X axis problem change in direction

[sowjoe]

I'm looking for the simplest way to build a gantry that travels back and forth on the X axis only without it being hooked up to a computer. It will be belt driven and controlled by an 18v dc drill motor.

I was thinking about using an "on-off-on" switch at each end of the gantry slide to reverse the direction of the gantry, but I think the motor won't last long if I do that. I need something to slow the motor at each end before it reverses direction.

I've thought about Hall sensor's, 555 timers, limit switches etc., but they seem complicated being that I don't know much about them. Just want something simple.

Any idea's?

 

[betwixt]

The simplest method would be switches a little way in from each end that short out a resistor in line with the power when the gantry is between them. At each end the short would be removed so the resistor could limit the current.

Brian.

 

[sowjoe]

The simplest method would be switches a little way in from each end that short out a resistor in line with the power when the gantry is between them. At each end the short would be removed so the resistor could limit the current.

Brian.

I thought about using a switch to cut the motor at a certain point and using the inertia of the carriage to carry it to a reverse switch, don't know how well that would work.

Or, using reed switches to shut the motor off and another to reverse direction.... what do you think?

 

[betwixt]

Assuming the motor is fixed and the carriage travels on the belt, it would be difficult to do that. It becomes more complicated when you have to restore the power as well as reverse it. Simply changing polarity with change-over switches is easy but the wiring gets much more complicated when you have on/off switches in line as well because the inertia of the carriage will push it away from the switch that restores the power needed to turn the motor on again. It is still possible to do but requires more switches and more wiring.

The reed switch idea is good but they are normally only rated at very low power so as a position sensor they are fine but to control the current you need a switch with higher rating. How big is this thing and is at a school/home project or to be built comercially? If it isn't too big, the easiest solution may be to use the gantry to interrupt a light beam when it gets close to the ends of it's travel.

Brian.

-

 

[sowjoe]

I've thought about this over night and came up with 3 ideas....

1) Use 2 motors instead of one. When the carriage gets to one end it turns off one switch (motor A) and at the same time turns on another switch (motor B).

2) Use 1 on-off-on switch and one motor. Install the switch on the gantry. The switch will reverse polarities on the motor when it reaches each end of the gantry. To smooth the transition from forward to reverse I could use a spring.

3) Install (bolt) a "knob" to the belt to "push" the carriage along the gantry. Have 2 blocks on the carriage (one on the inside top another on inside bottom). As the belt turns the knob contacts the top block and pushes the carriage to one end, then the knob rides around the pulley and comes into contact with the bottom block and pushes the carriage to the opposite end. This eliminates the need for switches.

What do you think?

 

[betwixt]

It's good you are using 'lateral thinking' - often I come across students who have one fixed idea in their mind and they do not consider alternatives, even when their idea is an obvious non-starter. I recently dealt with someone who had a simple machine with a simple problem they just couldn't see a way around. It was an air operated piston that travelled in or out of it's cylinder by about 5mm, they had to use it to operate a small electrical slide switch which was about 1 metre away, also displacing 5mm and moving on a parallel axis. They spent days buildng experimental linkages with push rods and pulleys but couldn't overcome the slack or flexibility of materials over that distance. They became so determined to make it work they stopped looking at the obvious solution which was to move the switch to the piston and run wires back to it's original location!

Back to your issue, I wouldn't recommend using springs as cushion brakes because they store the energy they absorb and may release it in a manner you can't control, also the motor has to drive harder as the spring compresses. Don't forget that if you turn the power to the motor off and then load it's connections you can brake it electrically.

Using the top and bottom of the belt to carry it back and forth is an option but the gantry itself has to decide which side to grab and there may be consequences of height difference between forward and backward motion. Is there also a support beam to bear the weight of the gantry or is it suspended entirely by the belt?

Can you explain what this gantry is doing and exactly how much control you need over it.

Brian.

 

[sowjoe]

It's good you are using 'lateral thinking' - often I come across students who have one fixed idea in their mind and they do not consider alternatives, even when their idea is an obvious non-starter. I recently dealt with someone who had a simple machine with a simple problem they just couldn't see a way around. It was an air operated piston that travelled in or out of it's cylinder by about 5mm, they had to use it to operate a small electrical slide switch which was about 1 metre away, also displacing 5mm and moving on a parallel axis. They spent days buildng experimental linkages with push rods and pulleys but couldn't overcome the slack or flexibility of materials over that distance. They became so determined to make it work they stopped looking at the obvious solution which was to move the switch to the piston and run wires back to it's original location!

Back to your issue, I wouldn't recommend using springs as cushion brakes because they store the energy they absorb and may release it in a manner you can't control, also the motor has to drive harder as the spring compresses. Don't forget that if you turn the power to the motor off and then load it's connections you can brake it electrically.

Using the top and bottom of the belt to carry it back and forth is an option but the gantry itself has to decide which side to grab and there may be consequences of height difference between forward and backward motion. Is there also a support beam to bear the weight of the gantry or is it suspended entirely by the belt?

Can you explain what this gantry is doing and exactly how much control you need over it.

Brian.

I often find in the things I invent that the initial start of an idea is almost always complicated, over engineered, too much stuff. After I draw it up I begin to simplify it, trying to get it to its most basic state. Once I achieve that, I build a prototype and test it.

I've learned over the years how to look at problems. In this case the problems are the ends of the gantry (reversing the motor smoothly). I can kick myself for not learning electronics when I was younger. I know it's never too late to start, but the field is so vast these days I wouldn't know where to begin. So that leaves me with talking to nice people like you. People who will take a few minutes out of their busy day to help someone else.... and I want to thank you for that Brian.

Brian:
.... I wouldn't recommend using springs as cushion brakes because they store the energy they absorb and may release it in a manner you can't control, also the motor has to drive harder as the spring compresses.

Joe:
I agree, but if you vary the compression strength of the spring it may work (but it's a cheesy way of doing it).

Brian:
Don't forget that if you turn the power to the motor off and then load it's connections you can brake it electrically.

Joe:
I didn't know that. Turn the motor off? I know nothing about electronics (kick, kick).

Brian:
... but the gantry itself has to decide which side to grab and there may be consequences of height difference between forward and backward motion....

Is there also a support beam to bear the weight of the gantry or is it suspended entirely by the belt?

Joe:
The belt with the knob bolted to it would be centered between four aluminum beams (2 on each side of the carriage). The beams are approximately 4 feet long. The carriage is basically a box with 4 rollers attached to it, the box rides left and right along the aluminum beams. It's the same as CNC machines they use in woodworking, except I don't need the Y and Z direction just the X direction.

Brian:
Can you explain what this gantry is doing and exactly how much control you need over it.

Joe:
The gantry will roll along a table while the carriage moves back and forth, from one end to the other. Speed can be adjusted by the drill trigger switch or some other controller. I hope that answers your question. That's about all I can say. If this thing works I want to leave the option open for a patent.

Thanks Brian

 

[betwixt]

Thanks for the compliment and detailed explanation.

The braking is a basic feature of most DC motors. They work (almost) equally well as generators, if you rotate them they produce electricity. The braking comes from using them in generator mode with an electrical load attached, this might be nothing more than a resitor across the motor wires. Do a simple check to see if your motor exhibits this characteristic - disconnect the power from it and see how easy it is to rotate by hand, then short the connections out and see if it gets harder to rotate. A short circuit is the ultimate load but at the speed you can turn it by hand there should be no damage. If it doesn't work then either the motor isn't suitable or there is some control circuit in series with it which is not allowing current ot flow outwards toward the wires.

I appreciate that you don't want to reveal too much about the idea but I have to ask another question: If it simply moves side to side continuously while traversing the table, it's locus is a diagonal zig-zag, it this want you want or should it step forward, sweep, step forward, sweep and so on, with 'square' sides in it's motion?

Brian.

 

[sowjoe]

Brian:
... disconnect the power from it (motor) and see how easy it is to rotate by hand.

Joe:
At the moment, this endeavor is in the preliminary stage or "throw around idea's" stage. I will put that on the "to-do" list once direction is determined as how to proceed.

Brian:
.... then short the connections out and see if it gets harder to rotate.

Joe:
How do you "short the connections?" (kick, kick) What do I use to short the motor, the 18v battery?

Brian:
If it simply moves side to side continuously while traversing the table, it's focus is a diagonal zig-zag, is this what you want or should it step forward, sweep, step forward, sweep with 'square' sides in it's motion?

Joe:
There's 2 ways you can do it, both have problems...

1) Step and sweep.

2) Slow continuous movement while the carriage sweeps back and forth.

The problem with the "Step and Sweep" is the "jerkiness" of the step because the carriage is in constant motion. So the step would have to be quick thereby causing a jerk motion when it advances.

The "Slow continuous movement" problem is timing the advancement of the gantry with the side to side sweeping of the carriage. This should be easy to fix by adjusting the speed of the sweep with the cordless drill trigger or adjusting the advancement of the gantry.

 

[chuckey]

the way I see it it is to have a limit microswitch at each end, physically "inwards" a bit. There is a double pole motor supply reversing relay, a timer (555) and a supply on/off relay. So the logic is :- Supply reversing relay, de-energised. supply on/off relay de-energised. Motor trundle off, actuates end micro switch, this actuates the supply on/off relay so motor de-energises and coasts, also energises the supply reversing relay which latches and the timer. When the timer has timed out, the supply on/off relay de-energises and the motor goes off in the opposite direction. When it actuates the other end micro switch, de-energises the supply reversal relay, energises the supply on/off relay, triggers timer, when timer times out, supply on/off relay energises and the motor starts off again. The micro switches must be actuated by a piece of material with an angled front edge and a length longer then the coasting distance of th motor. The timer(s) must have a time period greater then the coasting time of the motor.
Frank

 

[betwixt]

I can see we are talking cross-purposes here. In your first post you referred to "an 18v dc drill motor" which I assumed to be just the motor itself but in post #9 you refer to adjusting the speed with the drill trigger. I assume this means you are using the entire drill unit and it has an adjustable speed control. It makes a big difference because the speed control board will almost certainly stop you using dynamic braking. It may also try to regulate the speed according to the trigger position (to cater for varying load) and that would make an attempt to slow it with springs pointless as it would simply 'push' harder to overcome their resistance to compression.

At it's simplest, your drill probably has a reverse switch which you could extend out to the gantry, either by linkage or by extending it out with wires. When it reached the end of the gantry it would push the switch to go in the other direction and vice versa.

What I meant when I suggested shortng it out is the actual wires feeding power to the motor itself should be linked together. Obviously, this isn't possible if there's a speed controller in line with them unless you open it up and make a direct connection. If you try that, please make sure there is no chance of the 18V being turned on at the same time, especially if you are using a battery as the source because pressing the trigger could cause a high current to flow and potentially damage the battery or even explode it.

Brian.

 

[sowjoe]

I think what I'm going to do is build a test model using one on-off-on switch. The 2nd on will reverse the polarity of the dc motor thereby reversing the direction of the carriage. Simple and easy. If I have a problem with reverse-direction smoothness (which I don't think I will because of the friction of the belt, motor and other things) then I'll try something else.

Here is the belt drive design I want to use....

https://bell-everman.com/products/linear-positioning/servobelt-linear-sbl

Video.... https://bell-everman.com/products/linear-positioning/servobelt-linear-sbl

 

[betwixt]

I'm always cautious of things described "compares favorably to high-end linear motor drives costing thousands more" which kind of implies it costs thousands anyway. I'm afraid my Internet isn't fast enough for video but I did gather it works by applying the motive force in the moving gantry itself. Wouldn't it be easier to make the gantry ride the belt and drive it from one end.

Would your idea lend itself to other methods of driving, for example being screw driven or pulled by a draw wire which might be easier because it utilizes rotation directly without having to turn it into linear motion before use. I've used screw drives in machines up to about 3 metres width before but there is no technical reason why it shouldn't go further. At 3m length the gantry could be positioned to about 0.1mm accuracy and it used a controlled acceleration and decelleration to ensure no time was wasted. It was a long time back but from memory it could go from end to end in around 1 second, fast enough that we had to warn visitors not to be frightened when it lurched toward them!

Brian.

 

[sowjoe]

I thought about a screw drive but they're too slow and too expensive, especially for a prototype.

It would be easier to drive it from one end. It would also make the carriage lighter and a lighter carriage would mean a smoother transfer (less inertia) when moving from forward to reverse. The option is still on the table.

A one motor design with a reversing switch at each end is probably the simplest design and easiest to build. But installing the motor on one end would scatter the wiring from one end to the other. I wanted to keep all the wiring, switches and motor in one place (carriage).

How high would the learning curve be if I were to use a micro-controller to control everything?

 

[betwixt]

A one motor design with a reversing switch at each end is probably the simplest design and easiest to build. But installing the motor on one end would scatter the wiring from one end to the other. I wanted to keep all the wiring, switches and motor in one place (carriage).

I'm not sure that logic works because you still have to get power to the carriage and the power cables would have to be fairly thick to carry the motor current while still being flexible enough not to impede the motion. In industry there are two methods of doing this, one is to drop the wires from above, central over the travel area so the average flexing is as small as possible, the other is to roll the cable out from one end in a flexible tube.

Again, more ideas come to me -

1. Is it possible to use the guide rails to carry power? If it is, the motor on the carriage may be a good solution.
2. Is it possible to fit something on the carriage that interrupts a short light beam (~5mm)? I'm thinking of a protruding 'flag' that can be used to send notification of position back to a controller box. As the carriage passes certain points the light beam would be broken and a controller would then know where it was and could use that information to control the speed or reverse it.

Incidentally, the screw method can be very inexpensive, nothing more than a few 1m lengths of threaded rod welded together. One end is attached to the motor shaft, the other sits in a ball race. The carriage is then fitted with a nut that moves up and down the thread as it rotates. Being a fixed distance per rotation, as long as there is a known start position and something to count the rotations, it's possible to calculate the carrieage position without any wiring beyond the fixed motor itself.

I write controller software all day so I'm probably biased but in truth a machine as you describe is very easy to control. The microcontroller software is easy, it's more of a problem to switch the power to the motor but how difficult depends on the motors characteristics. I'm going to guess the motor draws about 1 Amp in motion, about 10A to get it moving and has a stall current of about 20A so you would be lookng at a H-Bridge driver with overcurrent protection. That isn't anywhere near as complicated as it sounds, it can be done in a few inexpensive components. You will see lots of references to these on Edaboard but don't confuse this with H-Bridge circuits in power supplies and inverters which have different requirements.

Brian.

 

[sowjoe]

I'm not sure that logic works because you still have to get power to the carriage and the power cables would have to be fairly thick to carry the motor current while still being flexible enough not to impede the motion. In industry there are two methods of doing this, one is to drop the wires from above, central over the travel area so the average flexing is as small as possible, the other is to roll the cable out from one end in a flexible tube.

Quick post....

Remember, this is a cordless drill. The 18v battery is already hooked up to it, no wiring necessary.

I'll finish answering your post later today, don't have time at the moment.

Thanks

 

[sowjoe]

....so you would be lookng at a H-Bridge driver with overcurrent protection.

Brian.

I looked up what an H-Bridge driver is and it seems like it can be used as a reversing switch. I don't see the benefit of this. The thing I like about using on-off-on switches is that I can move them if I ever had to (i.e. carriage travels 36 inches instead of 48 inches) and, for the motor to reverse it has to go through the "off" position of the switch, which means there is no current going to the motor at the moment the reverse switch kicks in. The "off" would act sort of like a momentary coast phase or "slow-down" phase for the motor. Again, I know nothing about H-Bridge driver's so I really don't know what I'm talking about.

Incidentally, the screw method can be very inexpensive, nothing more than a few 1m lengths of threaded rod welded together. One end is attached to the motor shaft, the other sits in a ball race. The carriage is then fitted with a nut that moves up and down the thread as it rotates. Being a fixed distance per rotation, as long as there is a known start position and something to count the rotations, it's possible to calculate the carriage position without any wiring beyond the fixed motor itself.
Brian.

You would have to get an ACME thread of around 2 or 3 in order to get the travel speed up to what a belt can do. And, when thread sizes are that low the thickness of the rod increases thereby increasing weight and price. I understand, you can get higher count thread sizes but you'll have to increase motor speed in order to increase thread speed. I like belts because they're lightweight, easy to use, easy to install if they break, easy to clean if debris gets in them and quiet. Threaded rod you have to keep greased/ oiled... higher maintenance, more noise, heavier.

 

[betwixt]

I think this boils down to the degree of precision you need. I respect that you don't want to give much information away.

There is nothing fundamentally wrong with using the drill as a source of motion but it's speed will be difficult to control and it will drop as the battery discharges. I may be wrong but I'm also wary of the idea of cutting the power by releasing the trigger and expecting it to coast to a controlled stop. I see the simplicity of what you are proposing but I think in real operation it will have problems you are not fully considering. I visualize someone finding the battery goes flat with the carriage half way across and then the "@@$%%! spare battery pack isn't charged up so I'll have to wait an hour before I can resume" scenario. There is also the potential problem of the drill model becoming obsolete and the mountings having to be re-designed to cater for a new type.

Personally, I would still go for a system where the motor was fixed and reversible. It has more overall versatility and can be permanently powered for reliability. H-Bridges can be used in diferent ways, the simplest is diagonally opposing switches to allow the output polarity to be reversed but you can also turn all the switches off to isolate the output (your coasting mode) or turn both top or both bottom switches on together to short the output and provide dynamic braking. With care you can rapidly (thousands of times per second) switch between these modes to give a controled output power to set the motor speed.

Brian.