Posts by sixdegreesofconfusion

Quote from Alejandro

Hi all,
has anyone solved this problem? I have the same problem, I'm trying to acquire an analog voltage input with my DX200 controller but it doesn't work properly. It seems that the measurement with both channels has no sense.
I also have a JANCD-YCP02 board, someone knows with which M register is associated these analog inputs? I'm trying to perform the control of an external element so I need to handle the robot speed depending on the reading of an analog input.
Thanks

Alejandro
I was confused by the same thing for a long time. The key is whether or not the analog input is going into the YCP02 board or the YEW board. My understanding is the the YCP02 board is interfaced directly to the motion planning hardware (EAXA board). Being tightly coupled to the EAXA board allows the analog values (or channels) to be directly connected to the robot motion. Unfortunately this also means that these signals are not accessible within the programming and control environment (jobs/CIO). Conversely, the YEW board is interfaced into a different 'layer' of the control which allows access to the signals (via the registers) but prevents direct interaction with robot motion.

For one of my current projects I need a way to access a laser distance value via motocom and preserve the ability for the laser output to alter the robot motion. I am looking to split a 4-20mA signal between the YCP02 and YEW01 boards using two analog signal converters. Has anyone else had to do this or something similar?

YasMas,

Thanks for the info. I tried setting the parameter but it didn't work. It may be something on the particular DX100 I tried it on. Perhaps it's dependent or conflicts with another parameter elsewhere.

Does anyone have the equivalent parameter for a DX100? I only have S2C0575.

On the topic of parameters can we make a sticky where we compile a list and what they do? Obviously FD parameters are off the table. No one person may have a comprehensive list but collectively we can build something very useful to us all.

I am interested in using these dual line scanners to track a feature for a project l have. I'd like to know about your experience with them. Also, what is your exprience with the Digimetrix library? We are currently producing vision guided paths using a custom visual basic app and Motocom.

I have only investigated how to do active path contol, otherwise i have zero experience in this area. Based on what I've gathered the only way to do active position correction without discrete point operations is to use the ACORON function. This function is built around scaling analog voltage units into robot distance units. They have some nice features in the function as well which youcould set up as fail safes if your pc will be doing the work. It would be super fantastic if the ACORON function could look at user acessable registers or I/O but that does not seem to be the case. (Motoman, consider this a feature request.) Maybe you could output six voltage channels from your labview rig. Not elegant but it could get the job done.

Not sure the speed you are running at but if I vaguely remember that with regular ethernet on the DX100 it was taking 40 to 60ms to get or write a point. You might be able to do something with math operations on the points with data from variables or IO or using the SHIFT function. Our process requires us to scan the complete length before we cut so we can "look ahead" and evaluate the sampled points as a whole path.

Potis, where is the rconf parameter?

I think we have a somewhat similar situation. On two of our DX100s the R and the T can end up flipped when returning home even though the path is programmed with several taught MOVJs in sequence. The only way the R and T can be taught to a specific rotation is to carefully manage the RBT configuration. (Perhaps pulse values do not matter because we are using relative jobs?) The work is at nearly maximum reach so it is common for the wrist to get flipped from part to part. It's a real nightmare to program since we have to be very careful in planning and testing the path to avoid pulse limit errors.

potis, thank you for your tip, I'll compare our robots and see how this affects things.

We are developing a vision guided cutting operation using Motocom with 6 DX100s and we are seeing some unusual behavior with value overflows and periods where the network response becomes exceptionally slow. On four of the robots we are communicating via an EtherNet/IP card while the other two are using the on-board ethenet. I am wondering if there is any way if these two errors are linked and what can be done about it.

We are seeing the overflow when returning a double using BscGetVarData. We are using motocom32.dll version 5.3. BTW: Is there any reason to use BscHostGetVarData instead of BscGetVarData if you are not handling strings?

On the network side we have taken steps to limit the amazing amount of EtherNet/IP traffic generated by two network valve blocks. These two devices were responsible for nearly all of the network traffic. We have had a several instances where the robots have EtherNet/IP cards have had thrown major communication errors. Anyone else seen this happen on their DX100s?

Motoman offers an option called MotoEIP. The software allows the controller to work directly with Proface panels. It is a paid option. Never used it, I know that the software exists.

This thread is a bit old but I just wanted to leave a message for the developer. Thank you for creating this software. I have found it very useful for analyzing the ladder in our robots. I am curious about adding the simulation capability. The ability to confirm the operation of the ladder would be very useful.

That Yaskawa is even asking such a question of it's user base gives me hope.

"Features" I'd like to see:
[list type=decimal]

• Documentation

• Willingness to work with customers to find workarounds and solutions not covered by the documentation

Between 1 and 2, and places like robot-forum.com, most functionality I might want will present itself in due time.

If I had the opportunity to specify the robots in our operation Motoman would not be my first choice for two reasons, the quality of documentation, or lack thereof, and Yaskawa's absolutely maddening culture of secret knowledge. I have never seen a company where one has to beg for information on what options are available to purchase. So far my experience has been that the people who want to help (i.e. Motoman North America) cannot get the information you need because the people who can help them (i.e. Yaskawa Japan) are unwilling to share much of anything outside their own inner circle. Compare this to Nachi who is giving a robot to one of our vendors to use for development. Whatever functionality they need Nachi works with them to figure it out. Meanwhile they cannot even get access to Motoman manuals for a project involving a major automotive vendor. While Yaskawa jealously guards user manuals it's competitors are handing out robots and detailed technical support.

If Yaskawa were to open up they would find that their customers will be one of their best sources of innovation and that stories of their customer's success to be their best sales tool. Ex. Factory robot guy is looking how to automate a difficult operation with a new part. His current vendor does not think what he wants to do is possible. Somewhat dissatisfied the guy goes to robot-forum.com to look for answers. Robot guy finds a thread where a user describes a similar application and how he solved it with a Motoman. Guy goes to boss, "We should talk to Motoman."

Thanks, that helps. Looks like I'll have to either pull the quadrature signal off the drive or from a standalone encoder. Now the question is if we can push the RS422 signal over 80 meters of cable.

I have an application with a DX100 where we are exploring the use of conveyor synchronization. The documentation indicates that conveyor synchronization requires a quadrature signal into the YCP02 board. The controller for the conveyor cannot output a quadrature signal but it can transmit the numerical position via a network connection. For our purposes the network can provide position updates with plenty of speed.

I am wondering if the count from the encoder inputs on the YCP02 card are written to a user accessible location. I'd like to avoid the overhead of capturing this quadrature signal and routing it over separate cable if the data can be written directly to a register. I could not find anything regarding these inputs in the Conveyor Synchronization or the Concurrent I/O manual.

Does anyone know were the YCP02 inputs are mapped to and if they are accessible?

That sounds very impressive. I am interested in similar approaches to find the apex of a curved panel. I have sent you a PM.

I am interested in this encoder converter for my application. Any info on it?

I have the exact same issue. I am working on a project which uses a very large number of addresses for I/O and data in over two dozen programs. I am looking for a tool where I can collect which addresses are used where so we can maintain and update the code.