[MrMetric]

I'm looking for *any* information from someone with knowledge about how a tool presetter probe is interface to a Mitsubishi control. Ideally I'm looking for information on an older Meldas M3 control and a Renishaw TS27R, but I suspect that this is not all that necessary. The machine I want to adapt is a Dyna Mechtronics DM4400m, although, again, the type of machine probably doesn't matter that much.

The TS27R is really nothing more than a switch. On my MC161 CPU card the switch connects to CNA18 #1 and #7. The latter is +24VDC and the former is the input pin. Now the problem.... The M3 control (well, all of the Mitsubishi controls) used G37 as the Auto TLM process. I am 99% sure that I have the physical interface working properly. But when I initiate the G37, I never get a length recognition when I manually "trip" the length sensor. There could be a few different reasons for this. First, the manuals make sound like the input I'm using is shared by several SKIP inputs and you somehow tell the control which one to look at. This means I might need an additional command maybe in the code? Is that right? But the manual doesn't mention anything about an input selection. Or maybe this is all vague because it is part of the PLC? I don't really know.

I also know that Renishaw has a CD with a bunch of macros. I'm not entirely sure how this fits with the functions built into the Meldas control. So, one of my questions for *anyone* that has a Meldas control is.... Do you have macros installed that manage your tool presetter (or touch probe), or do is everything managed by the built in functions of the control?

Basically, I'm trying to reach out to someone that has some type of probe connected to a Meldas control. We can go from there.

Thanks!

[MrMetric]

I figured this out. I ended up integrating into the CNA18 interface's SKIP inputs. In this fashion G31 and G31.2 provide the means to interrupt the movement. The BIG problem was that the firmware on the M3 could not support the Renishaw macros. That was a complete pain to concretely determine. I ended up getting a new firmware cartridge with C0. I don't know when Mitsubishi updated the firmware so it didn't cause problems, but I can concretely say that A3 and old do *NOT* work. I hope to eventually make copies of the EPROMs in case you can't find a newer version.

[MrMetric]

I didn't completely answer my original request for help, and I firmly believe in trying to give back to the community so that someone else can self-help. To that end, let me explain a bit..

CNA18 is an interface port on the MC161-1 CPU card. CNA18 has the pins that are high speed interrupt pins bound to the SKIP control. A SKIP control is basically a means of interrupting an operation. This is done by inserting a G31.x (where X can be 1, 2, 3, etc... one for each input pin... G31 by itself implies G31.1). If you've watched any of the thousand videos out there with a renishaw probe, you'll see it advance until it hits a surface. In G code land, that is just a Gxx to move at a specific feedrate BUT there is a G31 added to the command as a modifier. So, even though the move command says to advance 50mm, if the input changes state at, say 10mm, the movement stops/interrupts. Personally, SKIP is kind of a lousy name, in my opinion, but that is all that is really happening.

Renishaw tool setters and probes work with a set of macros. These handle all of the movements, which are, of course, interrupted by the G31 input changing state. The Renishaw macros then read the current location through common variables and perform whatever action is appropriate. In a tool presetter case, the tool offset will be calculated and the value put into the variable location that is bound to what you see on the screen. Think of these locations as dual ported memory. You can either access them through a graphical user interface (keyboard and screen) or programmatically via a bound memory location. That is *all* there is to it.

Now for the Meldas M3 specifically.... What I found is that I could not run the macros without getting format errors. It was frustrating.... An out of the box macro would error out in the first 15 lines or so, and the solution was to just remove a comment line. That isn't good... A comment line is functionally irrelevant, so removing it should do nothing at all, but it made all the difference in the world. The implication, of course, was that there was likely a problem with the firmware (basically the operating system of the CNC itself) was malfunctioning. The firmware is stored on EPROMs on this control, and the labels showed this version:

MC433 261W000-A3
MACRO 261W610-A1

Eventually I found another plugin module with the 12 EPROMs that had this sticker:

MC433 261W000-C0
MACRO 261W610-A2

That version doesn't have all the format issues that the original one had. Note that there *are* other firmware versions found out other M3 controls out there that have different part numbers. I cannot tell you which ones work and which ones do not. I only know that I tried two different firmware cartridges which A3/A1 and neither of those worked (likely eliminating corruption of the firmware itself), while the C0/A2 version seems to behave much better (I haven't fully vetted this yet, so there is definitely a TBD in here).

Feel free to reach out to me if you have a question. I'll help if I can, but please realize that I *am* an amateur. I have learned this stuff by painfully clawing my way up a cliff of knowledge. :-(

[Seven30]

I didn't completely answer my original request for help, and I firmly believe in trying to give back to the community so that someone else can self-help. To that end, let me explain a bit..

CNA18 is an interface port on the MC161-1 CPU card. CNA18 has the pins that are high speed interrupt pins bound to the SKIP control. A SKIP control is basically a means of interrupting an operation. This is done by inserting a G31.x (where X can be 1, 2, 3, etc... one for each input pin... G31 by itself implies G31.1). If you've watched any of the thousand videos out there with a renishaw probe, you'll see it advance until it hits a surface. In G code land, that is just a Gxx to move at a specific feedrate BUT there is a G31 added to the command as a modifier. So, even though the move command says to advance 50mm, if the input changes state at, say 10mm, the movement stops/interrupts. Personally, SKIP is kind of a lousy name, in my opinion, but that is all that is really happening.

Renishaw tool setters and probes work with a set of macros. These handle all of the movements, which are, of course, interrupted by the G31 input changing state. The Renishaw macros then read the current location through common variables and perform whatever action is appropriate. In a tool presetter case, the tool offset will be calculated and the value put into the variable location that is bound to what you see on the screen. Think of these locations as dual ported memory. You can either access them through a graphical user interface (keyboard and screen) or programmatically via a bound memory location. That is *all* there is to it.

Now for the Meldas M3 specifically.... What I found is that I could not run the macros without getting format errors. It was frustrating.... An out of the box macro would error out in the first 15 lines or so, and the solution was to just remove a comment line. That isn't good... A comment line is functionally irrelevant, so removing it should do nothing at all, but it made all the difference in the world. The implication, of course, was that there was likely a problem with the firmware (basically the operating system of the CNC itself) was malfunctioning. The firmware is stored on EPROMs on this control, and the labels showed this version:

MC433 261W000-A3
MACRO 261W610-A1

Eventually I found another plugin module with the 12 EPROMs that had this sticker:

MC433 261W000-C0
MACRO 261W610-A2

That version doesn't have all the format issues that the original one had. Note that there *are* other firmware versions found out other M3 controls out there that have different part numbers. I cannot tell you which ones work and which ones do not. I only know that I tried two different firmware cartridges which A3/A1 and neither of those worked (likely eliminating corruption of the firmware itself), while the C0/A2 version seems to behave much better (I haven't fully vetted this yet, so there is definitely a TBD in here).

Feel free to reach out to me if you have a question. I'll help if I can, but please realize that I *am* an amateur. I have learned this stuff by painfully clawing my way up a cliff of knowledge. :-(

First although this is an older thread thanks for taking time to detail your work. I thought Id mentioned a couple of problems Ive run into on my Mitsubishi TC8B lathe and see if you might have any suggestions.

The control is a M325LV W/MC-161b cpu. The mc-161 died within a few hours of receiving the machine before I could make a parameter backup but I did have a paper copy. I was new to Meldas so it took a while to get it running again. Seems good with exception of built in Auto Tool Measure does not work. Cannot get it to even pop out of its door when I press manual tool set button. Im studying the ladder which I have but wonder if Ive lost some macro or perhaps an option setting.

This machine also has the graphics board and screen but attempting to draw toolpath no tool tip shows up. Thinking that might be some option parameter.

[MrMetric]

You didn't say where you are from, but judging from your English, I'm guessing that it is in North America? If so, I'd say that your first bet would be to contact Mitsubishi itself. Sign up for their MEAU support website as they have a plethora of information out there and a moderately decent search engine to help you find documents. Unfortunately, if you are not in North America, I don't think you'll be able to register because they will IP block you. You may need to use a VPN in that case.

If that doesn't answer your question, then I'd send a support request email to the US site. Again, they've been surprisingly helpful with me.... But now for the rub. What you are asking about is almost certainly a macro. I won't say that conclusively because I don't know the lathe control, but pretty generically speaking, the controls are fairly "dumb" on things that are outside their control realm. They rely heavily on integrations, which is through a variety of methods. First, there are the IO assignments, which are parameters. The next level would be the PLC. These are typically in EPROM, so unless your EPROM took a hike (bad day!) then you probably just need to activate the macro properly (parameters). Finally, there are macros. These can either be triggered by PLC or entirely managed by calls made within your g-code. This last one would be setup/managed by your post-processor on the computer. This is how all of the Renishaw stuff works.

Mitsubishi MIGHT be able to give you more help than a non-Mitsu machine simply because they built the whole thing (unless it was a contract build, in which case they might not be able to help). All I can really say is that it would be a good step to at least send them an email to see if they have any support tips.

As for graphics.... Again, this is getting a bit out of my element. I can tell you that there are a host of non-published configuration parameters that will manage things like the amount of memory and/or graphics. It could be, therefore, that the parameter you seek is not set. These are typically paid for parameters, however, and are generally setup by Mitsubishi. They should have records of your machine so that they can get you back to a base state. And given that it is a mitsubishi machine, they may actually have the original distribution parameters. They seem to have pretty good records. They were able to tell me exactly what machine a spare control I have came from (down to the serial number of the machine).

Another thing... Far too many people freak out and start formatting too much when the 161 "dies". It is worth noting that the parameters are actually written to EPROM, which will not go away when the battery dies UNLESS you start formatting too deeply. The battery just keeps the SRAM alive. But that is program memory, and usually people have backups of their programs, tools, etc. Unfortunately, it is also macro memory though......

Finally, not all PLC is in EPROM. I know Mitsubishi has a module where the PLC was all in RAM and you would read it in. That would be dependent upon a battery. I have never actually heard of someone having this flavor, but it definitely did exist. It might have been more for the Mazak side of the world as Mitsubishi sold a lot of this stuff to them. The software has a different frontend, but much of the deeper stuff is still the same.

Anyhow, hopefully this will give you a few ideas....

[Seven30]

Thanks for the quick response! Yes USA/Colorado so Ill create MEAU account and see if Mits may have records for this machine.

Before I swapped out the MC-161 as a last ditch attempt I had tried the SW 1,3,5 reload attempt. No idea if that did
anything as the old cpu never started again.

There are about four pages of extensive ladder logic for the tool measurement system. I can see the tool measure button
toggle in the ladder monitor but before pulling my hair tracking it all down Ill contact Meau/Mits.

I noted with some amusement the MC68020/MC68881 and TMC320C25 signal processor. Pretty hot stuff for a CNC in the
early 90s.

EDIT: "Finally, not all PLC is in EPROM. I know Mitsubishi has a module where the PLC was all in RAM and you would read it in. That would be dependent upon a battery. I have never actually heard of someone having this flavor, but it definitely did exist. It might have been more for the Mazak side of the world as Mitsubishi sold a lot of this stuff to them. The software has a different frontend, but much of the deeper stuff is still the same."

I think this is what this TC8B lathe also does. Part of the recovery process is to reload battery backed ram from the rom/eprom module. I do need to replace the batteries at some point and they SAY a super cap will allow enough time to pull board and swap out battery. gulp!