Parallel Port Voltage lower with Ubuntu 10.04

More
28 Nov 2010 19:46 #5705 by Woody
Hi,

I'm moving from Mach 3 to EMC2 and have had a few issues.

1) getting a ChargePump output to Pin 14 - solved that by removing some .HAL text linking the estop pin with the ChargePump pin as solved by others - why would it do that!!??? (and still be doing that years after it was noticed?)

2) with Windows2000 and Mach3 the Charge Pump on pin 14 is a rather gittery but functional 0.3 to 5V wave form at about 10kHz BUT with the same PC and same BIOS settings and interface hardware the parallel port wave form is a perfect 10kh (no noticeable jitter at all) but 0.5V to 3.6V and so the charge pump circuit does not quite run.

Any idea why? I can't see anywhere to adjust anything related to this. Where are the Parallel port driver setting like in Windows? Sorry, I'm new to Linux...

So, frequency and waveform are perfect, just voltage level is 3.6V with Ubuntu 10.04 not 5V as with W2k, (just software changed)?

3) why does the chargePump output not stop when eStoped or power down machine? Is that a HAL setting I need to change?

All inputs from the mill to EMC2 work fine. Just outputs are lower voltage than with Windows

If I can get this fixed then I'm really looking forward to playing with EMC2. Looks very professional.

Thanks,

Woody.

Please Log in or Create an account to join the conversation.

More
29 Nov 2010 13:26 #5722 by andypugh
Woody wrote:

2) with Windows2000 and Mach3 the Charge Pump on pin 14 is a rather gittery but functional 0.3 to 5V wave form at about 10kHz BUT with the same PC and same BIOS settings and interface hardware the parallel port wave form is a perfect 10kh (no noticeable jitter at all) but 0.5V to 3.6V and so the charge pump circuit does not quite run..


That seems really quite bizarre. P-port voltage is purely a hardware thing, Linux/emc/windows/mach have no control over it as far as I know.
The only possibility I can think of is that the EMC2 config files are leaving more pins high at any one time than Mach was, and the P-port is over loaded.

Your charge-pump problem should be easy to solve with a 1k pull-up from the input pin to 5V. It is much more effective to treat the P-Port as a set of switches-to-ground rather than a source of 0 or 5V.

Please Log in or Create an account to join the conversation.

More
29 Nov 2010 14:19 #5727 by Woody
Thanks, that's an excellent suggestion I will try a 1K resistor this evening.

Yes, maybe it is the default "rest of pins" settings which I've not checked. But with Win2K I could switch all the digital outputs on (spindle, flood etc) and the charge pump was still fine, so not sure if the electrical loading is too high but I'll definately check it.

I wondered if there was different modes for the parallel port to be configured in on OS startup which had different pin configs?

Any idea why the charge pump is always on when EMC2 is started irrespective of eStop or machine on-off? Would that be in my HAL file setup?

Woody.

Please Log in or Create an account to join the conversation.

More
29 Nov 2010 14:36 #5728 by andypugh
Woody wrote:

I wondered if there was different modes for the parallel port to be configured in on OS startup which had different pin configs?


Not as far as I am aware, but then you could fill Wikipedia with the stuff I don't know.
It is vaguely possible that EMC2 is starting the P-Port in EPP mode and Mach was doing so in ECP mode (or vice-versa, or some other combination) and that the p-port runs in different voltages in different modes. 3.6V is too high for a 3.3V port, and I don't think 3.3V is a standard anyway, even if some laptops only give you that much.

Any idea why the charge pump is always on when EMC2 is started irrespective of eStop or machine on-off? Would that be in my HAL file setup?/quote]
Didn't you say you had deleted the link to the e-stop circuit?

Anyway, on all the time is right. All that the charge-pump says is "It's OK, the computer is running and in control. you can stay live". If the charge pump is running then all the pins are in a known (and correct) state and there is no reason for the drives to power down. If you want the drives to power down then you should use the amp-enable pins too.

Please Log in or Create an account to join the conversation.

More
29 Nov 2010 21:33 #5750 by Woody
Tried inverting all my EMC outputs I either get all at 0.44V or all at 3.6V which made no difference at all to the charge pump waveform on the 'scope so I don't think its a loading issue.

My board needs the parallel port to pull low to connect the opto coupler to ground so the problem seems to be the parallel port is not switching off the output fully so the board cannot pull the pin up to 5V. 1k resistor to 5V made very tiny difference, got to about 3.65V. Its doing this on all the outputs. Very strange. No such problem with Win2K.

( I believe the parallel port is open collector and needs pull up on the thing that is plugged in?)

I fudged the charge pump on the board so the Xylotex board would enable and I can jog!!!!! So, at least that side of things is OK with only 3.6V. (And very smooth jogging, so I'm very happy!)

But charge pump, and digital outputs no go at the moment.

I've got a PCI parallel port board coming so I'll just give that a try. Perhaps the motherboard is sensitive to some other unrelated Ubuntu setting somewhere.

If that has the same problem I'll try 8.04 (although I've no idea how yet). If same, I make an interface just for those output pins, USB 5V powered.

Woody.

Please Log in or Create an account to join the conversation.

More
29 Nov 2010 22:00 #5752 by andypugh
It might possibly be worth looking in the BIOS settings, maybe limit the mode to ECP or Legacy, see if that makes a difference. Linux/EMC2 might try to set the port to EPP (which is important for the Mesa and Pico P-Port boards).

Generally I would suggest getting a Mesa 7i43 as a P-port upgrade, but if there is any doubt about the port's ability to drive it then perhaps that would be an expensive error.

Please Log in or Create an account to join the conversation.

More
30 Nov 2010 22:19 #5783 by Woody
Got it working!!!

Yes, it was the BIOS settings. Changed from EPP to ECP then to "Bidirectional Parallel Port" and finally it worked, giving exactly the same waveform as Mach3 and board then worked perfectly.

Apparently it is possible for the OS to change the PP config after boot up so perhaps Mach3 forces the port into std bidirectional mode without it being obvious and just works (for me) whereas EMC uses what it gets as set up by the BIOS. I can't think of any other explanation.

Maybe this will help others if they see a voltage restriction with their setup.

Now just have to figure out why Ubuntu 10.04 is so slow on my P4 1.7GHz machine where as W2k was lighting fast. I've just got rid of all the startup stuff and checked "no fancy graphics", which has helped but still very slow even when not much CPU activity. Is there noticeable difference between 10.04 and 8.04?

Thanks for your help.

Woody.

Please Log in or Create an account to join the conversation.

More
30 Nov 2010 23:27 #5786 by andypugh
Woody wrote:

Apparently it is possible for the OS to change the PP config after boot up so perhaps Mach3 forces the port into std bidirectional mode without it being obvious and just works (for me) whereas EMC uses what it gets as set up by the BIOS. I can't think of any other explanation.


I think it might be that EMC2 forces the port into EPP mode unless the BIOS tells it that it can't. EPP mode is essential for some of the common EMC2 hardware. (That which uses the p-port as a communications bus rather than a set of pins)

Now just have to figure out why Ubuntu 10.04 is so slow on my P4 1.7GHz machine where as W2k was lighting fast. I've just got rid of all the startup stuff and checked "no fancy graphics", which has helped but still very slow even when not much CPU activity. Is there noticeable difference between 10.04 and 8.04


I can't make a direct comparison, but 10.04 is rather snappy on my Atom board, and 8.04 is a little slower on my dual (old)Xeon.

Please Log in or Create an account to join the conversation.

More
01 Dec 2010 12:15 #5793 by Rick G
Woody,

Now just have to figure out why Ubuntu 10.04 is so slow on my P4 1.7GHz machine where as W2k was lighting fast. I've just got rid of all the startup stuff and checked "no fancy graphics", which has helped but still very slow even when not much CPU activity. Is there noticeable difference between 10.04 and 8.04?


You might want to explore the video driver troubleshooting section and try some of the options there. Apparently 8.04 and 10.04 handle setting up the video drivers differently.

Rick G

Please Log in or Create an account to join the conversation.

More
01 Dec 2010 14:10 #5794 by Woody
Thanks Rick, yep I think that is it. Looking, I've got the "Nvidia tnt2 model 64" board which seems troublesome and no specific drivers get found. Looks OK and everything works, but I think its not using hardware acceleration. If it does not become a problem for EMC, I'll live with it. If not, I've found reference to a generic Nvidia driver that may help but I can't work out how to force its use yet.

Anyway, EMC2 is working and I've got something to play with now. Very different "feel" to Mach3.

Woody.

Please Log in or Create an account to join the conversation.

Time to create page: 0.181 seconds
Powered by Kunena Forum