[nvdave]

Hello,
I am into my first retrofit and need some help. I plan to start a build thread and have been taking pictures, but I need to get my ducks in a row for that. The machine is a 1990-1992 Ram Mill 13x42 CNC vertical knee mill. This is my first "real" machine tool. It came with a Hust control which eventually lost all but MDI 1 step operation. After 2 years or so of shopping and reading, I finally purchased the Ajax/Mach3 setup for 3rd party drives. Scott at Ajax said he would take a look at the manual for my inverter to hopefully help me out with a diagram of how to connect it to the Ajax system. Well, 10 days later of 12 hour days working on the mill has put me way past the point where I would've liked to have had the answers I'm seeking. I bet this is a real simple thing for many of you. The inverter I have is a Mitsubishi A024 - 3.7K. I do not have the PU unit for it, so I cannot change parameters. I do have the manual for the inverter. Any help would be greatly appreciated!!!:drowning:

Thanks,
Dave

[cncsnw]

You will want to buy the keypad for it sooner rather than later. An inverter you can't program or troubleshoot is less than worthless.

If for some reason that is impossible, you could replace it for about $350. How much is your time worth?

That said, assuming yours is programmed in a more or less default configuration, you can probably just use the diagram in the manual (section 6.3 in the copy I found online):

Close SD to STF to run forwards.
Close SD to STR to run reverse.
Close SD to RES to reset faults.
Apply 0-10V analog speed reference to terminals 2 and 5 (analog output to 2, common to 5).
The Fault output (wired to a PLC input so you can detect when the inverter faults) would use either A and C or B and C, depending on whether you want a normally-open or a normally-closed fault signal. You probably should use normally-closed, which is probably B and C.

[nvdave]

Thanks for the help! Some of that I did actually understand out of the manual, but the alarm contacts definitely threw me, and I wasn't positive about any of it. I'm hoping that I can run the inverter in the external control mode and not need to change parameters.

There is also a "MRSAT" terminal on the inverter- would I use this to stop the spindle? I have mounted a maintained switch on my control panel for spindle on/off that I actually wired into a contactor feeding the 3 phase to the inverter. The control for this contactor also runs through the e-stop relay. I'm thinking maybe I should have the contactor controlled only by the e-stop and for normal spindle on/off operation, I should wire into that MRSAT terminal. Any thoughts on this?

Also, I find the Ajax documentation lacking. There are conflicts on the pinouts and other needed info is just plain missing. Where on the Ajax GPIO4D do I wire these connections to the inverter??? I followed Scott's diagram to connect my AMC drives - that ate up input 13, which the documentation, I believe, says spindle enable, for instance???


I know I'm asking a lot of questions, but, hopefully, they are simple answers.

Any additional help will also be appreciated!
Thanks Again,
Dave

[Al_The_Man]

You definitely need a PU unit.
Try ebay, is this one is a common one.
ebay #330514218755
Al.

[nvdave]

From the amount of times I've seen cncsnw and Al The Man post with awesome info, I know now I need the PU. I made an offer on one on feebay - didn't know if that FR-PU02E would work, manual lists the FR-PU03E, so that's the one I made the offer on. Thanks for the info.

By the Way, Al The Man, I saved some info from a thread where you and others helped someone out with a differential to TTL converter chip and the pinout connections to connect an old style (like my Sumtak) MPG wheel to Ajax. I purchased the chip, a socket and a small board and soldered it all up and connected it last night. I have one question that will probably make you laugh, but how do you know the chip orientation? I got the data sheet on the IC from Digikey and I see the indent on the schematic that matches the socket, but the only thing on the IC is the lettering on top and a little indented circle on top. Which way is which? I got 2 IC's in case I blew one.

[Al_The_Man]

Sometimes the identification is a bit obscure, the ones with the U in the centre of the end is usually more obvious.
But the indented circle should be in one corner, the nearest pin is pin 1.
Al.

[nvdave]

Thanks, Al! I'll check that out. Any ideas on the inverter "MRSAT" terminal and its use for the stop spindle function or on the Ajax specific connections for the inverter?

Thanks Again,
Dave

[cncsnw]

I doubt you need to use the "MRSAT" terminal at all.

You probably don't want or need additional panel switches or switch-controlled contactors. Let the CNC control start and stop the spindle for you.

It will start and run the spindle by closing SD to either STR or STF, and keeping it closed for as long as the spindle is to remain running. It will stop the spindle by opening that connection.

If you want a minimal level of E-stop protection, route the SD connection that comes from the inverter to the common point(s) of the PLC relays through a pole of your E-stop contactor.

If you want a more robust level of E-stop protection (but one that is harder on your inverter) then route the 3-phase power that feeds the inverter either through the E-stop contactor, or through a separate contactor that is switched by the E-stop contactor.

If you want additional front-panel buttons to start and stop the spindle, wire them to PLC input points and put logic in the PLC program to recognize them as manual-mode spindle start and stop.

You can use any of the 16 relays on the GPIO4D board to command spindle run and reverse. Your PLC program is responsible for closing the correct relays when it wants the spindle to run.

The most common default PLC programs I see for the GPIO4D have spindle run (either direction) on OUT7, and spindle reverse on OUT8. The wire from the SD terminal would go to the OUT7 common; you would install a jumper from the OUT7 output to the OUT8 common; the run-forward signal would go from OUT8 NC (normally-closed) to STF; and the run-reverse signal would go from OUT8 NO (normally-open) to STR.

[Al_The_Man]

I am not sure about that particular model, I have use the A200 that has a MRS input, this is a stop input, I usually set mine up for either fwd or rev.
So either on is on or off, so the MRS input function is not used.
Al.

[nvdave]

Al - Thanks Again, I'll explore that terminal's function a bit more.

cncsnw - thanks very much - this is what I need. I am concerned with having a very limited number of inputs on the GPIO4D; that's why I was looking to have my panel switches control the spindle on/off outside of the GPIO4D. I currently have an e-stop relay (24VDC coil) which is controlling a 3 pole contactor feeding the 3 phase to the inverter. I read in the inverter manual that it was hard on the inverter, as you said to use the contactor, but it was vague about whether it was refering to the contactor on the line or on the load side of the inverter. That's when I read something about the stop terminal and thought maybe I could use that for manual spindle on/ off. The old controller had the on/off button and I like having it. As far as the robust e-stop...absolutely! I actually also have the estop relay controlling another contactor which will break the AC inputs to each of the 3 AMC drives. My thinking was that anything that can do harm is controlled by the e-stop.
As far as the inputs go, maybe you would be kind enough to help me a little more. This is what I have for my inputs currently:

in 1. limit x-
in 2. limit x+
in 3. limit y-
in 4. limit y+
1-4 com. 24vdc common

in 5. limit z-
in 6. limit z+
in 7. (was thinking jog -)
in 8. (was thinking jog +)
5-8 com. 24vdc common

in 9. (was thinking cycle start)
in 10. (was thinking feed hold)
in 11. e-stop
in 12.
9-12 com.

in 13. x axis enable
in 14. y axis enable
in 15. z axis enable
in 16. (should this be spindle fault?)
13-16 com. 5vdc+

I also have 2 selector switches:
1. axis select - x,y,z
2. multiplier - 1x,10,100x
should the common of these switches go to 5vdc+ or 5vdc common?

I would eventually like:
1. spindle speed override
2. feed rate override
will this be possible with only input 12 left (assuming I didn't miss anything else)?

I also see on the mpg pinout diagram that there is aux input 1,2,3 and aux output 1,2 - can these be used? If so, for what and what do the commons go to - 5vdc+ or 5vdc common?

As I'm currently unemployed, I would like to avoid buying an input expansion board.

Any further help you can give me would be greatly appreciated. I can't get a return email from Scott at Ajax (can't even get the stupid activation code so I can post on the Ajax forum:boxing

Thanks Again,
Dave

[cncsnw]

My advice is to just use the keyboard jog panel features for jogging, jog increment select, manual spindle start/stop, and so on. Once you get the machine up and running, and can make some money with it, buy the jog pendant.

It will be an uphill battle all the way to try to roll your own operator panel without the benefit of an I/O expansion board.

One of the nice features of the Centroid/Ajax control is the integration of the jog panel pendant. You get all the standard jogging, cycle control and machine control keys, plus programmable Aux keys, all in a unit that requires no extra wiring; doesn't occupy any of your PLC I/O; and takes little or no custom PLC programming.

If $739 is out of the question right now, just use the keyboard features that are already available and reasonably functional.

[nvdave]

Am I really reaching that far? I thought I was pretty close to having all of this going. I've already made the control panel, mounted all of the switches and have them about half wired in. Whats mainly lacking is where to tie them in on the board. I know I will have to write brains for some of them and will need some help for any that are more complex than the simple cycle start/feed hold example from Scott. But I thought this was definitely do-able?
If I forget about the feed and spindle overrides, am I in the ballpark on the rest?

Thanks Again,
Dave

[cncsnw]

Well, if you don't bother to connect the inverter fault signal, then you have two PLC inputs available.

Your axis selector requires at least two inputs; your increment selector requires at least two. You could conceivably wire those things into the MPG header on the MPU11; but I think you said you also want to use an MPG.

You could leave three of your limit switches off the PLC (but perhaps still wire them as directional inhibits to your drives). The PLC only really needs to see the switches that you are going to home to. You can probably trust the software travel limits to keep you off the other end.

I don't think you need to tie up INP13, INP14 and INP15 with enable detection. If your objective is to detect when GPIO4D has set the axis enable outputs, your PLC program can just look at OUT17, OUT18 and OUT19 directly. The GPIO4D doesn't need to send them around to its own inputs just to see what state they are in.

So on review, you can probably free up enough inputs to make it work.

Feedrate and spindle override pots would require analog inputs, which you don't have at all.

[nvdave]

That's more encouraging - thanks cncsnw for taking another look!

On the mpg header, I wired into the pins for x,y,z axis select; 1x,10x,100x multiplier; wheel a,a-,b,b-;+5vdc,5vdc com. I assumed that these pins were for these connections. I saw in a post on Ajax where you said that the common for these switches should go to 5vdc com, so that's what I did - haven't tried it yet.

As far as the inputs 13,14,15 go, it seemed strange to have to lose 3 inputs just to loop a signal right back into the board, but I was following Scott's diagram for the AMC 30A8 drive, which has the exact same pinout as my 16A20AC drives and he confirmed to me on the phone, that it should be the same for my drives. You guys all know much more about this stuff than I do. I also downloaded the brain that he posted to make the diagram work, although I haven't installed it yet.

If my assumtions on the mpg header are right, I should have just enough inputs, I think. x-,x+,y-,y+,z-,z+,jog-,jog+,cycle start,feed hold,e-stop,enable-x,y,z,spindle fault. This is 15 of the 16 inputs - I hope I'm on the right track here.

Regarding the spindle fault: is this the circuit that goes through the A,B,C contacts (whichever is NC) on the inverter and do I feed it the 5vdc+ from the GPIO4D power header?

I am attaching Scott's diagram for the AMC drive as well as inserting a link to my build thread with some pics of my machine. I hope you will check it out and tell me what you think.

http://www.cnczone.com/forums/vertic...tml#post883802

I'm sorry to be such a pain, but it is very much appreciated!

Thanks One More Time,
Dave

[cncsnw]

Yes, the axis select and increment select inputs on the MPG header are for those purposes. Ordinarily they are used with selector switches built into an MPG pendant and apply only to the handwheel, but I don't see any reason you couldn't use them to control the function of jog buttons as well.

Perhaps there are limitations in the Mach I/O system that prevents it from reliably reading the states of OUT17-OUT20. I have only used CNC11 software, which can read any of the output states without restriction. That would be a question for Ajax or Mach.

You "feed" the spindle drive fault inputs with whatever voltage source you have chosen for that bank of inputs on the GPIO4D. For each bank of four inputs, you choose a voltage, source, and direction (sourcing or sinking); you wire one side of a DC supply to the input common terminal for that bank; you wire the other side of the same DC supply to the other side of the input switches (e.g. the C terminal on the inverter's fault relay); you wire the switches (e.g. the A or B terminal) to the input points; and you make sure you have the right SIP resistor installed for that bank: 470 ohm for 5V inputs, 1K for 12V inputs, 2.2K for 24V inputs).

You would wire the +5VDC from the GPIO4D logic supply to the common of the fault relay, if and only if you have wired the 0V side of that supply to the OUT13-OUT16 common and have installed a 470-ohm SIP for R4.

[nvdave]

Thanks, cncsnw! It's a relief to hear you say the mpg header wiring should work.

I believe the 17-20 inputs are all used for the x,y,z,a faults, so I'm not really messing with those.

After asking you about the spindle fault, I realized I have it going into a bank that utilizes +12v and that I will have to use that same voltage through the contacts on the inverter, I have the 102 SIPS in place for that bank already. I've been very careful to install the correct SIPS for each bank. I appreciate you giving me the heads-up on the SIPS and all the wiring help.

I just powered-up for the first time since initial testing last night. Output 1 wasn't closing, which kept my e-stop circuit open. I believe it's because I need to jumper out all the unused faults and inputs. I did see all the limits clear and the estop clear on the diagnostic screen in Mach. When I bypassed the Output 1 connection, my drive enable leds lit up, so it's looking promising, so far.
I printed out the CNC11 manual and am going to use the software as I have read Scott suggest to check all of my inputs, etc. I am going to connect the inverter as per your and Al The Man's suggestions and see how that works, if the other I/O testing goes well. I plan to connect the analog out on the GPIO4D to the speed reference (terminals 2 and 5) on the inverter. According to the inverter manual, it ships with a default setting of 0-5vdc, which can be changed to 0-10vdc. I'm not sure what the GPIO4D outputs, can't remember what I saw on that.

Did you, by any chance, check out the build thread?

Thanks for all your help; I'll let you know how it goes.
Dave