Hi Guys,

Thanks for reading on but I have a techy question for anyone who can help. I'm using some 40Kg servos to drive a hydraullic rig, direct onto the valveblock. However, I need some system or assembly which I can use to trip the system up if one of the control servos fails. So in essence, I need something which will be triggered should the servos physical position be different from the signalled one.

If anyone has any thoughts i'm all ears, so far all I have come up with is using a second smaller servo of the same speed which would shadow the main servo carrying some sort of electrical microswitch, which should they become out of sync' would trip the switch...but that's all so far.

H E L P ! !

thanks in advance

The second servo with limit switches is a cute idea. Mind you it does assume that both servos transit at the same speed. Or is the motion slow enough that the trasit times of the servos are simply not a factor?

A more elegant way to do it would be to have a pot or other "position sensor" on the arm or device and use a programable controller to compare the reading at the position sensor part to the position commanded by the rest of the system. Or, assuming your present control system is programable, to simply add in the routine to sense the arm position.

Originally Posted by bmki

I'm using some 40Kg servos to drive a hydraullic rig, direct onto the valveblock. However, I need some system or assembly which I can use to trip the system up if one of the control servos fails. So in essence, I need something which will be triggered should the servos physical position be different from the signalled one.

within a servo is a position sensor (e.g. potentiometer) used to compare the desired position value determined from the input signal. This comparison produces an error voltage that drives the motor. I'm sure there's some delay, but that error voltage should be close to zero during normal operation and will not be zero if there is a problem.

Of course, if there is a problem in the input signal path, then that error voltage will not be zero. A 2nd servo might be used without the motor to compare its position sensor to the same input signal used by the 1st servo.

Both are great points, which is why I asked! Inside the plastic casing is pretty well a mystery to me but I'm assuming that the pot in suggestion is on the output shaft? I think perhaps I should buy another servo and pull it apart....So just to clarify, the pot on the output shaft is presumably a 3 wire device which would show output if it wasn't tracking correctly?

Originally Posted by bmki

So just to clarify, the pot on the output shaft is presumably a 3 wire device which would show output if it wasn't tracking correctly?

yes it's on the output shaft. There should be a max and min voltage applied to two of the terminals, and the voltage on the middle terminal will be relative to its position. It doesn't indicate how well it is tracking. It indicates the position. If you know what the position should be, and have the means of relating this to a voltage, you may be able to determine if the output shaft is tracking correctly.

here's an article describing the motor control loop, but i don't see a description of how the control voltage is determined from the radio pulse

how familiar with electronics and control loops are you?

greg

Sadly I'm not electronics savvy beyond a 20 yr old physics class! However I'm reading as fast as I can to get the understanding. What I gather so far is that the output shaft pot will give a voltage which varies depending on it's variation from the signalled position? If this is the case I should be able to make a small circuit via a transistor and relay to operate as a safety circuit when a specific disparity between signal and postition occurs. I'm sure somebody out there can correct my assumptions?

thanks again clever folk!

Originally Posted by bmki

What I gather so far is that the output shaft pot will give a voltage which varies depending on it's variation from the signalled position?

not sure what you mean by signaled. It is not the error relative to the desired position. It represents the mechanical position of the shaft.

Originally Posted by bmki

If this is the case I should be able to make a small circuit via a transistor and relay to operate as a safety circuit when a specific disparity between signal and postition occurs.

maybe you should consider using a comparator. Here's an introductory web page. You should also look at the lm339 circuit.

i'm a bit concerned about your limited knowledge and the load on the device (40 kg == 88 lb). This doesn't sound like a toy,

greg

ok, my first failed assumption (you know what they say about assumptions!). My assumption was that the output shaft pot only output a voltage when it differed in position from the the position signal sent. Of course when thinking further i realise the servo uses that pot output to maintain it's output shaft position.


You are correct, it's not a toy. I've been using big servos on hydraulic valveblocks on my timber crane (like a hiab / loader crane) The system has a stop button which the operator can hit should a problem arise, however the inspectors are now demanding self regulation of the system, such that if a servo fails and doesn't return to centre the system selfs trips into a safe state (closed hydraulic flow). It just means that if a servo breaks with the crane moving that it'll stop itself.

Originally Posted by bmki

Of course when thinking further i realise the servo uses that pot output to maintain it's output shaft position.

i hope you also realize that the error signal you suggest is available within the servo. It is a pin 6 of the lf411 comparator shown in the attached circuit diagram.


But I assume that the fail-safe circuit needs to be independent of the control circuit. Ideally it duplicates the path from input device (controlled by a person) to the position of the device being controlled.

greg

Hey Greg, I'm Ben btw, thanks for your assistance so far.

What I currently have is this;

40kg servo operating each valve on a hydraulic valveblock, these (there are 6 in total) are controlled via a 2.4ghz receiver with operater input from a modified JR PCM9X (made a little more robust).

The safety circuit is a hydraulic solenoid which once energised allows the oil to flow through the valveblock discussed above, upon breaking the safety circuit via pressing the 'stop' button it closes the oil flow off and renders the crane static.

What I need to achieve is this; should a servo fail /sieze / stop moving during use, the operator would return the control to neutral /centred position (normally closing that valve off). As the failed servo will not move due to fault, I need a circuit to recognise the disparity between operator control signal and servo movement / position and generate an output which I can used to break the same safety circuit. I'm confident I can bring the output signal up to a level which will control my existing safety circuit, but getting this signal from within the servo is above my understanding.

I hope this helps?

I should also mention that these are digital servos, I've only just realised this looking for the if411.

XQ Power XQS5040D.

Ben

ben
this is certainly an unusual topic for this forum, and not one to take too lightly since safety is the issue. i am no expert in mechanical servo control and am only providing very basic information.

i suggest you create and post a diagram showing the components of the safety circuit and how it fits into the existing system. the diagram/design can be reviewed to help you make it complete (correct and unambiguous) and better understand the issues.

greg

a little envelop style, and the hydraulic path is a little simplified but it covers enough of the main elements.

With the safety switch circuit closed, the oil solenoid is energised, this permits oil to flow from the pump to the hydraulic valve (B).

If the TX lever is moved in 'N', the servo remains in 'N' neutral position & then the RAM doesn't move. Oil returns to the tank.

If the TX lever is in position 'X', the servo moves to 'X' controlling the hydraulic valve to move the RAM in the direction of 'X' - when the TX stick is returned to 'N' then the RAM movement stops - conversely with the stick moved to 'Y' the RAM moves in the oposite direction.

If the safety switch circuit is open (by pressing the stop buttoon), the oil solenoid is not energised and oil flows straight back to the tank and rendering the crane to be static. (A)

This open circuit also occurs should the TX signal be lost (fail safe)

--------------------
Currently for example;
...the operator moves the TX lever - servo to 'X', the RAM then travels in direction 'X'. The operator returns the TX to 'N' to stop the RAM but the servo fails and remains in position 'X'. The RAM would continue in it's travel until the operator hit the 'STOP' button. This would occur once the operator had realised there was a problem and thus rendered the circuit open via the 'STOP' button and stop the crane.


What I'm looking to do is remove the operator from this situation. I would like a system or signal which recognises when the TX lever position does not correspond with the servo position and thus trip the safety circuit to it's open state.

not sure if this is any clearer?

ben
it's clear that you need a fail-safe circuit that is reliably triggered when one or more servos is not responding to the radio input. At this point, I don't believe it matters what the servos are controlling.

we've discussed using a 2nd servo as a sensor to determine a position error, and you've suggested using a "transistor and relay to operate as a safety circuit". I'm asking you to post a diagram of the fail-safe circuit based on your understanding so far.

greg

You should be able to accomplish this with an off the shelf PLC (programmable logic controller).
You will likely have an Emergency Stop (big red mushroom switch) loop wired into the PLC as well...std practice.

Very common industrial controls application.