[cnczane]

I got through page 30 (THAT is gold-medal reading). Here's what I think they said. Later readers of similar bent/warp/OCD-nature might consider continuing this list from page 31 down through the present-day, by copying/editing/augmenting the contents of this.
case design
http://www.cnczone.com/forums/open_s...tml#post758004

Wiring schematic
http://www.cnczone.com/forums/open_s...tml#post758009

case DXF
http://www.cnczone.com/forums/open_s...tml#post845556
http://www.cnczone.com/forums/open_s...tml#post845708
http://www.cnczone.com/forums/open_s...tml#post864829

magnets from old CDROM drives
http://www.cnczone.com/forums/open_s...tml#post880023
http://www.cnczone.com/forums/open_s...tml#post851019

magnet testing
http://www.cnczone.com/forums/open_s...tml#post847057

magnet holder
http://www.cnczone.com/forums/open_s...tml#post850004

possible parts list
http://www.cnczone.com/forums/open_s...tml#post866325

possible wire source
http://www.cnczone.com/forums/open_s...tml#post899125

about the adhesives
http://www.cnczone.com/forums/open_s...tml#post977271

typical ebay order
http://www.cnczone.com/forums/open_s...tml#post983334

[Robert M]

I bow to you Dan.....thks for this summery....and your patience !...yep, a well deserved medal ;-)

[jm82792]

Hi all!

I have no website just yet. That costs money, and on a student budget right now, it's not feasible. I have posted information in the Product News and Announcement thread, located here for those interested. I'm not allowed to post the information here to sell them.

If you do hosting from Arvixe, use Wordpress, and finally Woocomerce then you'll be set for under $8 a month.
The $8 is for hosting, and the rest (theme, and shopping cart setup along with PP account) including Paypal is free.

[tjb1]

If you do hosting from Arvixe, use Wordpress, and finally Woocomerce then you'll be set for under $8 a month.
The $8 is for hosting, and the rest (theme, and shopping cart setup along with PP account) including Paypal is free.

Free Web Hosting with PHP, MySQL and cPanel, No Ads is free and have been very good for me, worth checking out.

[Flemilla]

Anybody want to verify the schematic?

Hi!
This is a bit offtopic, so sorry, if you prefer we can move to anther post...
I just receive today the BOB, I did not open the CD, just google a bit and found this post...
So at least seem that the XYZ will work out of the box, but not the endstop and B axis.
I will check the schematic, but I have one doubt on the error found in the B-axis. What is the error and how to fix it? we need to cut any trace?
Regards

[RonaldoNZ]

The bob will work (XYZA, by a fluke, but not as it should.) out of the box. However the fix is very easy to make it work properly.

All credit to John_100 for the discovery and the below images that detail the repair.

Connect pins 13 and 14 as show, just solder bridge the two pins.
Attachment 226010

Then run a small link wire on the back of the board to supply them with the correct +5V
Attachment 226006

If you need the display / B as well, take a look at this one, the two resistors need to connected end to end as they are in the other channels and the display connector should connect directly to the input pin of the 74HC244 and not to pin 1 and 2 of the D15. I neither use B, a pendant on the d15 or the display connector so haven't bothered.

[Falcon69]

LessPaul has notified me of a fix and after testing it, I am happy to say that the Limit Switch Kit I have developed (with help of people on this forum) now performs as it should. A limit switch with the safety factor of a normally closed switch. The switches, with the new fix, now will activate Mach3 when a switch has been unplugged or accidentally cut. Although they still perform in a normally open configuration, they will act as if they are closed. The switches are wired in parallel, and NOT in series. That should also make wiring much easier for anyone wiring these up.

Here's how I tested it.

I created a wire with a diode, 2K2 resistor, and an LED to simulate the G540's output/input for the limit switches. The G540 uses an Optocoupler operated on 12 volts with a 2K2 pullup resistor. this is the wire I create to simulate it.

I connected that wire to 12 volts, and turned on the kit with all the switches plugged in. With the wire for the G540 plugged into one, then all, one after another, I tested the limit switches by bringing the magnet close. Good, as soon as the magnet got close, the LED turned on, meaning the LED in the Optocoupler would activate the transistor, sending the signal to Mach3. So, as you can see, it thinks it is a normally open switch. Otherwise, the LED would be on, then turn off when limit switch is triggered, but performs like a normally closed switch, thanks to the series of ULN2803 Chips on the board.

BUT! When a switch is unplugged, any of the switches in parallel to the G540 wire, it turns on the LED. However, it only turns on the LED in the switch and on the front panel for ONLY the switch that was unplugged, giving a visual indication that, "HEY! You cut through X-Axis Home Switch!" for example.

Thanks to LessPaul for finding this fix!

I will post the new schematic for it, as soon as I have finished drawing it up.

SO, to recap, the Kit now operates as it should, with a safety factor that still sends a signal to the BoB (and then onto the software) that a switch has been cut or unplugged accidentally.


Falcon69

[RonaldoNZ]

And if your board is unpowered or faulty? No stop condition.

You are still relying on a logic high or power on to signal the bob.

I.E. How do you power that optocoupler to high if nothing is connected to it? for instance if there is a cable fault between your board and the bob or there is a power supply failure?
That isn't inherently safe I'm afraid, not in any normal sense. if your circuit for any reason fails the bob has no way to know it and will blindly drive past that stop condition that it can never register.

The only way to make an inherently safe electronic sensor of any kind is if the normal or non error condition is logic high. I.E. The input to the bob must be logic high and logic low = sensor active. That way under any circumstance if a fault occurs the error is thrown.

In other words as is found here

http://www.cnczone.com/forums/open-s...ml#post1347698

[Falcon69]

hmm, you're right. I didn't think about the whole unit losing power. But, then again, most likely the power for this would be coming from a USB plug from the PC. So, if the power for the PC fails, seems like there's other problems that would need to be addressed.

[RonaldoNZ]

hmm, you're right. I didn't think about the whole unit losing power. But, then again, most likely the power for this would be coming from a USB plug from the PC. So, if the power for the PC fails, seems like there's other problems that would need to be addressed.

I spent far more time thinking about the problem while developing mine than was healthy

Yes hopefully so re the usage scenario, in your own case you outline no problem. Although thats not necessarily the case for others right.
The reason I raise the point is this, and although speculative something to seriously consider I'd think.

My thoughts:

I don't know about US limitations on litigation but I've heard plenty of the horror stories, anyway it occurs to me that if you are selling something that is ostensibly safety equipment, IE intended to insure safe operating conditions of potentially dangerous gear and your product has within it a mode of failure that is known or knowable to the seller (and probably even if it is not) then that seller opens themselves up to serious consequences should loss of property/health/life etc occur as a result of failure of said gear to deliver on its purpose. Also as far as I'm aware you can't contract out of such obligations with disclaimers etc. Its all about the pecuniary gain I believe, if you are giving something away and the recepiant suffers a loss significant enough to make them come after you then "I think" you are ok (although even that may not be the case) however if you are selling it...then you are definitely in the cactus. I could be completely off on all of this of course but that is my understanding of how it works here, The US may be totally different. Know a friendly lawyer? no doubt $$$.

Any more familiar with US law care to comment?

[Falcon69]

Thank You RonaldoNZ for bringing to my attention of the power loss to the switch board. I had not even thought of that and you are correct, the machine would have continued to move when it should have stopped had the limit switch been operating.

I've managed to redesign the single switch (no panel LED's) to work in closed or normally open (selectable) and when unplugged, they turn off the LED in the optocoupler, disconnecting the circuit, as it should now. I tested it all on the breadboard, and in series the switches work now, as they should, triggering the optocoupler (turning it off when magnet triggers the Hall Sensor) in series in the Normally Closed position. When the switch is unplugged from power, the optocoupler's LED stops lighting, causing a disconnect.

In Normally Open position, the switches work as expected in parallel, but, in normal machine operation, the BoB's optocoupler's LED is off and when the hall sensor is triggered, it turns on.

So, the Normally Closed option gives the switch the safety factor you were talking about. When disconnected (whether accidentally or purposely), the optocoupler of the breakout board disconnects (LED turns off).

In Normally open mode, however, there is no safety features, as expected. But, the option is there if you prefer to wire the switches in parallel for the limit switches. (like for home switches)

Now, I am waiting on the other parts to try and make this all work with the original intent, of having a front LED panel as well and rear plug in for the control panel. I am also waiting on a bigger breadboard to test it out on.

Falcon

[nlancaster]

Falcon69, question on your limit switch kit. Not clear from the documentation I can found, how many output pins are there? Are individual outputs avialable to use for homing and squaring of a 4 axis gantry machine?

[Falcon69]

Hello nlancaster

Yes, there are 8 outputs, and the ground is common. Basically, it takes the signal from the breakout board, goes through the switch, then through all the Darlington Transistors, and then sinks to ground.

[crob09]

"Falcon69, question on your limit switch kit. Not clear from the documentation I can found, how many output pins are there? Are individual outputs avialable to use for homing and squaring of a 4 axis gantry machine?"

You can connect all the limits to one I/O, if any are tripped the machines will stop working or give you an indication of a problem by moving very slow.

[Falcon69]

if the machine is still moving after a switch has been tripped, even if moving slowly, something is wrong with the breakout board or mach3 configuration. when those limits are tripped, the machine should stop completely or halt the program. Depending on how you wired your system, the router may still be on and spinning, but the motors for the axis should all stop completely.

[nlancaster]

Right, but to square up a gantry on a dual drive machine, you need a seperate input for the slaved axis. That is what I was hoping to hear about your board falcon, I may be ordering soon

[crob09]

if the machine is still moving after a switch has been tripped, even if moving slowly, something is wrong with the breakout board or mach3 configuration. when those limits are tripped, the machine should stop completely or halt the program. Depending on how you wired your system, the router may still be on and spinning, but the motors for the axis should all stop completely.

It depends on the machine, for example the RepRap 3D printer using Sprinter Firmware will still allow the machine to move but away from the limit and not all that fast, it's just an indicator that the Limit is the cause of the recent problem, a broken wire acts as a tripped limit (optics) so for a person building the machine having that Movement-Feature is a big bonus.

Most machines, specially commercial units will not allow for any movement, however for troubleshooting some do.

[nlancaster]

Having a large cnc router keep moving after hitting a limit sounds like a horrible idea. Even reversing movement sounds bad, in that it can ruin a work piece that might be recoverable if the machine just stops.

[crob09]

Having a large cnc router keep moving after hitting a limit sounds like a horrible idea. Even reversing movement sounds bad, in that it can ruin a work piece that might be recoverable if the machine just stops.

Didn't say it was a good idea, just said that some are setup this way.

[nekidfrog]

2 Shielded XLR Male to Female Mic Microphone Audio Extension Cable 25 ft Foot | eBay

would these be ok for cables for doing these homing switches?