plugging encoders directly into LPT1

[justCNCit]

Hey, I just ordered the encoder board from Rogers machine. Plugged it in and haven't gotten it to work at all, was wondering if anyone had experience running it here.

Sent an email to Roger 2 days ago and am expecting a response at any time. But since I am impatient, is there anyone here who would know something about this?

Basically, the point of this board was to get encoder functionality going with Mach 3. I plugged the board into 115v and plugged in the LPT1 to the Xylotex controller, 4 axis box. From there loaded the provided DLL file into Mach 3 and enabled it. I'm running a laptop computer with no secondary parallel port. So I didn't enable port #2 in Mach 3. Maybe I have the pinouts wrong for the interface. I'll double check that but pretty sure I checked over their help file for it.

if this board doesn't end up working out, can the encoders be plugged directly into the LPT1 port? and if so how difficult is it to know which pins to plug them into? I guess the xylotex manual would tell me which pins are available.

It would suck though, not getting this interface board working. Let you guys know if/when Rogers gets back to me on this one. Thanks for any help.

[Al_The_Man]

Not probably all that relevant in your case, but as far as reading and testing an encoder from the parallel port see the PC DRO.
http://www.lindsayengraving.com/othe...rests/dro.html
BTW, I used direct encoder input without the buffer transistors shown.
Al.

[antichip]

I was hoping to hear that answer. No I just got to find the pinout for the encoder I found.

[justCNCit]

i think the problem with my setup is that I am using a LPT1 on port 1 but Romax seems to be setup to run a second parrellel port, because looking at pins 1-9 on the Xylotex they use the same pins for motor output as the Rogers interface board seems to use for encoder input.

That means I might have to rewire the LPT1 cord?

[Torchhead]

The xylotex card inputs do not go straight through to the 25 pin cable. If you put the inputs into the right screw terminals then you can pick it up on the parallel port but it's no difference from making a direct parallel port connection (no isolation, no pullups). I am pretty sure the Rogers board uses the 2nd parallel port inputs since you only have 4 open inputs on port 1 in MACH. It won't work even directly connected if you need more than 2 encoders or you want any normal inputs like Homes or limits.

The dll for the Rogers card would assume you are using specific pins on a specific port. The custom screens with User defined DRO's and the math the compare the User Dro's with the Axis DRO's would be in that plug-in as well. You are not going to see anything on the normal XYZ Dro's because they are hooked to the MACH pulse engine.

TOM Caudle
www.CandCNC.com

[Torchhead]

Originally Posted by justCNCit i think the problem with my setup is that I am using a LPT1 on port 1 but Romax seems to be setup to run a second parrellel port, because looking at pins 1-9 on the Xylotex they use the same pins for motor output as the Rogers interface board seems to use for encoder input. That means I might have to rewire the LPT1 cord?

Port 1 cannot be the pulse port (step & dir) AND an expanded Input port unless you do some fancy mux stuff like we do with our UBOB card. Rewiring the cable won't help. You can NOT make the port "listen" to pins 2-9 in the software!

You cannot even set port 1 in mach up with 2-9 as inputs. You could "fool" the software and use port 1 address for port 2 in MACH and just use it as a DRO. You lose all other I/O functions in MACH

The Rogers cards is designed to use two parallel ports and needs 6 (maybe more) open inputs. those have to be on pins 1-9.

What are you trying to accomplish?

[justCNCit]

What do you mean by custom screens? I enabled the software but was not aware of custom screens.

Also from the sounds of it you're telling me I have to use a second parallel port in order to provide enough pins to use both systems at once. Was trying to avoid that, but whatever.

thanks for the help.

edit: Just noticed your second post.

[justCNCit]

Originally Posted by Torchhead Port 1 cannot be the pulse port (step & dir) AND an expanded Input port unless you do some fancy mux stuff like we do with our UBOB card. Rewiring the cable won't help. You can NOT make the port "listen" to pins 2-9 in the software! You cannot even set port 1 in mach up with 2-9 as inputs. You could "fool" the software and use port 1 address for port 2 in MACH and just use it as a DRO. You lose all other I/O functions in MACH The Rogers cards is designed to use two parallel ports and needs 6 (maybe more) open inputs. those have to be on pins 1-9. What are you trying to accomplish?

I'm trying to keep track of the steps on the machine. As you know the Xylotex board is being used. Everytime I run a program it loses steps and I am too deep into steppers to easily switch over to servos.

as unwise as it sounds the encoders are meant to be insurance to keep from messing up parts (because I will have to constantly monitor the machine)

hopefully it gets running good enough to where it doesn't miss steps too often. But with the stupidity of steppers it seemed unavoidable to tack encoders onto the ends in order to get something done on this machine.

[justCNCit]

today I plugged a second parallel port into the PC with Mach 3 and put the hex address from device properties into Mach 3, with no results. Installed the drivers for it too. Installed the plugin and activated it. Where am I going wrong? the interface guy isn't responding to my email.l

[Torchhead]

Originally Posted by justCNCit today I plugged a second parallel port into the PC with Mach 3 and put the hex address from device properties into Mach 3, with no results. Installed the drivers for it too. Installed the plugin and activated it. Where am I going wrong? the interface guy isn't responding to my email.l

Did you get a manual or instructions with the package?.

Losing steps, much like backlash, is best dealt with in the stepper and machine design rather than in software. The kit that you bought is supposed to setup OEM DRO's (real position) and compare them against the MACH DRO's and stop motion if the error becomes too great. It probably won't work with any serial output protocol (like USB) since the actual position and software position will always be different.

Properly designed and setup (and properly used) stepper systems do not lose steps. I have a stepper based router table I use in daily production and we have NEVER lost steps unless we had a crash (which is usually caused by doing something stupid).

It's important that the step & dir signals to your drives are properly buffered and clean.

If you can't get support for a product and you can't figure it out yourself then you may want to consider another solution.

You can use the MACH diagnostics screen and the Raw Port Bits to determine if the encoder signals are even getting to MACH. In most cases it has to do with not having the 2nd port setup correctly. In the Raw Port Bits that is the actual signals on each port pin. You should be able to see them flash when you operate the encoders. If you do and they don't show up on the screen in the proper DRO's it's a problem in MACH setup or the plug-in. IF they don't flash the port pins in diagnostics then you probably don't have the port setup right.

We sell a multi-function Hand Controller for MACH that uses 6 of the port 2 inputs and I have a special screen for testing and diagnostics of the product and it's the raw port bits decoded. Most often when we get a Hand Control that does not work either the parallel port is not setup or the plug-in is not loaded and enabled in MACH.


TOM Caudle
www.CandCNC.com

[justCNCit]

we did a test (me and Roger from the Rogers & Romaxx machine site)

established that the encoder and board worked by attaching a analog multimeter to the encoder posts, found that the encoder is pulsing normally.

It looks like a software problem, drivers most likely. Though the drivers off of the LPT2/3 CD for parallel port expansion seem to install fine, copying the hex addresses from device manager/properties to Mach 3 has resulted in no changes on settings/diagnostic screens.

The plugin has been enabled, I've tried the recommended port 2 pinouts to get feedback going in/out, and have gotten no positive results on the diagnostic screen from any inputs.

Roger seems to think the hex address for interfacing the second port is wrong, but I am at a loss to explain this

the only thing is in device manager, 'PCI SIMPLE COMMUNICATIONS DEVICE' is showing the drivers have not been installed and is not working, though it does show that under Ports (lpt/serial) subtab for device manager, the chipset for LPT2 and 3 have been installed.

any help appreciated.

[Torchhead]

Use the Raw Port Bits on the diagnostic screen. Port2 inputs are (mostly) the bottom row. IF the parallel port is at the right hex address most of the bottom row will light up. You have to make sure that you are not using the hex address of the "multifunction card" it installs. On a NETMOS card it installs two drivers and only one is the actual LPT port. You have to use the RESOURCES under the actual LTP2 (or 3) port it installs. if you have to, do a manual install of the card using the Add/Remove Hardware in Windows. Make it use the drivers off the CD that comes with the card. You should be able to test inputs by using a jumper in the plug from the parallel port and sort between ground (pins 18 - 25) to the numbered pins. It should turn the LED's off/on in the RAW PORT BITS.

Make sure the port 2 in MACH is enabled, has the right port address and the Use pins 2-9 as inputs is checked. If you get no changes on the RAW Port bits before or after the port card install and setup MACH is not seeing the card. The first Hex address in the Resources Tab for the LPT are the correct ones.

Since RAW Port BITs are the actual port pin signals that MACH see's pin mappings to functions (or even a bad plug-in won't effect the RAW PORT BITS).

TOM Caudle
www.CandCNC.com