Router related questions.

[jazzcnc]

I don't think it is possible to do this with less than two inputs and two outputs. You have to monitor each home switch independently as it will never be known which will hit first. Based on which is first hit, you will have to disable that driver, and then continue with the second. Using the backoff technique realy only works with different axises. OK, I suppose you could use the backoff technique, but then you are purposely raking the gantry to stop hitting the first home switch.

The solution I offered wasn't a theory I've done it on other controllers so know can be done with 1 input 1 output.
But like I said the switches have to be offset so yes you are racking the gantry very slightly. However, we are talking tiny offset just enough so doesn't trip the second switch.
This method is actually how Cslabs gantry squaring Homing works. It requires one switch is slightly behind the other to work. It then measures the offset and corrects accordingly or to some other figure if the user requires.

[jazzcnc]

I think that you could do this with 1 input and one output using a DPDT relay. One side of the relay toggling between 2 home limit switches, the other side toggling either the enable or step signal, if enable is not available

Relay (output#) inactive
Home (Dual motor axis) as per normal routine.
Back off switch ~.125"
Activate output (relay)
Enable to primary drive is turned off (or step signal if no enable)
Signal from primary home swith open, close to secondary motor home switch
Seek home, apply adjustment
Deactivate relay to primary home/limit switch, toggle on enable or step to primary motor

Gary, This is exactly same as my suggestion except there is no need to back off the first SW1 other than normal backoff. Because the SW2 is offset behind SW1 so doesn't get tripped on the first approach.

Exactly same but doesn't require SW1 be taken out the system while seeking SW2, provided not large offset so racking doesn't trip SW1, but if it does then setting a small backoff to SW1 fixes this.

However, as you probably know better than me this shouldn't be required because can be done thru software by De-coupling motors from slaving and homing again as separate axis. Which is probably how Centroid gurus will do it along with code to measure and apply correction if needed.

This is how I wanted to fix the Cslabs controller but they don't allow access to the De-coupling routine so had to work around it.
If the Centroid allows De-coupling, which think Keith said it does, then if next release is taking too long to appear this is how I will approach making it work for me.

[mael]

Great stuff guys. Thanks jazzcnc.
You guys know better than I do. I cant wait till i get my hands on my Acorn and do the bench tests. Just thinking ahead in the mean time.

Cheers!

[mikes]

I don't think it is possible to do this with less than two inputs and two outputs. You have to monitor each home switch independently as it will never be known which will hit first. Based on which is first hit, you will have to disable that driver, and then continue with the second. Using the backoff technique realy only works with different axises. OK, I suppose you could use the backoff technique, but then you are purposely raking the gantry to stop hitting the first home switch.

The solution I offered wasn't a theory I've done it on other controllers so know can be done with 1 input 1 output.
But like I said the switches have to be offset so yes you are racking the gantry very slightly. However, we are talking tiny offset just enough so doesn't trip the second switch.
This method is actually how Cslabs gantry squaring Homing works. It requires one switch is slightly behind the other to work. It then measures the offset and corrects accordingly or to some other figure if the user requires.

I believe you... So you are using one output to disable the "first" driver (the one with the offset home switch). Is that correct? What happens if the "second" side of the gantry is raked enough that its home hits first, even with the offset? At that point won't the state change be interpreted as the "first" side home being tripped?

May be I am not understanding your solution. However, I am sure you would agree, if I/O was not a limiting factor, you would probably not jump through such hoops.

[DICKEYBIRD]

if I/O was not a limiting factor, you would probably not jump through such hoops.

If a frog had wings...

[mikes]

if I/O was not a limiting factor, you would probably not jump through such hoops.

If a frog had wings...

You know frogs are pretty adaptable. Some are able to change sex, others be frozen and then thaw without issue. I bet we could mess with them a bit and get some wings on those bad boys. May be via their USB port.

[jazzcnc]

I believe you... So you are using one output to disable the "first" driver (the one with the offset home switch). Is that correct?

Yes, output turns on a relay which disables the first drive, however, it's the second switch which is offset backward from first.

What happens if the "second" side of the gantry is raked enough that its home hits first, even with the offset? At that point won't the state change be interpreted as the "first" side home being tripped?

In it's simplest form then yes but it's not difficult to work around this.
2 options really. Simplest is use 2 inputs.
But to maintain the 1 In & 1 Out setup then we would call upon the Relay NO contacts to disable SW2 while on the first run, then when disables the first drive it swaps SW1 for SW2 for second and does a state check. If high back off until low and vise versa.

May be I am not understanding your solution. However, I am sure you would agree, if I/O was not a limiting factor, you would probably not jump through such hoops.

Agreed more is always welcome, But the problem isn't limited I/O because would still need 2 x inputs for slaved axis anyway. Plus like I mentioned above this is do-able thru software if you are capable of creating/modding macros. But all your saving is 1 x output.
However the guys are working on it so hopefully, this won't be an issue for long.

[jazzcnc]

I want to build a 4x4 router. But I am going to likely go with one screw right up the middle for the Y axis in order to avoid the issues you have. Heavy linear guides and trucks on the sides.

Don't do it, Marty, it's not worth it. Twin screws are much better on a wide gantry and provided you use Stepper drives which have Stall detect feature then there is nothing to worry about. Anyway, correctly set up and tuned stepper based machine should rarely losses steps so not an issue.

Leadshine AM or EM series Digital drives are excellent for routers and feature stall detect. AM882 which are classed as superseded now by EM series now but still made and easily available are much cheaper than EM and work just as good.
I've fitted 100s over the last 4 years and super reliable only ever had 4 dead units. Also very well protected and tolerate voltage fluctuations very well which I find is what kills drives the most.

[martyscncgarage]

I want to build a 4x4 router. But I am going to likely go with one screw right up the middle for the Y axis in order to avoid the issues you have. Heavy linear guides and trucks on the sides.

Don't do it, Marty, it's not worth it. Twin screws are much better on a wide gantry and provided you use Stepper drives which have Stall detect feature then there is nothing to worry about. Anyway, correctly set up and tuned stepper based machine should rarely losses steps so not an issue.

Leadshine AM or EM series Digital drives are excellent for routers and feature stall detect. AM882 which are classed as superseded now by EM series now but still made and easily available are much cheaper than EM and work just as good.
I've fitted 100s over the last 4 years and super reliable only ever had 4 dead units. Also very well protected and tolerate voltage fluctuations very well which I find is what kills drives the most.

The rouiter I am working on currently is a Phoenix/Magnum 4'x4' CNC Router driven by Baldor NEMA 34 DC brush servos. It is hell for stout. Heavy steel tubing 1-1/4 round linear rails. I plan on using 30mm Linear guides and trucks. I am planning on using servos....
Marty

[Gary Campbell]

I want to build a 4x4 router. But I am going to likely go with one screw right up the middle for the Y axis in order to avoid the issues you have. Heavy linear guides and trucks on the sides.

Don't do it, Marty, it's not worth it. Twin screws are much better on a wide gantry and provided you use Stepper drives which have Stall detect feature then there is nothing to worry about. Anyway, correctly set up and tuned stepper based machine should rarely losses steps so not an issue.

Marty...
I agree wholeheartedly. Single drive = 2' wide max