Machsupport Forum
Third party software and hardware support forums. => CS-Lab => Topic started by: Hood on October 19, 2012, 05:45:06 PM
-
Hi Andrew, I was wondering if there was a feature or will be in the future, in the CSMIO/IP-A to do Index homing on amplifiers that do not support it?
Most of the drives/amplifiers that would be used in a retrofit with the CSMIO/IP-A would, I imagine, be of the dumb kind, in other words it would have been the control that referenced the machine and set the machine coords when the switch, then index, was seen. As the CSMIO/IP-A already has the encoder inputs it should be possible to do this I would imagine?
Hood
-
Hi Hood,
As encoder with INDEX is connected to CSMIO/IP-A then homing on Index is possible without any problems. Moreover it's possible in CSMIO/IP-S too. We have pdf for it (it's for Delta drives as example). Here is the link: http://www.cs-lab.eu/en/artykul-21-CSMIOIPS_Download.html (http://www.cs-lab.eu/en/artykul-21-CSMIOIPS_Download.html)
Is that what you meant?
Andrew
-
This is interesting! However, I am not sure that reducing the encoder PPR to widen the index pulse would be practical in all cases.
Dan
-
Ok not quite clear on the manual but here is what I think you are saying in that document.
You have the axis home switch connected to a CSMIO Input and configure it in Mach for that Axis Home switch. You then take the drives buffered encoder Index output and connect that to an Input on the CSMIO and then configure the plugin to look at that index input. The buffered Index output need divided in the drives software to widen the pulse and that is all?
So if the above is correct then what will happen is Mach instructs CSMIO to home, it moves to the switch, reverses then instead of telling Mach the switch has closed it waits until it sees the index then stops and sets Machine Coords?
If that is correct then that is exactly what I am wanting :)
Only thing is I am not sure what kind of output the Index pulse is on my drives, also as Dan has mentioned you may not be able to divide the outputs on all drives. I am not sure if the older Osai drives I am contemplating using allow that but will look.
Hood
-
Ok had a look at both drives I have, one is the older Osai analogue command drives and they hae an encoder output from them but no means of dividing it, also I think they are just unbuffered outputs and thus will be line driver as that is what the encoder is. So seems that will not be any use on both counts?
The newer drives can have both unbuffered adn buffered encoder output signals and the buffered can be divided so that is fine but the problem is they to are line driver outputs so looks like they would not work either?
If that is the case is there another way this can be done?
Hood
-
Hood,
What do you mean when you say the outputs are line driver? Could you not just use the Z (omitting the Z-)?
Dan
-
Yes Dan but from that Document I understood the Z output needed to be open collector, maybe I am wrong?
Hood
-
Yes, I find the document a bit confusing. They are saying "line driver" and "open collector" output in the same line. I think they were meaning something else. The output spec they provide on the next line is a misleading again - why would they provide the drive's output spec? Would rather expect the CSMIO input spec instead (if anything).
Anyway, I believe it shouldn't matter if the output is open collector as it's only a matter of a pull-up resistor.
Dan
-
Dan
I think what they are meaning is the encoder is a line driver but the buffered output from that particular drive is an open collector output.
On one of my drives I think the encoder outputs from the drive are not conditioned in any way and are just a direct pass through from the encoder and thus will be a line driver. On the other drive I have the choice of unbuffered or buffered but I think both of these will also be line driver outputs.
I am not sure how they would fare being connected to a CSMIO input as I think it is expecting somewhere in the region of 24v but possibly it would work? Sadly I am not great at electronics so I am not sure, hopefully Andrew can clear things up a bit.
Andrew,
you said the pulse needs to be widened by dividing the Index output, how much time is needed? If a drive was not capable of dividing the encoder outputs would reducing the homing speed have the desired effect? If that is the case then it may even be possible for that to be done in the plugin, ie home at a normal rate but have the back off from the switch at a much lower rate?
Hood
-
We have corrected the document. Thank you for your vigilance :)
-
There was a little bug in index connection manual. OCZ output in Delta is not a 'Line Driver' but 'Open Collector'.
Index pulse must be widened only for IP-S. In most cases servo drives has separate settings for encoder output and electronic gearing, so we configure Delta to 10000 step pulses/rev and encoder output to 500 pulses/rev then we have precise index homing without losing resolution.
Currently IP-S need index signal to be active at least 1ms. Decreasing homing speed makes index wider, so if it's not possible to change encoder output pulse number then try decrease homing speed.
Move from homing switch is already much slower, so even for 10000 pulses/rev or something like that index homing should be possible with sensible speed.
However I'm planning to improve this function with a little help of FPGA chip :) Then the problem should practically disappear.
Note that all of above is only for IP-S.
IP-A has fast differential encoder inputs and there are no limitations or special requirements for index homing.
Andrew
-
Andrew, thanks for the clarification.
That seems good news at least with the IP-A if I am understanding correctly as it means it can handle a line driver Index input and no need to divide the encoder output, hope that is what you are meaning? If that is correct it means I can use the older drives I have sitting here, worst comes to the worst I have more modern drives I can use.
The only problem I see for me using an IP-S is the open collector requirement, is there a way I can convert a line driver output to open collector so that I can then input to the IP-S? I am interested in this because I am presently fitting the IP-S to a wee lathe and the drives do not have homing features in them like my other machines have.
Hood
-
Was just thinking about this, on the Bridgeport when I used the Scorpion controller I had to connect a MPG via a board to convert it to 24v as the inputs were all 24v. It worked well so that should allow me to use the Index pulse from my drives and feed them into the + of an input.
Hood
-
Hood,
And how it's done in their example with the Delta drive? Seems they're feeding the 5V signal directly to the controller? Don't think the drive outputs 24V for the encoder signals.
Dan
-
Dan, from my understanding the drive in their example has an output for the Index which is open collector. This connects to the Negative terminal of your chosen Input on the CSMIO and you connect 24V+ to the +ve terminal for your chosen CSMIO input.
Hood
-
Ah... OK. Makes sense now. What was that board you used with the Scorpion?
Dan
-
It was a driver board I had that Ed Gilbert (CNC Building Blocks) used to make, pdf attached.
Hood
-
Thanks.
Dan
-
Hood, yes, with IP-A you don't need to do anything. You just connect index differential pair from drive to controller and that's all.
With IP-S I agree that sometimes opencollector requirements can be a little bit problematic. If driver does not have opencollector output the simplest way is to use optocoupler.
However we're preparing some useful converters and they soon should appear in our web store.
Andrew
-
Good news all round :)
Hood
-
TechCNC,
The CSMIO/IP-A looks good, but I was wondering about the manual. Any estimate of when it will be available?