Issue with old Novamill and VR Milling 5

[GillianS]

I have a Novamill with ATC dating from 1997. The controller shipped with the standard DOS software but was subsequently upgraded to the VR Milling 2.1 ROM. The controller is the Original SmartStep/3 and serial only controller.

I borrowed my Microrouter compact laptop and VR Milling 5 setup to try to drive it and I don't seem to be getting very far. I've backed up the machine config, selected an old legacy Novamill and edited the settings to enable the ATC. When run, the controller is detected, it starts probing for the controller software, and invites me to download a new mint file from the PC. The PC Mint on the hard disk is 214.3 and the mint card version on the machine is a suspicious 631.632.

I say "Yes" and after watching the download progress and a starting message, I immediately get an error message that the machine cannot be communicated with.

I've checked the voltage rails for the card (+,-12V and 6V) the 2.1 Rom is seated, the controller is seated, I've used an FTDI serial adaptor with correctly linked pins as well as what I think is the correct long "Lindy" 1812GG Denford Machine Link cable in conjunction with another USB PL2303 adaptor and directly to the laptop serial port.
Short of using a desktop relic with a true +-12V swing serial port, I'm not sure what I can do.

By way of diagnostics, the red LED indicates L for loading and never seems to change. I'm concerned that I don't have a DenStep setup and the other Novamill machine choices are for NextMove controllers judging by the serial port settings.

Am I flogging a dead horse? Will an original SmartStep board and controller support VR Milling 5 with the 2.1 ROM? Is the software expecting me to be running a DenStep wired driver board with it's additional IO? Am I falling foul of using a too fast a laptop that is stuffing data down the serial (19200 baud) that the controller drops data? Might I have a dead controller?

Pointers welcome.....

Gillian

As a follow-up: I know the SmartStep drivers and mill mechanics work fine as I've exercised everything using a Mesa card but I'd quite like to get the VR software running. Now if someone has a working example of the Denford 6 pocket ATC for Linux CNC they can share then, perhaps, I could be persuaded to the dark side

[Martin]

It may be down to the internal RS 232 lead.

It would be best to make up a 9 way to 9 way lead to connect the machine directly to the computer.

viewtopic.php?t=231

[GillianS]

I till try the direct link and bypass the existing wiring. I will grab my box of serial patch cables and have fun at the weekend.

Thanks for the pointer to the post on the wiring.

Gillian

[GillianS]

I tried a "laplink" cable from laptop to baldor controller and it didn't help.

I refitted the original bulkhead DB25 and connected my FTDI based USB to serial controller back on.

It still refused to communicate so I looked a the advanced settings for the COM4 port and changed RX and TX bytes to the minimum 64 and set the latency timer to the minimum value of 1.

Success!!!

I've added the teaching keyboard using another USB->DB25 converter and tried a quick three tool engrave in a block of wood.
One Novamill with ATC now up and running again. My to-do list includes looking at the ATC and fine tuning the stop position and index rotation to minimize the catching of the tool as it's returned to the pocket. I will read up on the serial driver and look to disable the selective suspend.

My current com port settings attached