UIROBOT Stepper Drives
 

Hey:D
This is Tiger Huang from Shanghai, Product Manager of Stepper Motor Controller. Our Product Engineer Doctor Zhu who study and work in the USA developed a size coin, which is only 1.665*1.665*0.53 inch Stepper Motor Controllers with Microstepping, RS232 or CAN (protocol).
http://www.uirobot.com/userfiles/240_06.jpg

Welcome communicate with us tiger@uirobot.com Thanks!

This does not look like it would be of much use to a Mechmate builder.

Tiger, we need to drive at a minimum of 3 Amps and 40Volts - can you do that?

http://www.uirobot.com/robot-archive-a-344.html

Yes they can.

Well, to be fair, they claim they can. 8A 40vdc max. Sites in chinese, but some of the specs are in english. If they can send me 6 for testing..... :)

Edited to add: The drawings indicate they are meant to me mounted on the back if the NEMA motors. This would present some redesigning of the electronics.

Me too me too.. I will test them for FREE :D
Please send them to South Africa Mr. uirobot I will test them on a Mechmate in the HOT South African conditions.

I will take the 8A version..

Translation of the web-page

UIM24002/UIM24004/UIM24008 is a series of ultra compact high-performance stepper motor drives. Its greatest features a small volume, driving skills. Plus the corresponding flange, be fixed directly to the 42/57/85/110 of stepper motor. Its thickness less than 14 mm. UIM24002 can provide 0-2A adjustable peak current; UIM24004 can provide 0-4A adjustable peak current; UIM24008 can provide 0 ~ 8A adjustable peak current. Its high speed current compensation feature, be compensated high-speed rotation back-EMF. This series of drive use 10V-40V DC power supply.

Absolutely we can.
Welcome contact us tiger@uirobot.com for specific information.
Using our CAN protocol stepper controllers, you can control at most 112 motors at a time. Who can tell me how many axis?:p

You know Chinese? Great:D
Our English version website is under construction. Our Product Engineer Doctor Zhu is going to open an branch company in the US.

I know HTML and Google translate. I'm rather hopeless at Chinese. I can correct poor english though, if you're looking to hire a final editor. Or a controller tester... ;)

I think the key question is how to integrate these CAM or RS232 protocol drivers into Mach3 or EMC2.

Tiger, we normally use four motors on three axis. The X axis uses two motors. In some cases, five motors are used for four axis.

Can you tell us more about how these motors are interfaced to a (standard Windows) control computer, please?

and yes whats the current pricing :)

There are a lot of factors that are hiding in the background. Most of us use the Geckodrive products, particularly the G203v, which has 2000 steps per rotation, automatic current reduction, LED indicator (yellow) that shows when the drive is pulling too much current, and fail-safe operation that protects the user from self-inflicted self-destruction.

What does this drive offer that the Geckodrive does not offer?

If their input is CAN (Commercial Area Network) OR RS232 then you can't use them with any Step & Dir signaling. I have seen several of these type drives out there but they take custom control software that talks to the drives serially and sends them motion commands rather than S & D.

It's hard to design the solution when you don't understand the porblem (:-)

TOM Caudle

Had a quick glance at the specs,

0~40V operating voltage
There are 3 models, 0~2A, ~4A & ~6A.
Selectable peak current,
current compensation (??).
Selectable 1, 1/2, 1/4, 1/8, 1/16 micro steps, (i.e. 3200step/rev.)
uses step, dir, Ena.
Opto input

I'm bilingual so hope this quick summary will help someone.

I don't think you need to know how it interface with Windows, really... who can do that???

I received more info from them in pdf format. They have 3 types....CAN, serial and parallel. Just like Gecko they can also have STEP DIR and ENABLE pulses from your BOB.

Attachment 9167

Attachment 9168

Attachment 9169

Attachment 9170

Price for UIM24008 is 120USD/PC and price for UIM24004 is 80USD/PC under 10pcs.

I am sure the 4 Amp model at US$ 80.00 will do for me....if I go that way

Sounds good to my bank too :)

and about electrical noise? and wire lenght?

I do see some advantages, by mount the driver on the motor, we will eliminate any potential noise within the control panel, just wire the DC power & smaller shielded signal wires from the bob to the motors.

Cheaper cable cost as we can use cat 5 instead of those expensive big 1mm2 shielded power cables... the power cable can be a 2 core 1.5mm2 normal cable....

Maybe we don't even need the cat5 for the signal to the motor...

Maybe we can build (smaller) dedicated power supply units & install them right next to each motor....

I think the CAN would be the Controller Area Network protocol that was started by the automotive industry and which has very quickly become the dominant network protocol technology there.

Ken,
Don't forget thought that yes you move the drivers out of the control box, but now those low voltage step and direction lines are now big antennae and if not shielded well, they will pick up loads of noise and cause you much grief.

Heath,
So some CAT5 cable should sort that out? Still a big saving on the cable.

Cat 5 is not shielded cable. It is only twisted pair. I would not like to count on that to keep 5v logic level signals free of noise. Shielded cable would still be the way to go for that.

My mistake. Then low cost shielded microphone wire should do the job then.

Let's review some basics.

1. On most systems, a PC is connected to a controller through a parallel cable. To keep electrical noise to a minimum, the power supply in the PC is grounded to the PC's case and the case is grounded to the main ground lug on the controller. Usually ribbon cable is NOT used to connect the PC's parallel port to the controller because of noise problems, instead a good quality, shielded round cable is used.

2. Most controllers use a break-out-board to amplify the signals from the parallel port. On most systems, either the break-out-board or the actual stepper driver uses opto-isolation on the step/direction signals. The length of wire between the break-out-board and the stepper drivers is usually less than 36 inches. All of that wiring is inside a metal case that is grounded to the master grounding lug.

3. Most systems use a stepper power supply that is isolated from the computer's power supply. Only the grounds are connected, so noise from the stepper power supply will not affect the computer's power supply or the break-out-board's power supply.

4. The cables connecting the stepper-drivers to the stepper motors are high quality shielded cables with the shield wire connected to the main grounding lug on the controller.

That type of system is robust and reliable. Electrical noise problems can be controlled by proper grounding. If all inputs to the break-out-board are optically isolated and all outputs from the break-out-board are optically isolated, you will have done just about everything that you could possibly do to minimize most forms of electrical noise.

(Look at the schematic of the G540 mother board to see why it is so reliable. All signals are optically isolated from the outside world. The schematic is posted on the Geckodrive forum on Yahoo groups.)

If you mount the stepper drivers directly onto the stepper motors, in my opinion, you will introduce problems into the system. The low-voltage step and direction signals will have to travel a long distance between the controller and the motors. Those signals will be passed between the controller and the stepper drivers on cables that lie in a cable path that is only an inch or two away from the cable that powers the router or spindle. The little 5VDC signal going to the stepper drivers will have to compete with the large 120VAC or 240VAC signal going to the router or spindle. Remember that an AC signal is measured using RMS, so a 120VAC signal is really 1.414 X greater, or 170V peak to peak. A 240V signal is 340V peak to peak. Also remember that a 5VDC TTL signal is considered good if the low level is less that 0.7V and the high level is greater than 2.2V, so the actual signal range can be as small as 1.5V! What chance does 1.5V have against a 340V peak to peak spindle signal?

There are ways to ways to handle that situation. If you put optically isolated line-drivers on both ends of the step/direction cable and then use shielded twisted pair cable to carry the signals, you will have a better system. If you use fiber optics between the controller and the stepper drivers/stepper motors, you will have a better system. Both methods are complex and possibly very expensive.

For our CNC routers, having good quality stepper drivers mounted inside the controller that are optically isolated from the break-out-board, is the least expensive, most reliable system available to the average user who hasn't spent most of his life tinkering with electronics. Mounting the stepper driver directly to the stepper motor can introduce a lot of noise problems that will not be easy to fix.

Ken, I mentioned a "windows" computer way back in an effort to express the idea of a standard PC that could be running Linux and EMC2, or Windows and Mach3, as opposed to a dedicated machine control computer of one kind or another. This would be particularly relevant in a CAN bus system, which would likely be very robust if you found a way to get it working with our current GCode engines.

Now that it's clear that there is an option for a step/direction interface, how to connect it is obvious. How to connect it reliably is less obvious, as Mike points out. It strikes me that one straightforward solution would be to acknowledge that most of our machines have TWO umbilicals running to the Y car via different routes, and so relocating either the high voltage or the low voltage lines onto the dust collection tube would reduce the noise problem. Another cute idea is to move the stepper power supply onto the gantry. You're still going to want to star wire the power to the drivers, and this will save on the wire lengths.

In terms of competitive cost, we're still at 4 * $80 + ?? for a Bob = $320+ for a set of these, supplying 40V * 4A = 160 watts per motor. Compare that with G540 at $300 for a Bob and 4 drivers supplying 50v * 3.5A = 175 watts per motor.

I was hoping to be able to demonstrate that a CAN bus interface would be practical for our application, but just like with USB, the latency is killer. CAN bus can run up to 1Mhz, and a worst case message takes 134 bit times, or 134 microsec. Assuming a clean CAN bus constantly sending signals to four motors, we're at a cycle time of 134 * 4 = 536 microsec = about 2,000 updates per second. That would be not quite 200 IPM on a 10 microstep ungeared motor. Too slow for bit twiddling, and once you move beyond bit twiddling, the integration with Mach or EMC gets rather complex.

These drivers are an interesting solution, and show the potential of the new breed of driver chips ( See also Allegro ), but they aren't yet a complete solution for our application.

Mike, Brad, Thank you so much for elaborating. Really helps. Before this I only have a faint idea of what to do with our MM electronics, now, I'm getting a clearer picture of what I'd done correctly on my control system. Still need more time to digest the information fully before I throw out more stupid questions.

Thanks again, really appriciate it.

Ugh, and I made two errors. One, the step signal needs to go on and off for a full cycle, which is two updates, and two, Nyquist sampling limits, which are also approximately 2 to 1. So in my hypothetical CAN bus case, it's actually less than 500 updates per second, which would be under 50 IPM on an ungeared motor, and closer to 10 IPM on our common configurations.

Hey guys, Mounting directly on the motor?? What about heat buildup? Wouldn't the driver and stepper be frying each other on a hot summer day in the shop? Or, did I just ask a stupid question?:D

It's a good question! Anyone have a theory?

How about Don't run your motor too hot...