Best way for 2 motors one axis

[rhfurniture]

Hi,
I have a (much modified) PRT benchtop with the 4G board. I am considering adding a second Y (moving table) motor to improve stability as I do a lot of heavy hardwood cuts, and I am rebuilding the frame to accommodate 1200mm/48" a la buddy48. As I want to use the spare (5th) gecko slot for a second Z in the near future, and my X2 is already running my second gantry carriage (X axis) motor:
1. would it be practical to run the 2 motors off the one gecko driver.
2. what would be the performance disadvantages of doing so (as opposed to 2 motors, 2 drivers), and how could I best deal with it - (eg upping the amp output from the shared gecko)?
3. Has anybody tried this?
4. If it is possible, any tips re cabling, connecting etc
5. Is it dangerous - eg if one motor lost a step and not the other, could damage be done?

R.

[gerald_martin]

Ralph: I do not know if you can run two motors on one driver - does not sound like a good idea to me. Shopbot has available a 4G board with a plug for a remotely mounted additional (6th) gecko - I had one of these on my previous PRT96 with dual y motors and dual z's. Might want to find out if you board has this plug-in or if you could upgrade to this board and transfer your existing gecko's. The output channel on the additional port is selectable, just like on the 5th gecko slot.

Best wishes.

Gerald

[richards]

Ralph,

Only one motor per driver. Even the best motors have small variations in the windings which would eventually fry the motors or the driver.

Use two drivers and, if necessary, a one-chip signal booster so that each driver gets strong signals. If you need help, send me an e-mail and I can send you a simple schematic that might help.

[dana_swift]

Mike, what difference would the subtle differences in the windings make, presuming the driver has the current and inductive drive capacity to drive both coils (of the same motor model) in parallel? Stepper motors "should" move the same angle with each step without regard to the coil winding. Or so I have been led to misunderstand..

And the driver needing "stronger" input signals? Isn't that handled by the LS-TTL fanout rules? (I believe that is the logic family used on the bot control board, that should be verified tho.)

If there is any problem it is that when the system powers up the orientation of the motors is unknown, and when the coils for the "two" motors are energized, one might try to move 180 degrees in phase from the other one. That would load the drivers with a bind for the duration of that run. How severe that turns out to be would also need to get a little study. Probably with some clever design that problem could be overcome.

It wouldn't be the first time I have been undereducated, and always welcome a better understanding.

Thanks

D

[rhfurniture]

Thanks, I will work on doing it with just 1 motor, and improve the mechanics. Just thought I'd better ask in case it was an OK practice that I didn't know about. It was just that my 2 motored gantry carriage works really really smoothly. However you look at it 2 drive pressure points must give a steadier motion than one. Maybe 2 pinions on a shaft belt driven by one motor. Or a really rigid guide rail set up for horizontal stability, to supplement the vertical rails and rollers already there. Part of the problem is that the rack is positioned tooth up, (as all prt bt's), gathering dust and greasy grott, so I will work at turning that first.

Thanks for replies,

R.

[richards]

Dana,
The question was asked on the Gecko forum. Mariss, the designer of the Gecko stepper driver, gave an elaborate explanation why two motors can not be connected to one stepper driver. I haven't been able to find that post today, but I'll keep looking.

Depending on the stepper driver being used, the fan-out of one TTL type signal cannot feed more than one input on the Gecko. The G201 and G202 stepper drivers SINK 15mA of current. The G203v requires 2.5mA of SOURCE current. On a data sheet, the item IOL (current, output low) shows the maximum current that an output can SINK. The item IOH (current, output high) shows the maximum current that an output can SOURCE. A standard TTL (not LS) ouput can usually SINK 16mA or SOURCE 0.4mA. Many people see that 2.5mA spec on the G203v and automatically assume that several G203v drivers can be driven from one TTL output. They don't realize that the G203v stepper driver requires a SOURCEing signal.

[dana_swift]

Mike- thanks for the info. I downloaded the G203v manual to see what was going on, the answer is the inputs are connected to opto-isolators, all inputs sharing a common ground.

My hunch is they did that because the typical user doesn't understand digital logic where sinking current is the more powerful state. They specifically note in the manual that the common rail is NOT +5 volts which would have an advantage of better TTL compatibility, but most folks would have been quite confused.

I agree with you that standard LS-TTL can only drive one gecko input directly, adding a pull-up to +5 would probably allow driving two.

As to the paralleling of motors.. and "why not". It would be because of the output is a constant current drive. Load sharing may be unfeasible with bi-polar drive. If it is unipolar, a set of diodes per motor winding (one for load sharing, one for quench) would make load sharing workable.

Ralph- I yield to Mikes wisdom. He is correct even though it sounded unusual. I have to agree with him that you need to use two gecko drivers and some form of current sourcing output booster between the control board and the drivers due to the optical isolators.

Be aware that the stepper motor phase-split problem can occur with this driver arrangement also. (Perhaps worse)

Ralph- keep us posted with your results. One experiment is worth a thousand expert opinions!

Mike thanks for the extra info and steering me to look at the driver documentation and specifics, it just isn't a standard logic interface. I'm better educated now, which is all I can ask!




D

[richards]

Dana,

Unfortunately, the LS-TTL chips can only sink between 4 and 8 mA, so they are not suitable for use with the Gecko stepper drivers. The 'L' in LS-TTL refers to low power. There are a lot of economical buffer type chips that deliver about 25mA per pin that can be used as easily as TTL, so finding a chip that works is not a problem. For those that don't mind a little extra effort, small transistors can be substituted.

[rhfurniture]

Mike,
Are you suggesting that I can take the output from one set of driver connectors on the G4 board and with a little electronic boosting run 2 drivers mounted somewhere else in the control box? This might be an alternative to Geralds option above, as I do not see any obvious connector for this on my board)(though I havn't yet stripped it right out to check fully).

R.

[richards]

Ralph,

Yes, that's the way that signals are "buffered" and "amplified". In the TTL world (transistor-transistor-logic), one output signal can usually drive up to ten input signals on similar devices. So, a 7400 output signal could normally be connected to ten standard 74xx TTL inputs. A 74LS00 output could be connected to ten 74LSxx inputs, and so forth.

In the case of the Gecko G201 and G202 stepper drivers, the inputs act as if they were roughly equivalent to ten standard TTL inputs, so only one Gecko can be driven from a standard TTL output. Driver chips have outputs that can sometimes handle 2x or even 4x the current of a standard chip, so sometimes a buffered chip can drive more than one Gecko input. However, even the buffered chips have a problem. The total current that the entire chip can handle at one time is limited. It's not unusual to see a buffered chip that has eight 25mA outputs (200mA total) specified as a 100mA max chip. In that case, if more than four outputs were active LOW at one time, and if the devices connected to the buffer chip drew full current, the chip would be damaged.

That's a long explanation to describe why you might need to add external circuitry to a device.

How you add the circuitry is up to you. Normally, I try to find chips or components that use the same voltage levels as the original device. By doing that I can connect the grounds together and the VCC (ISO standard designation for the supply voltage to an integrated circuit) together. One output from the original driver connects to two or more standard inputs on the auxiliary device. The auxiliary device takes that one input and uses two or more output signals to drive two or more Gecko input signal lines. An added advantage of handling multiple motors with an auxiliary buffer/driver is that you can easily invert the polarity of one of the motors so that one motor turns CW while its opposing motor turns CCW (like on the X-axis when you use two motors). If the direction signal is not inverted, then you have to wire the coils differently. Wiring coils so that motors run in the opposite direction is not hard, but it's just one more thing to track and trouble-shoot when things go wrong.

One other question that sometimes gets asked when people learn that an opto-coupler or opto-isolator is used is why can't you just change the resistor on the LED side of the opto device so that the LED only draws half as much current so that one TTL output could drive two opto devices? The main reason that you don't want to do that is related to speed. The more current that an LED can pull from a circuit, the brighter it is, that also means that an LED that is turning on, reaches a threshold brightness faster. That threshold brightness is the point that the photo-detector that is watching for the LED to change states, knows that a state has changed and that it can safely change the polarity of its output. The Gecko stepper drivers are normally rated at 250,000 steps per second, so they like a strong current source for the opto devices.

The Gecko G203v is different from the G201 and the G202. It expects the output signal to SOURCE current (remember that the G203v uses a common ground signal). It only requires 2.5mA of current to operate, but that is about 6x more current than TTL chips can SOURCE. Driver circuits for G203v stepper drivers commonly have driver chips that can either SINK or SOURCE about 25mA of current per output. If that's the case, then you could connect two G203v inputs to the same driver output without the need to build any additional driver circuitry.