Hi,
I do not have a Hicon so what follows is my best guess, there would need to be several confirmations of various features
before this could be considered practical.
In the diagram attached note that the Hicon unit (which includes the add-on boards) receives Step/Direction input from Mach and compares
that to the Linear Encoder feedback. The Hicon would be required to process that via its PID processor and output an analogue voltage
to your servo/drive. The servo/Drive would have to be in Velocity mode. The rotary encoder of the servo would still be hooked up to the drive
per normal, its just that it is not used as a positional measure but rather velocity measurement. The output shaft is connected to the axis and finally
the linear encoder is connected to the axis and measures the table position.
This arrangement is shown with the Hicon but could equally describe a Galill controller..
Lets take a few components and look at the requirements to make it work.
The first is the Linear Encoder. It will be hooked to the Hicon so it will require an output that the Hicon can decode, namely A and B quadrature
channels and probably an Index signal. This could be a problem, I have seen linear scales with quadrature output but don't recall seeing one with an Index
signal and yet I suspect that the Hicon requires it. This would best be discussed with Vital Systems.
The second issue, still related to the Linear Encoder is the quality of the encoder itself. Have a look at Hidenhan and Reinshaw websites to see what they
offer, and make sure you are sitting comfortably when you read the price. What sort of corners have to be cut by Chinese (and others) manufacturers
to sell linear scales at the prices they do?
I'm of the opinion that to get decent linear scales, not Heidenhan or Reinshaw, just decent ones is likely to be quite expensive, I'm guessing $200 per axis.
The Hicon itself is not cheap, the basic board is $600 and the addon board and activation cost another $600.
I assume you already have analogue input servos and drives?
What is not certain is that the Hicon can do the PID part without additional on-board programming. The Hicon does offer that low level programming
but could require some very clever programming on your part. There again I would rely on Vital Systems advice as to what is possible 'out of the
box' and what extra may be required to get this setup to work.
Note that this general arrangement is not uncommon, it is uncommon in CNC lathes and mills, but is used in semiconductor wafer processing.
Semiconductor machines require extreme precision. The absolute best (Heidenhan or Reinshaw) linear scales are probably only just good enough.
Some use LVDT (Linear Voltage Differential Transformer) transducers and all the high precision (and tricky) stuff use laser interferometry. Zeeman splitting/stabilization
will get you to about 10nm and iodine cells are required to get you to a few nm.
I make the point that in those situations where linear position feedback is employed the feedback components (scales/LVDTs/interferometers) are very
VERY expensive. While you may not require the sort of precision these technologies seek to achieve I think you can rely of the fact that you will spend
significant money trying to do linear feedback. I suspect rather more than it would cost to fit C5 or better ballscrews.
Craig