ANCA 2000 Controller DNC problem

[derekj308]

Hi Guys

I'm trying to get an ANCA 2000 controller to DNC and have communications problems.

We have been able to transfer using straight RS232 upload and download but didn't use the settings as recommended in the manual we have. The manual says 8 data, 1 stop, no parity which transferred data but incorrectly. Using 7 data 1 stop even parity gave us accurate transfer back and forth except for "(" at the start and ")" at the end [no quotation marks, just the brackets] when downloading from the CNC. Easily edited out, not a real issue. When we try to DNC we get "COMMUNICATION ERROR' repeated every 5 or so lines as the file is being transferred using the DNC software. We are trying EasyDNCXP as it has a free 60 day evaluation.

We wired our 25 pin cables as per the diagram in the manual, we can use RS232 to transfer small files so I would think cable is not our issue.

Also the manual give an example of reading a memory address W4000:0022 to show the baud rate configured in the controller. When we look at this address we see FFFF:-0001 which tells me the address has been wiped. I checked W0001:0001 just for the hell of it and there are values in that so would this mean it isn't a battery backup issue? If that is the problem ie the value in W4000:0022 needs to be set how is it done?

Any assistance would be greatly appreciated.

Cheers

Derek

[HarryDaniel.com]

Hi Derek and fellow CNC zoners

My first posting, just joined today. No better way to introduce oneself than with an answer to a previously un-replied issue, followed by a query of my own.

First the answer:

As you say, the baud rate for an ANCA 2000 controller is specified as a word (2 bytes) at segment 4000, offset 22Hex, namely address W4000:22 in "ANCA speak". This is achieved by entering the command "I" from the Diagnostics Page, which is Page 7, from the Main menu, that you can drill back to, via th ESC key. At the Diagnostics page, you will see a cluttered full screen of:-

1 = DISPLAY OUTPUT LOGICALS
2 = DISPLAY INPUT LOGICALS
3 = DISPLAY OUTPUT PHYSICALS
4 = DISPLAY INPUT PHYSICALS
5 = DISPLAY AUXILIARYINPUTS etc.

You want to Inspect memory and alter if if necessary at address W4000:22 (Hex). Do this via the I command at this screen, which results in the following line, in the middle of the existing, uncleared screen, requesting the address of the memory location in question, namely:-

memory locn ? _

to which you reply W4000:22 (followed by a carriage return key). i.e:

memory locn ? W4000:22 <cr>

which yields the word at that location, which say for 9600(decimal) would be:-

4000:22=02580H,9600 ?_

since (16**3)*2 + (16**2)*5 + (16**1)*8 + (16**0)*0 = 9600.

Entering a carriage return <cr> will maintain the contents and step onto the next word address and display it's contents. Typing in the new hex value followed by the <cr> will replace the word contents and step onto the next address. On my ANCA, the contents are as shown for an arbitrary 6 word range from offset 22H

4000:22=02580H,9600 ?_
4000:24=02580H,9600 ?<cr>
4000:26=00000H,0000 ?<cr>
4000:28=02580H,0000 ?<cr>
4000:2A=0000FH,0015 ?<cr>
4000:2C=0000AH,0010 ?<cr>

I'd also like to know what offset 24H determines, also 9600 as shown.

On the other hand, if you are looking at a byte addresses, the W prefix is not used. eg the byte address for segment 4000 offset 8A is simply 4000:8A which happens to be the byte of value 04 that controls 8 data, No Parity, 1 Stop bit. I'm yet to discover the bit definitions for this byte. I'd also like to know why one should use 8,N,1 for DNC communications and 7,E,2 for any other kind of terminal emulation at the PC end, without changing anything at the ANCA end!

Hope this helps, if not yourself then anyone else who is following this thread with a similar issue, since there have been no replies documented to date.

Now my query:

My ANCA has decided to stop being able to receive files via the RS232 link, after having the controller serviced at great expense, for a battery replacement and lost settings. It fails the acid test, namely the "^T Test Communications" from DNC on the ANCA with a "!! TEST FAIL !!" at the ANCA after 2 secs, connected to a PC running DNC5. The test also fails after a "T Test Communications" at the PC end, with an instantaneous "!! TEST FAIL !!" at the PC. My ANCA Com speed is 9600 as above and I have tried permutations of 9600:N,8,1 9600:E,7,1 9600:E,7,2 at the DNC5, PC end, to no avail.

Running a Terminal Coms program at the PC end shows "[!T/<cr>" (excluding the quotes) being instantaneously received at the PC end, in response to a ^T at the ANCA running DNC. This is only if the PC coms is set for E,7,2 or E,7,1 otherwise garbled if N,8,1 (as per my reference above).

Divorcing DNC at the ANCA (namely option 5, form the page 8 , OUT Program option) from the equation and just doing an RS232 INPUT (option 3) and sending data from the PC fails but OUTPUT (option 4) to the PC works, with the PC set for E,7,2 or E,7,1


It's not a wire continuity or crossover issue. An RS232 breakout panel shows correct levels for each Tx line and a loopback connector substituted at the controller board cable, verifies coms at the PC end with an echo. Almost indicating that the Rx input to the comms port on the controller is dead.

Does anyone know of a local loopback test at the ANCA end, to verify that it's Tx can talk to it's own Rx line?

The only fly in the ointment, is the PC end has been replaced with a faster unit, but with an identical disk image to the original, after the old PC died. (RIP). The serial Com port on the newer PC echoes nicely to the null modem, replaced at the ANCA controller end, for test purposes.

Nothing has changed with the previously successful serial line between the ANCA and PC, which I believe is a 3 wire connect 2-3, 3-2, 7-7 between them and a local 4-5, 6-8-20 at each end of the D25's. I've even tested with a laptop right at the ANCA controller and fastidiously connected. I'm clutching at straws at the moment and thinking maybe a rise time issue on the PC Tx line, which may need capacitor smoothing. No too silly but I'm running out of tricks.

Any constructive suggestions most welcome.

Cheers
Harry

[derekj308]

Hi Harry

Thanks for the reply and thanks for the detailed post. I think I know what your issue is. It's the fly in the ointment. I was told by a guy who has experience with ANCA controllers that they don't work with anything other than 286's and early 386's. I'm not sure of the reason why but its what he told me.

Cheers

Derek

[HarryDaniel.com]

Thanks Derek

I've also heard this and attributed it to sloppy programming in the DNC software, probably, simulating delays by unsophisticated countdown loops.

So I've divorced DNC from the equation and used conventional Terminal Emulation software, to test communicate with the serial line on the ANCA. This separates the added complexity of having to obey whatever added protocol DNC and the ANCA have agreed upon, in order to stay in sync with each other. I have a personal preference for Kermit to do this Terminal Emulation but this is irrelevant. I'm using RS232-In and RS232-out at the ANCA end to communicate from and to the Coms port, which I've appropriately connected to my Terminal Emulation Coms port on my PC.

Whereas RS232 is a universal standard at the character level and should be able to communicate with any other serial device, whatever the underlying CPU type or speed, as long as the line characteristics are mutually compatible, namely baud, parity, stop bits, handshaking etc. This is vindicated by the fact that I can read serial coms from the ANCA on the newer PC, using Terminal Emulation but my problem is that I can not write to it, hence my suspicion about the Rx line of the ANCA's D25 Com port. It would be nice if the ANCA had a (hidden) test routine to loop test to itself via the RS232 line, needing the attachment of a test loopback connector at it's D25 outlet.

Wish I had an old 286, 386 just to eliminate the theory, but they are rare as hens teeth and the one I have needs fixing. Thanks for the reply and please keep in touch with any ideas.

Cheers
Harry

[hatcat]

Hi guys.

I am new here and have just come across this thread.

I don't know whether you've solved your problems one way or another, but if you haven't, I have a copy of the "2000 SERIES CNC FMS DNC MANUAL" which may possibly be of some help. It is release 5.0 and dated 21/9/1990.

It says that to get the CNC to test the communication, send it a ^T. Since I don't currently have an actual controller, it is difficult to ascertain whether the manual is talking about the CNC controller or the PC software at this point. It says to:-

1. from the main directory select page 8 (EDIT MODE).
2. select menu (O)UT (PROGRAM I/O)
3. select option 5 (DNC) to display DNC menu.
It shows a ^T entry in this menu, so the ^T may be sent from here, or perhaps over the RS232 from the PC?

The CNC should then send out a test message which it expects to be echoed back to it. It will then finish by displaying TEST FAIL or TEST OK.

Interestingly (after reading this thread), the manual states that data bits are fixed at 8. Only stop bits and parity are variable.

Let me know if you need more.

Cheers,
Peter

[derekj308]

Hi Peter

I think I had access to that same manual and we did try ^T and got TEST FAIL.

Thanks for the info anyway.

Cheers

Derek

[hatcat]

Hi Derek,

I noticed on another (old) thread that you were looking for a copy of ancadnc.exe. Did you ever get hold of it?

I am also looking for whatever I need to do DNC on an ANCA 2000. I've got a machine, but unfortunately, none of the computer or software which used to talk to it.

Thanks,
Peter

[derekj308]

Hi Peter

No I never did get a copy. In the end I decided to do nothing with the controller or trying to talk to it again other than to swear at it. When I get a spare moment I might go open up the cabinet and see if I can isolate the drives and bypass all of the ANCA control crap. The objective would be to run it with Mach 3. Even if I couldn't isolate the drives I would consider making a new control cabinet and just use the original motors and encoders.

Cheers

Derek

[SprinterTRD]

Hi Everyone,

This is my first post in this forum and a newbie to milling My background is mainly electronics & race cars.

I recently got my ANCA 2000 mill up and running. I have been able to manually enter G codes and run the programs and send G codes via RS232.

The Mill came with a PC that has software to send code to the CNC, as well as some other software which I haven't worked out how to use, Its for the DNC function.

The following manuals came with the Mill

ANCA 2000 Series Operators manual release 3.1
ANCAMill 2 2 axis Universal Milling MC release 1.2
and another one which is not for this Mill
ANCA CNC200 & Series E CNC

I am going to do some reverse engineering on the RS232 to see if I can get the DNC side working.

Hatcat,
I would be interested in obtaining a copy of the 2000 SERIES CNC FMS DNC MANUAL that you have, it may help me.

Steve

[hatcat]

Hi Steve.

Welcome to the world of CNC (not that I've been here long myself).

At least you got a computer with software with yours! I didn't, and haven't been able to get hold of a copy of the ANCA DNC program which it appears you have. I've written a very basic program up to the point where it allows me to transfer programs back and forth between the CNC and PC. Sure beats developing a program to do something, and then having to vapourise it to make room for the next one!

I'm not really understanding where you are at concerning the manual, and why you feel the need to reverse engineer the RS232. Is the DNC program not working, or are you just after the instructions for using it?

My background is also mostly in cars and electronics. I presume from your login that you are into Toyotas?

Kind regards,

Peter

[SprinterTRD]

Hi Peter,

I have some DNC software but I am not sure if it is the Anca one. I think the name is DNC7 or something like that. I will get the name of it tonight. I only played with it for a short time and could not get it working. I was hoping the manual you have may relate to the software I have.

There is another bit of software to transfer code Via the "RS232" recieve mode and has a built in editor however it has a dongle on the parallel port to get it working.

I wanted to write windows software for the controller, which is why I wanted to see whats going on in the comms dapartment.

Yes I am into Toyota's, I have one of the TRD/TTA Group A Sprinters that raced back in the 80's

[hatcat]

Hi Steve,

Attached is a pic of the screen shown in the manual I have, so you can tell whether it matches what you've got or not. I think it is the first startup screen, but as I said, I don't have the program, so I can't tell you for sure. Sorry about the quality of the pic - it is a scan from the manual which I suspect might have been through a fax machine at some stage.

As far as I know, the ANCA program is ANCADNC.exe, so it may be different.

What make and model is your mill?

Kind regards,
Peter