Hello Guest it is March 19, 2024, 02:14:48 AM

Author Topic: I beleive this is a PC Hardware Conflict  (Read 5624 times)

0 Members and 1 Guest are viewing this topic.

I beleive this is a PC Hardware Conflict
« on: June 23, 2008, 01:41:23 AM »
I have been using Mach for 3 years now on many machines I have built three Hotwire, 3 Router tables and 1 and a half water jet, the half is the one that's eating me starting at my brain.

Here's the problem I have a industrial PC (from China) It's got a built in Keyboard and flat screen LCD PC has On board Video and the parallel port is out putting 3.3v but I have a C10 breakout board. All of the inputs and relay outputs are working fine. I have the A is slaved to the Y axis with matching gear rack, motors, gears, drives (Y and A have separate drives) I'm using G340s (Step multiplier x 5) with 500 Cnt Encoders. on 50:1 Planetary gear boxes so I need 5000 steps per rev. and 150 Vel, and accel 30 Motors were tunned with a Oscilloscope with a pulse generator at 40 msec pulse rate

Checked, double checked, Constupid checked the Configuration I had the kernel set at 25k and increased it to 45k in which improved when it was set to 25k if I hit the arrow key for x axis jog the A axis jogged roughly and error-ed out the G340s after 1/4 th of an inch of travel, Y axis jogging moved the X and Y and error-ed out the Geckos after and 1/8 th of an inch and racked the gantry. I had a G100 on this but had too many conflicts with the Chinese computer.

But now I'm at wits end. I have the following in the XML file:
 Axis   Port    Step Pin     Dir Pin
 X          1          2           3
 Y          1          4           5
 A          1          6           7

As I said I get complete limit, E-Stop, and Input#1 checking good.

I have a PCI 32 bit multi I/O parallel port card coming from http://rogersmachine.net/PCIport.html but what else could cause such a problem, Oh I forgot to mention that I was jogging on 5% of 150 inches per minute when it was wigging out any higher and it errors instantly.

I'm posting the xml file.

Wayne
Re: I beleive this is a PC Hardware Conflict
« Reply #1 on: June 23, 2008, 01:44:37 AM »
I forgot the Xml File

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
    • View Profile
Re: I beleive this is a PC Hardware Conflict
« Reply #2 on: June 23, 2008, 02:38:23 AM »
Have you run the drivetest to see what the pulse looks like?
Hood

Offline da21

*
  •  87 87
    • View Profile
    • M-Machine
Re: I beleive this is a PC Hardware Conflict
« Reply #3 on: June 23, 2008, 03:06:29 AM »

Some opto isolators do not like 3.3v parrell ports ,( as 3v is at the midway point of the voltage swing of 5v , the opto sometimes has difficulty deciding which logic level is correct and so gives an incorrect pulse , the higher the port speed the more this can happen )  and are a better match to 5v  , so your problem could be the parrallel port and or including  the opto isolation switching being poor at the frequencys being used  , try bypassing the opto isolators for the axis step as a test , as inputs or relays etc  are at either logic high or low ( 0v or 3.3v )then this problem will not show on switched limits etc .

their are convertor chips available that can handle the 3.3v to 5v logic levels , if this is found to be the problem , or even replacing the opto's for a different type more suited to 3.3v may help .

also check your dc supply to the bob , make sure you have enough current 

Dave   
 

Offline jimpinder

*
  •  1,232 1,232
  • Wakefield, West Yorks, UK
    • View Profile
Re: I beleive this is a PC Hardware Conflict
« Reply #4 on: June 23, 2008, 04:09:40 AM »
If you look at it logically -

You say you have built several machines - and I assume they all perform satisfactorily - so

If this machine doesn't perform - WHICH BITS ARE NEW TO THE SCENE -  and I assume it must be the chinese computer, although that is a guess. All the other stuff appears fairly standard, and if you have used them on other machines in this type of scenario, then all would seem to point to the computer, if my assumptions are correct - although the spec seems very good.

It would seem therefore, to be a waste of time messing about - try another computer on the same machine and see if that performs properly. If it does, you know where the fault lies.

A smooth stepper output may be the answer, if you can get hold of one of those to try.
Not me driving the engine - I'm better looking.
Re: I beleive this is a PC Hardware Conflict
« Reply #5 on: June 23, 2008, 11:21:35 AM »
I just ran a drive test on it and at 25k, 35k, 45k and 60k and very clean, Which is pointing more to the parallel card. The C10 Breack out board doesn't have any Opto Iso on it I have a Bob Campbell board on the way. And I put a Smooth stepper on oder but they are two weeks out. I Shoud have my new Parallel card tomarrow I will try it on the computer tomorrow.

Thanks for everyones input, Wayne

Offline da21

*
  •  87 87
    • View Profile
    • M-Machine
Re: I beleive this is a PC Hardware Conflict
« Reply #6 on: June 23, 2008, 11:40:11 AM »
ok , i'd agree  by the tests you have done  it looks very much as it's the parrell port. it's becoming a common problem

Dave   
Re: I beleive this is a PC Hardware Conflict
« Reply #7 on: June 24, 2008, 09:44:34 AM »
Rich,
Do you think adding a video card would possibly fix that issue? There is a two week wait for SS.

Wayne

Offline Hood

*
  •  25,835 25,835
  • Carnoustie, Scotland
    • View Profile
Re: I beleive this is a PC Hardware Conflict
« Reply #8 on: June 24, 2008, 10:01:55 AM »
It may and it may not, usually however if your onboard card is causing problems the DriveTest will be bad. But note I said USUALLY as there are no hard and fast rules :(

Hood

Offline da21

*
  •  87 87
    • View Profile
    • M-Machine
Re: I beleive this is a PC Hardware Conflict
« Reply #9 on: June 24, 2008, 05:10:39 PM »
I doubt if a video card etc would solve the problem you are experencing , it's a bit more deep routed as a motherboard design issue
and then matching the p port to the breakout board  , i'd wait for your ss , a much more approprate solution all round in my opinion
 
Dave