On Mon, 2007-09-10 at 22:28 +0000, Neil Buck wrote: > I'm half way through retrofitting a Cincinnati VMC with a MESA m5i20 setup. I > would welcome some advice as to the best way to control the spindle. My goal > is > to be able to do rigid tapping and fully automatic tool changes. My thoughts > are > that since I am planning on adding a 4th rotary axis very soon I would like to > not use the remaining spare axis on the m5i20 card. So I was thinking about > using one of the digital outputs via a PWM HAL component. Then filtering the > output with a simple RC low pass filter to provide the analog 0-10V that the > spindle drive requires. The direction control is trivial using 2 relays driven > from a couple of digital outputs. I wonder would this setup provide a command > signal accurate enough for rigid tapping - or would it even matter given that > the spindle generates an index pulse. Incidentally the spindle also has an > output that indicates when the spindle is upto the requested speed, I was > wondering whether it's worth using it or whether it's overkill. ... snip > Neil
I have limited knowledge and experience with electronics, but here is some of my recent experience with my spindle VFD. Just for testing, I used a parallel port out put with an RC filter and found that to smooth the signal, the capacitor had to be quite large which also slowed the signal response. So, if you command a 100% to 0% speed change the filter output will actually take time to go from 10 Volts to 0 Volts. Contributing factors to the large capacitor are; the parallel port's PWM frequency is fairly low and the VFD input is a high impedance input. My test actually worked okay but it seemed that I could not make speed changes fast enough to get the VFD to utilize dynamic braking. I was hoping for an order of magnitude better performance, so I didn't pursue this any further. (Are there better "active" filters?) Currently, I am using the DAC option available for my controller and "it just works", so there isn't much to say about it. Without a braking resistor, I got VFD fault-outs when trying to get one second stops from full speed. With the resistor, these stops where not a problem. Be sure to consult your VFD manual when sizing the resistor because you can easily fit a resistor that will overload the VFD's braking circuit. For direction control, I use two parallel port outputs connected to a pair of opto-isolators wired to sink the VFD's inputs. My VFD's input sink option provides the pull-up resistor. This setup has worked well. (Now that I have reread your post, I am thinking that a digital output PWM frequency from the 5i20 might be much higher, making an RC filter usable. I think there is some 5i20 PWM information here on Ander's website : http://www.anderswallin.net/2006/06/first-steps-with-brushless-servodrive-microchip-dspic-irf-irams/#more-62 ) I haven't finished my conversion yet, and have not done much testing or any threading or tapping. As far as the "up to speed" signal - for tapping, the active axis is electronically geared to the spindle, so no matter what the spindle is doing the axis will try to follow it. I am thinking that the "up to speed" signal might be used for fault sensing for spindles without encoders, but I don't think any work has been done on EMC in this regard. If you have pictures or specifications on your machine, please post a link. Kirk Wallace ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
