PDA

View Full Version : Timer Over-run



kamen
07-02-2005, 12:04 PM
Hi again
Now I have another problem.
When I swich my spindell on in
Control Box Prooblem window appears mesage
No Return Value Timer Over-run

I tryed to run my ShopBot on Win98
Is this software problem or is another grounding problem?
The Colombo spindel only gets power supplay from control box, and have no other connection with control box.

Regards

paco
07-02-2005, 01:24 PM
Hi Karmen!

The first thing that I think of is power connections in the control box... I get this message when I shut off the control box BEFORE ending SB3...

Can you duplicate this EACH time you try to start the spindle?

Do you have to restart SB3 after this happen?

Software? don't think so... grounding? maybe check it out to be sure if connections are OK at the control box.

gerald_d
07-02-2005, 01:30 PM
Kamen said "I tryed to run my ShopBot on Win98"..... Did it run correctly with another Windows before?

richards
07-02-2005, 01:37 PM
Check the USB cable. Sometimes mine gets partially disconnected (when I step on it) and I get that error.

bleeth
07-02-2005, 02:12 PM
Usually that goes away just by clicking to close it once or twice as long as the control box is on. I will get it sometimes if I turn my control box off and then back on. Sometimes I will get rid of it by turning the controller off and on! Sometimes I need to restart SB3. I also run Win98 on my shop floor computer and I believe the latest version of SB is pushing it for most systems running 98. By the next SB version we may need more computer on the floor. I spoke with Frank about it a while back and he had me check for 98 updates. It helped some.

Dave

billp
07-02-2005, 02:13 PM
Have you tried "the Fix" which Jim Zaccarro posted a few weeks back on this Forum? Apparently there is an "issue"with some of the USB pigtails coming out of the control box (they are not grounded properly). The "fix"is to connect a wire with an alligator clip to the connection where the USB pigtail, and the cord from your computer connect, and then connect the other end of that cord to your control box. It eliminated 90% of my Alpha issues, and I know of a few others who have had success with it as well.

bleeth
07-02-2005, 02:15 PM
Addendum to above: My spindle is running from a dedicated 220-1ph circuit and my controller from a different 110 circuit. My computer runs from a third circuit.

Dave

bob_lofthouse
07-02-2005, 03:23 PM
I only get this message when I turn off the control box before shutting down the control software.... which is the same as Paco

gerald_d
07-02-2005, 03:38 PM
My impression is that Kamen was operating on a later Win version before he struck this problem with Win98. I am increasingly finding that Win98 cannot keep up with recently re-written software - the programmers seem to be leaning towards stuff that will run properly only on true 32-bit Windows (NT, 2000 & XP). I have converted two of my PC's to Win2000 recently and there is a big improvement in some applications.

Who has their Alphas running happily on Win98?

ETA. With Win98, you must also make sure that you have the latest DirectX.

Ryan Patterson
07-02-2005, 06:55 PM
I just install a 3hp spindle and am having the same problem. Talked to SB they said it might be noise from the VFD. They suggested that the e stop wire was acting as an antenna and had me disconnect the e stop. (Had to jump wires from es1 to es2 and es3 to es4) This did not solve my problem. I now think it might be in the USB connection. The only wiring I have for the spindle in the control bos is the control wire to turn it on and off. The main feed is not going through the control box. ShopBot also suggest to reroute high voltage from low as best as I could. I am going to try a new USB cable and will let you know if it solved the problem.

richards
07-03-2005, 01:29 PM
Ryan,
Electrical noise is a MAJOR problem in many industries that use process control computers to control electrical/mechanical equipment. I've fought the noise problem for thirty+ years and have never found a 'standard' method to eliminate the noise. Although I don't have any experience with a spindle yet (mine arrives on July 7th), I've started preparing my alpha for its arrival. All low voltage lines running on the flex-strap-bow that spans the Y-axis will be moved as far away as possible from the spindle's power cables. If necessary, I'll wrap the low-voltage lines with a braided copper wire (similar to the wire/cable that can be found in the dust collection section of your local Woodcraft store) and connect ONE end of the copper wire to frame ground. If the problem still persists, I'll replace all low-voltage lines with shielded CAT-5 cable (twisted pair cable) and attach ONE end of the shield to frame ground.

Working with the USB connection is a good idea. The general design of USB is noise prone. (In my opinion, a much better approach would have been to use fiber-optics to connect the computer to the control box - at a cost of about $200-$500, but without the noise problem.)

One last thought: Neatness counts when there is noise; however, it is VERY important to NOT coil cables. Coiled cables can act like a transformer and actually increase the noise problem. Also, instead of laying a cable(s) along the floor in a serpentined loop, vary the angle and distance between 'loops'. Having cables lie parallel to each other can induce noise.

As many others have said in many other posts, proper grounding of ALL parts of the machine is critical. Keeping low-voltage lines away from current carrying lines is critical. Checking all connections for solid crimps/solder/screw-terminals is critical.

richards
07-03-2005, 01:57 PM
An oscilloscope is a valuable tool in finding and eliminating noise. Connecting a 'scope to the low-voltage lines will show the jitter that causes the false triggering of inputs. Remember that a low TTL signal must be between 0.0 and 0.7 volts while a high TTL signal must be at least 2.2 volt (assuming the standard 5VDC VCC supply - CMOS and other types of chips might have higher or lower voltages). The rise time of each signal transistion should be in the nanosecond range (the time that the signal takes to go from low to high or from high to low). Watching the 'scope while the machine is running will show you what kinds of moves/speeds/switching causes the problem. After finding the cause of the problem, the 'scope will show whether moving cables, grounding, crimping, etc. removes the problem.

By their very design, stepper motors and the VFD of a spindle generate electrical noise. The switching power supply in the computer generates electrical noise. The low-voltage input/output lines connected to the control box act like classical antennas - and antennas pick up signals. Finding and fixing electrical noise would normally be considered to be outside the normal responsiblities of a woodworking shop, but when you're the only person in the shop that responsibility has to be faced and the problem has to be fixed.

Ryan Patterson
07-03-2005, 05:35 PM
I tried a new USB cable it did not work. So I started to move the low voltage wires around and found it may be the Z zero plate that is the cause of the problem.

Mike,
Would a small capacitor help if I installed it on the zero plate.

richards
07-03-2005, 09:56 PM
Ryan,
It's hard to say if connecting a capacitor to the Z-zero circuit would do the job. A capacitor blocks DC and passes AC, so a static signal like the one the Z-Zero plate generates (similar to a switch closure) might not work. Attaching one end of a capacitor to ground and the other end to a resistor and then connecting the resistor to the signal side of the circuit might work, but it might also slow down the signal's transition from high to low too much. At the very least, that could give a false reading and could easily break a bit.

If I HAD to modify the input circuit, I would tie the input signal line high through a resistor (2K to 4.7K) to +5vdc. That would normally hold the signal line in its positive state until the bit made positive contact with the Z-zero plate. If that didn't entirely clear up the problem, I would add the capacitor to the circuit, between the signal line and ground.

What I've done with my Z-zero plate is to NOT attach it to the Y-carriage, but to mount its holster to the control box. I always measure the Z-axis height at the XY,5,5 position, so having the holster very near that position works great for me (and I don't need to run the Z-zero cable across the machine - so it has less chance to pick up electrical noise). Whether it will still work with the spindle will be a mystery until my spindle arrives.

kamen
07-05-2005, 03:38 AM
Here's the situation.
As my Shop is not finish yet I started the ShopBot with the same old computer with win98. It's not the best but it,s worked. When first hit limit swich problems accurred I reinstall win98 on the same computer because think that this is software problem. The new instalation on the same machine gave the control box problem mesage.
In the same time Shop Bot wopk well with my notebook.
I have some scary experience yesterday vhen my Step motors just going mad and make random muvement vith very high speed.
I stopped the machine and examine all cables and control box connections. As I know step motors are Voltage controled devices, so I think that was wrong voltage generated from software or somewere in the box. One of the ground wire in the box was disconected so I check all the coupling in the box. Meny of the screws was loose, so I thing it's good idea to check youre control boxes too. After reconecting everything was wright and SB work properly. The limith swich mesage appears again only in the long programs. (I have one with 2500 lines)
I disconnect limit swiches and the problem disappear. Today i'll try it with shielded cables.
My Spindel cable is all shielded so I thihk that from the begining the problem was in the limit swiches.
As Gerald mentioned win2000 is better choice
so I'll try it too.

Ryan I tested widely ZZero plate. In my case it was not any influence over USB and the work of the system at all. I thing that grounding is the first thing you should check. If grounding is OK you can remove aligator clip from ZZero plate. You can use shealded cable. My Zzero plate is not attached to the y-carriage too.

Thanks for advices
Tomorrow I'll post the result of the test.
Thaks again.

gerald_d
07-05-2005, 07:09 AM
The bottom half of this (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=26&post=2408#POST2408) archived thread could be interesting. Then we were talking of interference caused by power lines, cell phones and fluorescent(neon) lights.....

wayneo
07-05-2005, 06:25 PM
Ryan,
I used my SB probe for the first time over the weekend. The first couple times I tried it would probe along for a while, and then all of a sudden it would quit with an error message - something about a bad probe - or whatever..... Re-starting, it continued probing longer, however it still stopped after a while with the same error message.

I thought "gee maybe it is trying to move to fast and jiggling the probe to set it off". So I slowed the move speed down. That seemed to help for a while, but it still stopped with the same problem.

I searched around the wire for the probe and discovered that the Z-Zero plate was very close to touching a piece of metal on the carriage. I repositioned it and that eliminated the problem.

You might check that something is not somehow jiggling and making contact with the Z-Zero Plate or some other similar situation.

Wayneo

Ryan Patterson
07-05-2005, 10:21 PM
I think I solved the problem. I think it was my first thought of the USB cable. I tried what was suggested by Jim Zaccarro and bill, so far it is working but I have thought it was working before to. The lost connection was not 100% repeatable. Some times I will get going without a connection problem. The only thing that was repeatable is when it starts acting up I will lose a connection every time the spindle comes on. I had removed all the accessories (e stop, zero plate,prox switches, output relays) and still had the problem. When I grounded the USB cable the connection was not a problem.

gerald_d
07-16-2005, 03:03 AM
This post (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=27&post=26111#POST26111) may also explain the problem.