Re: Cooling Rutex drives

>
> That's mostly what I wanted to know, how much heatsink to use. I'm
> familiar with at least the basics of thermal design so if I had an
> estimate of the power dissipation of each drive I could figure out a
> reasonable size heatsink.

If you do a little "finger-in-the-wind" type engineering and figure
that each motor will draw an average of about 450W (9A X 70V X .7
duty cylce) and that Rutex motor driver is 70% efficient then 135W of
heat is going to be generated. Given that part of the equation you
can use the heatsink mfg's curves under the conditions you will have
(forced air, free air, etc) to get a determination of the square
inches of surface area you need in a heatsink. This is a rough way
to do the numbers but it is better than holding your tongue on the
unit until you have no feeling! Use some thermal grease between the
Rutex case and the external heatsink to improve the thermal transfer.

>
> > A little attention to heat removal and RFI suppression on the

inital

> > design will save you having to redo it in the future.
>
> Hmm, I would welcome your thoughts on RFI too.

Basic RFI/EMI suppression techniques:
1. Keep leads as short as possible

2. Use Twisted Wires on wires that are low level signals

3. On power runs longer than a few feet use a large filter cap
(1000MFD and up) close to the LOAD.

4. Low level signal wires are the most susceptable to external
RFI/EMI and need to be run in shielded cable with the shield grounded
at ONE END ONLY.

5. Each signal needs its own ground return wire. Have a single
point ground in the Power Supply cabinet that all grounds terminate
at.

6. If you put your motor drive modules in a cabinet away from your
machine consider using shielded wire for the motor leads. Once again
ground the shield at the motor driver module end. Don't run your
encoder wires next to your motor leads, even if both are shielded.
Keep a few inches separation.

7. Keep your machine ground and computer grounds separated. Use
relays or opto isolators for logic signals from the machine (limit
switches, etc) Use a separate isolated power supply for the
sensors/switches.

8. Except on motor driver modules that have a common ground between
the logic and the power circuit and non-isolated inputs, do not
ground the negative side of your motor DC to the common ground in
your controller cabinet.

9. In EXTREME cases find or make a good earth ground close to your
machine and ground the machine to that. Run separate heavy copper
ground wires from the machine chassis to this new ground. If you run
non-differental encoder signals more than a few inches consider low
pass filters on the signal lines. Do NOT make a connection from this
new ground to the PC ground. RFI can creep through on the ground
lines of the AC. A large common-mode filer on the AC line to the
computer (it has internal filters that take care of most noise) can
help but don't count on it to take care of all of the problem.


While some have warned not to use any kind of Plasma Torch but the
Hypertherm because of other units noisy start arc, I have a vintage
Miller 500 running 15ft from my computer and it's next to my 6' X 6'
plasma table. It took some careful attention to noise suppression
but it works fine with both Gecko and Rutex servo drivers and a Gecko
201 on the Z axis.

>

>
> Unfortunately I can't test the drives under actual conditions

becuse the

> machine is still in pieces.

I have several machines in various states of upgrade...I understand.

>
> C|