crap from China (HK). You can't repair it if you fry it, and if you are new at this, it is quite likely that you will fry a driver at some point. Second, they admit right in the listing that they will falsify the shipping documents... "We always send it as a "gift" and describe cheap value on shipping label." If they will lie /for/ you they will lie /to/ you. Google around to find all the people who are having problems with those type of drivers... replacing chips... "fixing" the design... just not worth it.

[HomemadeCody]

I see what you mean, I've read up on them a little bit more and all i hear is crap crap crap. So i'm thinking they must be crap. Can anyone recomend a good driver? I've heard good things about Gecko drives but they seam rather expensive at around $100 per drive!

[James Newton]

To help estimate how much power you actually need, check out this page:
Stepper Motors

If you really need more, the Geckos are impossible to beat /at that power range/. They will move about 90 lbs at 1000 IPM given a well built, low friction, axis. That is an estimate calculated using formulas the Geckos designer, Mariss, published here on CNCZone and on the Geckodrive FAQ:
Geckodrive Frequently Asked Questions FAQ under "Should I use servos or steppers in my machine?"

If you don't actually need all that power...

:Cough: Ahem... I happen to sell a kit that will drive that motor at 3 amps per phase 40 volts = 120Watts, and is well reviewed and fairly easy to repair.
SLAm Stepper Motor Controller, Allegro SLA7062M, Unipolar $34.95 / axis quantity 3. It will toss about 60 lbs around at 1000 IPM.

And if you think you can get by with 2 amps per phase, the Linistepper is VERY well reviewed, high quality, well supported, easily reparable, and will perform as well as a Gecko, at lower power settings:
Linear Stepper Motor Controller $30 / axis quantity 3.
It will move about 35 lbs at 1000 IPM.

But please don't take me at my word... ask around... see what has actually worked for other people. Most will tell you the sun rises and sets on Geckos. The few who have tried the Linistepper, and who didn't really need a Gecko, will tell you it works very nicely for a smaller load and costs half as much.

[leaveme]

James is right, Geco is a brilliant choice but it is a bit costly.

Never go for Unipolar drive unless your existing steppers are Unipolar. Go for PicStep or TBA6560 based driver if you need something cheaper. But do not buy TBA6560 based driver from eBay.

Just a side note:
Most Unipolar steppers can be modified to Bipolar (parallel) by changing winding connections and it's actually easy.

[James Newton]

Unipolar is underrated. It has it's advantages over bipolar. And bipolar comes in 3 flavors, again, each with advantages and disadvantages.

This page explains all of the options, and what is good and bad about each.
Stepper Motor Connection Options

Part of the bad on bipolar is the extra cost, both of the driver, and of repairing a blown driver. And the rapid drop off in torque with higher speeds.

[Mariss Freimanis]

I agree with everything you say James except for the "rapid drop off in torque" part. That just isn't so.:-)

Step motors are constant power output motors and power is torque times speed. This means when speed doubles, torque is halved to keep the product of the two a constant. Power output on the other hand is entirely a function of supply voltage and motor inductance, specifically voltage divided by the square root of the inductance. It has nothing to do with the type of drive, be it unipolar or bipolar.

The only way a drive can affect this relationship is by not meeting the 'special' needs of a step motor such as mid-band resonance compensation and not morphing to a full-step sequence at higher speeds. Each deficiency robs the motor of its potential and causes it to deliver less power than it could otherwise.

I have no intrinsic prejudice against unipolar drives. They do use fewer power devices than a full-bridge (bipolar) drive. Bipolar drives must have 8 power transistors versus the minimum of 4 power devices for a simple unipolar drive or 6 devices for a more complex unipolar drive. This gives a moderately significant savings over a bipolar, dual full-bridge drive. Secondly, unipolar drives are immune to a bridge shoot-through current failure modality because,.. there isn't a bridge.:-)

I personally don't design unipolar drives (haven't for 25 years) because they have a very significant technical drawback causing two major design comprimises.

1) A unipolar drive generates a winding voltage that is twice the supply voltage. This means the power transistors have to be rated twice the voltage as what is needed for a bipolar drive. An 80V bipolar drive can use 100V rated devices while a unipolar drive has to use 200V rated devices for the same operating margin. Power device cost is a function of rated voltage; if the devices are MOSFETs, a 200V device silicon has to be 4 times as large as a 100V rated part for the same current rating.

2) There is no good way to recover the unipolar motor returned motor energy when the voltage is twice the supply voltage. It has to be dissipated as heat which is wasteful and makes the drive hotter than it has to be. A bipolar drive has no voltage in it higher than the supply voltage. Returned energy is easily and efficiently regenerated back to the power supply; the returned energy is at the same voltage! The drive stays as cool and efficient as theoretically possible; energy isn't being wasted as heat.

3) Bridge-type drives (bipolar) by design clamp all external voltages to the supply voltage. That is why a bridge drive effortlessly survives motor disconnects that destroy unipolar drives instantly. The unipolar topology has no natural current shunt path back to the power supply when a disconnect anomaly occurs.

All these are considerations a designer takes into account when designing a drive. Unipolar and bipolar topologies have their advantages and drawbacks and you weigh them as a designer. My particular balance scale tilts firmly towards bipolar designs; its complexity is outweighed by the manifold advantages you gain in return.

Mariss

[James Newton]

LOL... Mariss, I agree with everything you said except for the /lack/ of rapid drop off in torque. It's right there in every torque graph I've ever seen from a motor mfgr. The torque /starts/ lower, but /stays/ more even in unipolar mode and /starts/ higher, but drops off faster in bipolar modes.

Look at the attached torque vs rpm graph for the PK266-02 motor from Vexta (and you can find the same sort of information for most motors). On the left is the torque for bipolar modes at two different voltages (the red lines, solid and dotted) from 0 to 500 rpm. On the right, the graph for unipolar mode, but the rpm is from 0 to 2500! Tricky huh? As a result of the different ranges, the lines have an apparently similar slope, but in fact, the bipolar line is MUCH sharper drop off than the unipolar line. You can see that when you equalize the ranges, which I did by taking the 0-500 part of the graph on the right, and resizing it over the graph on the left. Now the (distorted) green unipolar line is being fairly compared to the red bipolar line. If they had continued the bipolar line past 500, you would see the unipolar driver beating the bipolar with the same sized motor.

Unipolar torque is MUCH more consistent as the RPMs increase vs bipolar.

Now, at this point, a lot of people will say: But that is only because unipolar starts lower! And that is absolutely true, but the unipolar driver/motor combination also /costs less/ and so for the same money, you can buy a bigger motor and driver, and start /as/ high, then /stay/ higher as the rpms increase, with unipolar than you can with bipolar.

It may well be that your Gecko's, because of the way they transition to full step mode, would not experience as much of a drop off in torque as the traditional bipolar drivers used to test these motors for the graphs, and if I had the money for a setup, that's what I would buy (yeah, I sell Linistepper kits and I just said I would buy Geckos if I had the money) but for a small to medium sized system, where the cost of the motors and drivers is a real factor, going with a cheap bipolar chopper driver instead of a quality unipolar driver and a slightly larger motor is a mistake. Period.

As to your points against the unipolar driver electronics:

1. Yep. But for small to medium sized motors, the higher rating necessary isn't an onerous problem. 100v TIP122's just don't cost enough to worry about. You can fry your Linistepper drivers and replace them for a few dollars. I sell all the transistors (drive and secondary) for $6.

2. Yep, Liniseppers get hot and need a big heatsink or a CPU fan. But as you yourself have said, a lower wattages, the lost power and lower efficiency isn't a serious issue.

3. Yep, the solution is to not disconnect the motor while the system is live! And gain, if you do blow the drivers, replacing them is cheap. I think "repairability" is also an underrated advantage. People /do/ manage to still blow Geckos right? Even though I'm sure it's much harder to do and you cover that under warranty right?

So here is how I see it: If someone is building a small to medium sized CNC machine, and they need to limit cost by DIY, the Linistepper and unipolar motors can't be beat. If the machine is medium large to large, or they have the $$$ for the best, they should go Gecko. I think the dividing line is around 70 Watts of drive, maybe 80oz-in of torque to move around 30 lb of load; up to that line, the Linistepper is a viable and more cost effective option. Above that, Gecko or go home.

In NO case should anyone EVER buy one of those crap bipolar choppers from China, TB6560 or otherwise. LOL... I think we can agree on that point, right?

[Mariss Freimanis]

James, we have very little that we can disagree about.:-)

I still maintain my objection to your contention about torque fall-off. Your post didn't include the mentioned speed-torque curves; I'd be very interested in seeing such data.

Barring that, the only explanation for this dilemma may be you were comparing a six-wire motor being driven bipolar full-winding against a unipolar drive. Were that the case, the full-winding connection develops 4 times the inductance of a half-winding and torque at highers speeds will be one-half the unipolar result. Perhaps that is what you are referring to.

If so, a fair "apples to apples" comparison would be a bipolar drive running the motor half-winding versus a unipolar driver.

Mariss

[James Newton]

Huh... they dropped the image I was trying to attach. You can find it here:
Stepper Motor Connection Options about half way down the page, or at the vexta web site. The "fake" graph at the top of that page is generally accurate, but not from a real motor. The one labeled "PK266-02" about half way down is the real data...

While I'm not totally certain what type of bipolar driver (and that was my other point, bipolar isn't necessarily bipolar which isn't necessarily bipolar either, and newbies miss that) the mfgr used, I would assume it was one that makes their motor look best: Bi-Polar Parallel.

Given my limited understanding of Physics, it makes sense to me that it takes twice as long to slow, stop, reverse and build the magnetic field in the opposite direction in bipolar parallel than it does to just start and stop the field in one direction in unipolar. As you need to do that faster and faster, it becomes impossible to overcome the inductance of the motor on each cycle. Unipolar only has to increase and decrease the field strength, not completely reverse it. You would understand that far better than I... I'm just looking at what the mfgrs say and what I've seen in my own limited experience.

Again, I'm just talking about medium small motors... perhaps in larger motors the physics are different.

[Mariss Freimanis]

Found the graphs and it's as I thought. They are labeled "Bipolar Series" and "Unipolar".

This might help regards unipolar vs. parallel:

The magnetic field has to be reversed in both types of drives every 3.6 degrees. It is the inductance of a winding that hinders the speed of field reversal while voltage promotes it. The relationship is V / L and the result is measured in Amperes per second. Unipolar and half-winding (bipolar) have the same inductance. You use each half of the winding 50% of the time, I use a single half winding 100% of the time. It's the same winding with the same characteristics in either case.

Mariss

[RomanLini]

Found the graphs and it's as I thought. They are labeled "Bipolar Series" and "Unipolar".

This might help regards unipolar vs. parallel:

The magnetic field has to be reversed in both types of drives every 3.6 degrees. It is the inductance of a winding that hinders the speed of field reversal while voltage promotes it. The relationship is V / L and the result is measured in Amperes per second. Unipolar and half-winding (bipolar) have the same inductance. You use each half of the winding 50% of the time, I use a single half winding 100% of the time. It's the same winding with the same characteristics in either case.

Mariss

Hi Mariss, yep that's correct that a bipolar motor operated with half winding per phase has the same inductance and high speed power as a unipolar motor. However in that case it also has the exact same torque!

The advertised torque gains of bipolar over unipolar are because of being operated in bipolar series mode. Where (as you stated) they have 4 times the inductance and reduced high speed torque, although they can have as much as 40% greater holding torque.

The holding torque difference with unipolar can be improved somewhat by running the motor at slightly higher amperage. With microstepping drivers the total power dissipation is reduced from full step mode and the motor can be run at slightly higher current without exceeding the dissipation spec, and unipolar run over-current allows a torque increase that is not usually possible (or not as much) with bipolar that are typically pushing the limit of magnetic saturation at their rated current.

From my experience (as a designer of only unipolar drivers since 2000) a unipolar setup can be run at similar dissipations to bipolar series with typically only 15% to 20% lower holding torque, and once spinning has the unipolar benefits of much improved top speed torque.

I'd also like to clarify a point you made about unipolar drivers;

... 2) There is no good way to recover the unipolar motor returned motor energy when the voltage is twice the supply voltage. It has to be dissipated as heat which is wasteful and makes the drive hotter than it has to be.
...

Which is quite wrong. The twice supply voltage effect of a unipolar driver (sometimes called see-saw effect) is without energy costs as there is no current within the open circuit half of the winding that rises to V*2. It is not dissipated as heat. Unipolar drives handle current regeneration as efficiently as bipolar (and more efficient than bipolar series due to half the coil resistance). The unipolar current recirculation in chopper drives is covered quite well in Jones on steppers and the current regeneration is seen in the usual way with the PSU voltage being "pumped up" on motor deceleration.

However you are correct that the unipolar switching FETs must be capable of handling the V*2 see-saw voltage.

I still believe unipolar has a lot to offer in terms of ruggedness and simplicity and high speed performance. For situations where holding torque is not the most important aspect of a design unipolar can be a very good choice.

What many people don't realise (but I'm sure you do) is that a large part of the push towards bipolar in the last 10 years was from cost cutting. Once driver electronics became significantly cheaper the cost of increased driver complexity became secondary to the costs of additional motor wires and higher motor construction cost, larger connectors, and larger motor frame sizes (holding torque).

Bipolar drivers have their benefits and I don't doubt they will continue to dominate the market. There is also the benefit you have mentioned in other posts where a bipolar chopper driver can have current protection on the high side, protecting from wiring shorts. But it should never be assumed that "bipolar is always better" which is an attitude sometimes adopted by the less knowlegable posters on the forum, I'll chalk that up to fashion.

[dominicburke200]

can i just check from someone that i am understanding correctly.
if i have a nema 23 4A motor and want to use a 3A driver running at 2.5A (so not to max it out) it will work the motor but it will only be taking the dialled current of 2.5A and running at about 50% total possible power.
i just don't want to fry my driver board
Thanks

[James Newton]

Yeah. I can't guarantee you won't fry your driver board, but if you do, it will not be for that reason. And the power shouldn't be 50 less... If you are running e.g. 10 volts 2.5 amps, that is 25 watts of power vs 30 if you were running 10 volts 3 amps.

[dominicburke200]

thanks, i am cannibalising my laser cutter drivers i replaced to use on my diy router . i have released the magic blue smoke from one driver several years ago with incorrect wiring but have now learnt better. i just wanted to check i wasn't trying to do something very stupid and expensive. once i confirm it all works i will look at buying some more drivers. i like the gecko ones but they are expensive and difficult to get in the uk but am not keen on the ebay chinese boards either.

i calculated at 24v at 2.5A = 60w rather than the motor total is (lots Vmax) eg 24v x 4A = 100w
thats what i thought but just wanted to make sure that the low impedance of the motor wouldn't somehow break another driver but trying to pull too much current without enough resistance.

thanks i will wire one up tomorrow and see what happens