[ger21]

How do you initiate the hard stop home sequence? Just a simple signal to the Clearpath drives?
I'd just do the hard stop routine, followed by the home all button in UCCNC.
No, you don't need switches. With no switches configured, Machine coordinates are set to zero at the current location.

You could probably combine them into one operation, but you'd have to put a lengthy wait before the Home All command, because UCCNC would not know when the hard stop homing was finished.

[1Jumper10]

Hi Gerry - There are 2 ways to initiate the hard-stop home with the CP's and it is set in firmware; 1) home on first power up, and 2) home on first enable. Obviously home on first enable requires UCCNC to be up and running.
When the CP's home they hit the hard stop and then back off a distance that you also set up in firmware. I could have them back off to the home position I guess? Except if they hard stop home at the far end of the table and I have them back all the way off to the front left corner, that's a lot of uncontrolled motion that I cant stop if there's a problem. The E-stop switch works with the software and during this movement the software is oblivious to whats happening.

[1Jumper10]

The E-stop cuts the "enable" signal. If I set them to home-on-first-enable the E-stop switch might stop them if thers a problem?? I also could have them hard-stop-home to the front of the table and this would be less of an issue. A little but harder than doing it the way I had planned but not insurmountable.

[ger21]

I'd have them hard stop home on first enable.

Not sure how you have your enables wired up or configured?

But you'd power up the machine, then Reset in UCCNC should enable the drives, and it should do a hard stop home. Only back off a small amount, say, where your limit switches or softlimits would be?
When it finishes, just click the Home All button in UCCNC. What I would do is set a Home Offset in UCCNC.

So, say your hard stops are at 97". Hard stop home, and back off 1". Then set your Home Offset in UCCNC to 96, and when you press Home All, it will set the current machine coordinate to 96.

THe one issue I see, is how to not trip the homing switches during the hard stop homing?

[1Jumper10]

That would work.
The homing switches will never be hooked up if UCCNC doesnt need them and the CP's handle the homing via hard stop.
Also, UCCNC ignores any home switch input unless its homing and since the CP's home independently of the software it wouldnt matter if it did trip one. I think??
This method seems less complicated then placing and configuring switches so it should be more reliable. Thanks Gerry, you've been a great help.
Anyone want to buy some almost brand new Pepperrel and Fuchs proximity switches :tired: ??

[ger21]

I meant to say limit switch, not home switch. Sorry.

[1Jumper10]

No limit switches either. The max travel limit away from the hard-stop home position is set in CP firmware. I might run with one home sensor and limit switch on the Z axis just for insurance. If the shaft coupler were ever to slip, the limit switch would warn me before the servo could try to drive the axis past it's travel limit. Other than that I don't think any switches are necessary?

Sent from my Pixel XL using Tapatalk

[1Jumper10]

Got the CP's mounted to the gearboxes. In checking them over after they were mounted, I noticed that one gearbox was slightly easier to turn than the other 2. I put this combination on the X axis so the units on the gantry would be as identical as possible. I may swap in one of the gantry drive units on to the X axis for auto tuning. I don't know how much difference it would make to the auto tuning routine. But I want it to duplicate reality as closely as possible. My inertia mismatch is only 5:1, so the slight difference in efficiency between gearboxes might not matter.


Sent from my Pixel XL using Tapatalk

[NIC 77]

Anyone want to buy some almost brand new Pepperrel and Fuchs proximity switches :tired: ??

Woah, those aren't cheap. $30 each + shipping and tax. You've got me thinking about prox switches now. How are they wired? They need to be powered with 10 to 30 volts? So what is that, 4 wires to each, or ??? Power and ground, then two for the circuit for the switch, no, it must be another way?

I have only used the mechanical switches, also from CNCRP, where it actually closes a switch, allowing current to flow through the two wires.

So these prox switches are 3 wire?

You should probably hold onto them until you've got everything working how you want it. Too expensive for me, I am poor.

I did just now do a search on EBay, I wonder what the quality is like between these switches and ones I can get on EBay 10 for $25 USD?

[1Jumper10]

These are it here: Pepperl + Fuchs Proximity Sensor Kit | CNCRouterParts
I'm going to keep them for now in case I cant get the Hard-stop homing feature to work. They are wired through the cable chains but I'm not going to terminate them on the MB2 unless I end up using them.
They are 3 wire, 12 -24vdc, inductive (?) proximity sensors. You can probably find similar sensors on E-bay but these are really high quality units. I was impressed with the sensors, cables and the connectors. Very well made. In this case, I did get what I paid for. I'd sell them for $100 shipped in case you were interested and I ended up not using them.

[NIC 77]

These are it here: Pepperl + Fuchs Proximity Sensor Kit | CNCRouterParts
I'm going to keep them for now in case I cant get the Hard-stop homing feature to work. They are wired through the cable chains but I'm not going to terminate them on the MB2 unless I end up using them.
They are 3 wire, 12 -24vdc, inductive (?) proximity sensors. You can probably find similar sensors on E-bay but these are really high quality units. I was impressed with the sensors, cables and the connectors. Very well made. In this case, I did get what I paid for. I'd sell them for $100 shipped in case you were interested and I ended up not using them.

For the last few minutes I've been doing some research on this, I haven't gotten into it before.

So it was a few years ago that I made my control box, I can't remember all I did on it now. I'm pretty sure I'm using Normally Open (NO) limit switches, because if I hit a limit inadvertently, I would unplug the wire from the back of the control box I made to be able to jog it away from the switch. Looking on EBay, there are 2 wire Normally Open Prox switches for sale, 6 to 36 VDC, I'm guessing those could be plug in replacements as far as my control box is concerned, no messing around with a third wire?

Do you remember those little square mechanical switches that CNCRP used to sell? That's what I have. I think I will leave them attached to my first build, and get some 2 wire NO prox switches for this one.

[1Jumper10]

If the switch requires power to operate then its almost certainly 3-wire and I think its more correctly referred to as a sensor; normally Hall effect (senses the presence of a magnet) or proximity (either inductive or capacitive). 2 wire switches are normally mechanical, or reed switches that sense the proximity of a magnet. I've used all types and the hall effect sensors I felt were best suited for use on a DIY router. Current type hall effect sensors are only 2 wire and require a magnet as a target. They're pretty easy to implement but they've become harder to find lately. Standard hall effect sensors require power to operate but they are dirt cheap and are usually very accurate. You'll probably have to make your own housing for them. I think, haven't tested, that mechanical switches are the least accurate type of switch/sensor.

[NIC 77]

I found some two wire proximity sensors on EBay. Not sure what the difference is? I realized I've started to derail your thread going off on a tangent about my own needs, so I made a post here:

http://www.cnczone.com/forums/genera...ml#post2091412

Hoping to find a two wire solution, or I will probably just use the mechanical limit switches.

[1Jumper10]

Pretty frustrating past couple of days. I installed and auto tuned the servo on the Z axis and that went well and is working good. I can jog or send it to positions with the MDI. Same with the X axis. The dual motor Y axis will have to wait awhile yet. This confirms to me that I've connected the step and direction inputs at the board and on the servo correctly. But I've got a problem with the outputs on the MB2. Not sure how but the outputs Y301 - Y304 failed first, followed by the second group Y305 -Y316. They fail to an on state. I could not figure out why my spindle run command was active on the VFD but not in software or the MB2. Started checking outputs with my multimeter and all outputs had -24vdc no matter if the LED on the board was lit or not. I began to realize that I had burned up 4 outputs somehow and shifted my enable circuits and relay output down to Y305 - Y316. Now, they are fried to. I don't think I hooked up anything incorrectly. If my servos weren't working as expected I would suspect I did. Not sure about the path forward from here. If anyone has suggestions I'm all ears.

Sent from my Pixel XL using Tapatalk

[NIC 77]

As you know from my other posts, I am no expert on BOB wiring.

What was connected to those ouputs? Was this because of the VFD connection to the board?

So this is a MachBob2 board?

Is one of these user manuals correct?

Rev 4
http://www.cncroom.com/downloads/MB2...al%20E16R4.pdf

Rev5
http://www.cncroom.com/downloads/MB2...al%20E16R5.pdf

It could be that your VFD has an issue that was the cause of this? Either a fault in the VFD causing too much current to flow, or a miss matched spec, or incorrect wiring? Could be your VFD is missing a resistor?

"Figure 12 shows 14 output terminals, each output can sink current up to a maximum of 100mA with a total maximum of 500 mA per group of 7 outputs, as shown in Figure 1 on page 4"

Do you think it got too much current?

Is there a way to test the current flow these outputs saw using a battery and multimeter connected to the VFD or whatever was attached?

[1Jumper10]

It is the MachBob2. I suspected the VFD at first. Not because it is defective somehow but maybe I had hooked it up backwards. The MB2 uses "sinking " transistor outputs (as near as I can tell but spelled out in the MB2 manual) The VFD has a jumper that you set to accommodate your inputs which would be "sinking". Set it, tried it, but they were probably already toast by then and I never did get it to work correctly. Eventually I gave up and shifted it down to the "CM" and "NO" of the onboard relay and then it was fine. Anyway, the current requirements of the enable circuits and the VFD and a couple external relays in total were far below what the MB2 is rated for : 24 miliamps for the 2 enable circuits, about 6 miliamps for the relays. MB2 is rated at 100 miliamps per output, 500 miliamps total. Both times it happened were right after turning on the power and I think that is what fried them.

I reached out to CNCroom to see if they could provide some insight.

[NIC 77]

How goes the battle? Were you able to get it working with your board through other inputs or are you now looking at needing a new board?

[1Jumper10]

Cncroom has been very helpful. Said the board could be repaired and they offered to send me the components to do it. I thanked them but declined. I asked if they could look over my wiring diagram to make sure I was doing it correctly. Weerasak had me send him the manual for the servos and the VFD and tell me how it should be connected to a UB1 board. It was going to take him a couple days to get back with me and in the meantime I got a new UB1 board ordered and on the way.
I am convinced there were only 2 things that could have fried the MB2 outputs: a sinking-sourcing conflict with the VFD, or a power surge from everything powering up when the contactor was closed. So, new plan is to separate the servo power supply and VFD from the UB1 with a 2 pole contactor controlled by the charge pump output. The UB1 will power up, followed by the software then the servo power and VFD.

It's a helluva way to upgrade my control board to a UB1. Can't say I recommend it.

Sent from my Pixel XL using Tapatalk

[ger21]

So, you were using an output to the VFD fwd input? I only used the MB2 relay for that.
I'm actually using an 8 channel 24v relay board for my contractors coils, so I don't accidentally blow up my board. Since I don't know what I'm doing. 😃

I managed to blow a DYN4 drive the other day, but it didn't damage the UB1. Now I have to send it to Canada for repair.

$$$

[1Jumper10]

Yes. Yes I was

It seemed so reliably simple and it will work that way but it is complicated. I switched it over to using the relay after it was too late for the MB2 outputs and that worked just fine too. I might leave it that way but Weerasak sent me a diagram of how to do it. He also pointed out that the Hitachi could use Y301 for the pulse stream instead of one of the axis outputs. Just assign it as the output for the Step/direction Spindle set up in UCCNC as the motor spindle step signal. Anyway, he's been very helpful and here's the diagram he sent me.