Using Servo Valves-How?

Hi! I have some questions I was really hoping you could help me with.
I am looking at using servo valves (Rexroth) to control hydraulic rams for a metal drilling machine.
The problem is I know nothing about them and need some help.
I will be using a Micrologix 5000 PLC and a Panelview.
The drill head needs to perform the following functions:
-From a HOME position, rapid advance ram to a FEED position.
-At the FEED position, advance at a slower feed rate suitable for the chip load of the tool.
-At two or more positions during the feed advance stroke, “PECK” or momentarily retract to clear the tool chip load, then resume advancing at the feed rate.
-At position DEPTH, rapid retract ram back to HOME.
-Upon return to HOME, I need to simulate a hydro-mechanical cushion so that the ram does not slam home and wear/damage the ram or the machine.
In the past, we would stall the hydraulic ram on an adjustable hard stop for full DEPTH. This stop is end-user adjustable for a number of reasons. How do I deal with that hard stop, whose actual position can be changed?
So, given all the above, here are my questions:
-Do I now only need a HOME switch for each ram?
-What encoder/LDT/???, do you recommend or have had success with? Keep in mind this is a metal cutting machine with flood coolant, so it needs to be durable in that environment. Also, do they make rams with the position indicator built in?
-What analog card(s) do you recommend? An Allen Bradley card? Or are there other brands that function easier/better?
-Are there other components/devices that I need that I did not mention here? Cables? Monitors? Etc???
And, as if that weren’t enough questions, I have another issue. All of my PLC programming has been using plain digital ladder logic. I know nothing about this analog stuff-including what type of logic to use. Yikes! If you could perhaps suggest a place to start to implement the controls I need above? Maybe even a sample of some ladder logic so I could study it and figure out how this is supposed to work?
Thanks for taking the time to read all of this, and thanks in advance for any help!

First, lets start with your choice of controller. The ControlLogix is a great controller, but its not nearly fast enough for hydraulic motion control using standard analog cards. There are motion controller cards for the ControlLogix so one option is to use those cards. An even better choice would be a Delta RMC controller, which will work hand on glove with your ControlLogix PLC (In fact it looks just like another AB PLC when put on your ControlLogix network). The RMC is super easy to program, and a profile like the one you have described will be very easy to set up.

Any good hydraulic servo system starts with a good hydraulic power unit design and system design. Screw that up and no amount of software is going to fix it. I suggest you go to
http://deltamotion.com/ and study the design resources there on designing a system that can actually do what you want it to do. They even have a system you can configure online and simulate.

Sensors: Use a magnetostrictive type sensor such as offered by MTS, Balluff, or Turck. The sensor mounts inside your hydraulic piston (external types are available) and gives superb resolution (up to 1 micron). It will be well protected that way. Choose one with an SSI output, which is a clocked output. The clocked output has the advantage that you know the position at a specific time, where with an analog device you have A/D conversion lag and don't know exactly where it was at a specific time. Since it is an absolute position sensor you can dispense with a home position switch and feed position switch.

Last, given what you have said about your experience level, you might be in over your head and this might be something you want to hire out.

Peter Nachtwey, president of Delta Computer Systems is a member of this forum and I'm sure he will be along to offer advice. Take it. He knows way more about this than I do and he and his people have helped us with what we perceived to be a difficult 6-axis control problem but which was probably pretty basic to them.

How many hydraulic actuators are you trying to control? It sounds like more than one but how many more?
Do they all the actuators need to drill at the same time or are they drilling different parts on different machines?

If you use a Temposonic or Balluff SSI rod you don't need to home the actuators because they provide 24 bits of absolute position with a resolution down to 1 micron. You didn't mention accuracy of the depth. I don't think it cost more for the extra resolution so get it if you can.

The motion controller take care of ramping up and down so the actuator doesn't need a cushion. The actuator can be programmed to ramp to home position without having to hit the end of the cylinder.

The depth and feed/drill speeds can be sent down to the hydraulic motion controller over Ethernet or serial from the Panel View. I STRONGLY recommend Ethernet/IP.

As TConnolly said, the hydraulic design needs to be right. It will make the setup of the RMC motion controller simple.

Here is a video with a few simple examples.
http://deltamotion.com/peter/Videos/PositionControl/PositionControl.mp4
The first one is similar to what you want to do because it changes speed in the middle of the motion. It doesn't do the 'peck' type of moves but they would be simple to add. Each 'peck' would be two more steps, the move backwards a small amount and the move forwards to where the actuator was. You would want to make the feed/drill speed slower than what I did in my example. You can also look at the programming for the RMC motion controller. It is a simple step based programming where a command is issued and then you wait for some condition to be met before going to the next step. Your example will be similar to the first program in the video with 4 extra steps, two for each 'peck'. There is a variable table where one can declare variables and access them as if the RMC is an AB PLC.

Now the cool part is the graphs and the ability to record exactly what is happening so you can check if the drill depth is actually achieved.

I'm pretty sure that you could do this cheaper and better using servos or even steppers. Maybe even linear slides??

TConnolly said:

First, lets start with your choice of controller. The ControlLogix is a great controller, but its not nearly fast enough for hydraulic motion control using standard analog cards. There are motion controller cards for the ControlLogix so one option is to use those cards. An even better choice would be a Delta RMC controller, which will work hand on glove with your ControlLogix PLC (In fact it looks just like another AB PLC when put on your ControlLogix network). The RMC is super easy to program, and a profile like the one you have described will be very easy to set up.

Any good hydraulic servo system starts with a good hydraulic power unit design and system design. Screw that up and no amount of software is going to fix it. I suggest you go to
http://deltamotion.com/ and study the design resources there on designing a system that can actually do what you want it to do. They even have a system you can configure online and simulate.

Sensors: Use a magnetostrictive type sensor such as offered by MTS, Balluff, or Turck. The sensor mounts inside your hydraulic piston (external types are available) and gives superb resolution (up to 1 micron). It will be well protected that way. Choose one with an SSI output, which is a clocked output. The clocked output has the advantage that you know the position at a specific time, where with an analog device you have A/D conversion lag and don't know exactly where it was at a specific time. Since it is an absolute position sensor you can dispense with a home position switch and feed position switch.

Last, given what you have said about your experience level, you might be in over your head and this might be something you want to hire out.

Peter Nachtwey, president of Delta Computer Systems is a member of this forum and I'm sure he will be along to offer advice. Take it. He knows way more about this than I do and he and his people have helped us with what we perceived to be a difficult 6-axis control problem but which was probably pretty basic to them.

Click to expand...

Ha! There is no doubt in my mind I'm in over my head-but then again, we've all been there before... Sink or swim.

I know we have some meetings with our hydraulics supplier tommorow, so I should also have a better understanding of some of this by then.

Thanks for the help!

Peter Nachtwey said:

How many hydraulic actuators are you trying to control? It sounds like more than one but how many more?
Do they all the actuators need to drill at the same time or are they drilling different parts on different machines?

If you use a Temposonic or Balluff SSI rod you don't need to home the actuators because they provide 24 bits of absolute position with a resolution down to 1 micron. You didn't mention accuracy of the depth. I don't think it cost more for the extra resolution so get it if you can.

The motion controller take care of ramping up and down so the actuator doesn't need a cushion. The actuator can be programmed to ramp to home position without having to hit the end of the cylinder.

The depth and feed/drill speeds can be sent down to the hydraulic motion controller over Ethernet or serial from the Panel View. I STRONGLY recommend Ethernet/IP.

As TConnolly said, the hydraulic design needs to be right. It will make the setup of the RMC motion controller simple.

Here is a video with a few simple examples.
http://deltamotion.com/peter/Videos/PositionControl/PositionControl.mp4
The first one is similar to what you want to do because it changes speed in the middle of the motion. It doesn't do the 'peck' type of moves but they would be simple to add. Each 'peck' would be two more steps, the move backwards a small amount and the move forwards to where the actuator was. You would want to make the feed/drill speed slower than what I did in my example. You can also look at the programming for the RMC motion controller. It is a simple step based programming where a command is issued and then you wait for some condition to be met before going to the next step. Your example will be similar to the first program in the video with 4 extra steps, two for each 'peck'. There is a variable table where one can declare variables and access them as if the RMC is an AB PLC.

Now the cool part is the graphs and the ability to record exactly what is happening so you can check if the drill depth is actually achieved.

Click to expand...

This machine will have 6 drill heads that fire off at the same time once the part is clamped. One dedicated machine, one part. Then they will all hit their feed rates, go to depth and come back on whatever timetable their particular tool/operating depth dictate. Only one tool will require the pecks, IIRC, its a bore something like 3/8Dia and 6" deep. The only other hydraulic on the machine are the clamps-and they will likely be just a regular spool solenoid.

I like the idea of not needing a home switch and an absoulute encoder-one less Prox switch to buy and wire should help offset the added cost of the servo setup.

Thanks for all the help-I'm surfing your website right now for more info as well.

On a side note-does anyone know how to make the post editor screen any bigger? It's like trying to read a document on a little phone screen.Between that and the bright borders on this fourm, my eyes are going haywire!

You have gotten some great advise so far.

Just to add to the info... you should know that you may well be able to use a high end proportional valve instead of a true "servo" valve. Manufactures have clouded the arena by sometimes using the term "servo solenoid" valve.

A traditional servo valve has a design called a "flapper nozzle". They can be made very dynamic but also require very clean oil and many contain screens that require extra maintenance to clean periodically.

A high end proportional valve or "servo - solenoid" valve is a high precision spool valve usually using a lapped spool and sleeve design with precision ground metering edges, on-board electronics to drive the solenoid coil and integrated spool position feedback which is used by the on-board electronics to ensure the metering opening is what it should be. These valves are able to tolerate dirtier oil more (but don't skimp on filtering!), and generally cost less and have lower lead times.

I spent about a dozen years at a company that made production drilling machines and my guess is you would do fine with a proportional valve. Since you mentioned Bosch Rexroth, if your flows are low, you could look at the 4WRPEH valve or if they are higher a 4WRLE valve. Your hydraulic supplier should know these. Linear "V-Type" spool cut is best.

I have a peck drill routine for a Delta RMC if you want it. I think I wrote it to back the drill out all the way on each cycle to clear the chips and then rapid back to just short of the previous drill position. Please PM me if you want a copy.

Last point I would make is on the feedback. Putting an SSI Balluff or Temposonics transducer in the cylinder is a great EXCEPT when you are making precise spotfaces, counterbores or countersinks. When you put the transducer in the cylinder, it is actually measuring the position of the magnet in the piston and not the end of the rod. We had a machine that had a .005" tolerance on spotface depth and a transducer in the cylinder. They would set it up in the morning and all was well. As the oil heated, the cylinder rod expanded and the spotface depth kept getting deeper and deeper until the oil temperature stabilized. No one believed me until I calculated the expansion of steel for the length of rod we had and it matched . We put an external SSI transducer on that measured the drill head position and no more drifting...

Good luck and if you go the Delta Route, their online training class is HIGHLY recommended to get you started in the right direction.

<disclaimer - I work for a distributor that sells Bosch Rexroth and Delta equipment>

Your application seems easy enough. The 6 actuators can be synchronized easily as TConnolly pointed out. He had to synchronize 6 actuators too.

Unfortunately the prox is a lot cheaper than servo valves but you must consider the extra precision will pay off in the long run.

Who is doing the hydraulic design?

Thanks for all the responses everyone!

ndzied1, you are spot on with your valve thinking. The 4WRPEH is what we seem to be looking at. Also your offer of some sample logis is very nice, and I will take you up on it-even if we do not go down this road, I will learn something new. I'll send you my E-Mail addy.

I don't do the hydraulic design-I just get to wire it all up. I do, however, get to pass my concerns on and try to back up the people who do the hydraulics.

Right now, Our hydraulic supplier is Motion Industries, and they have someone there who is helping us (His name is Tom, IIRC, last name starts with an F...). They put us in touch with Chuck Ziegenfuss over at engineering Technology Services, who seems to be the one person that we have talked to so far (present company excluded, of course!) who really knows about these valves and their applications...

I should have been a little clearer about the drill heads-they are not synchronized in any way, they all just get their start command at the same time, I only mentioned that because it does put a sudden drain on the system and I know if the pump/accumlator are not large enough, problems occur at that point.

I am sure that the prox is cheaper by miles, but then again, each head would have three of them (more for the peck cycle heads). Then there are the additional input cards, wiring, etc. Also, on the hydraulic side, we were using Vickers 3-position spool solenoids with a second 2-position valve to shift the head into feed rate through pressure/temperature compensated flow controls. All those solenoids and manifolds and flow controls took a lot of space, and the plumbing can be a nightmare-and take a lot of time to make and install. Whether or not servo-valves are actually cost effective with all this considered, I do not know, nor is it my decision. I believe the company owner is also trying to use this particular machine as a test bed to expand his machines abilities in the future.

Lastly, the videos you sent me, Peter, I could not view for some reason. I did, however spend some time on your website and have gleaned quite a bit of info there.

Thanks again for all the help!

VLC is the universal video watching application

Chicken said:

Lastly, the videos you sent me, Peter, I could not view for some reason. I did, however spend some time on your website and have gleaned quite a bit of info there.

Thanks again for all the help!

Click to expand...

I make my movies on a Mac Pro and export them in standard .mp4 file format. I make them compatible with iPads and most phones too.
Many older Windows systems don't support the play back of standard .mp4 files without a plugin of some sort.
VLC will play just about anything.
http://www.videolan.org/vlc/download-windows.html
You may need to download the file first to use VLC.