Whats good VFD for spindle control with LinuxCNC?

The SPINx1 s a one channel isolated digital potentiometer intended for spindle control. The SPINx1 takes three TTL level signals, PWM, DIR and ENA and has four isolated outputs: CW, CCW, ENA and VOUT. VOUT is a Analog voltage proportional to the PWM input signal and the Analog reference voltage (5 to 18VDC). The SPINX1 outputs 0 to 100 percent of the supplied voltage, its a potentiometer replacement. Input voltage must be from 5 to 15V.

JT

Robotwizard wrote:

Was planning to get a Teco FM50 for my lathe CNC conversion until I discovered that the 0-10 volt speed control input doesn't play well without an expensive DAC board to put in between. Are there any VFD's to look at that would be compatible?

In between what?

Are you wanting a VFD that will run from a parallel port? If not, can you be more specific about what you are trying to achieve?

You might find Modbus to be easier, and rather more flexible. I have found the Huanyang VFDs on eBay to be cost-effective, and they work very nicely with LinuxCNC, the hy_vfd HAL component and a £1 USB to RS485 adaptor.

Robotwizard wrote:

Yes, between the VFD and par port. The FM-50's I was looking at don't have Modbus/RS-485 so thanks for the advice. I haven't been able to find any manuals or specs for the Huanyang products yet tho.

Try this link:
www.jinlantrade.com/ebay/invertermanual.pdf

But you don't need to know most of that if using the LinuxCNC HAL component

If you prefer to use parport control then you ideally need to convert 3 of your output pins to FWD/REV or ENA/DIR and a PWM output for the speed control. You can potentially configure the VFD for 0 to 5V operation and use a simple RC filter to make the parport PWM into a variable voltage, but it is likely to be non-linear and isn't well isolated.

Better than using a simple RC filter would be to use an opto-isolator and then use the 10V supply from the VFD into your RC filter. This gives you the full 0 to 10V range, protects the parport and should be more linear if (and ony if) you select an opto-isolator with both current-sourcing and current-sinking outputs.
One of these will work well
www.farnell.com/datasheets/611920.pdf
Basically you connect the parport to the left-hand side and the VFD 10V and GND to Vcc and GND. Then a resistor from Vo to one side of a capacitor, the other side of the cap to GND and then connect the VFD 0-10V input to the top of the cap too. Choose R and C to suit the time-constant you need. (say 1/20 the PDM rate. (Use PDM rather than PWM for this)). The reason to use this device is that it will actively charge and discharge the cap through the filter resistor, whereas using a conventional 4-pin opto you need a bleed resistor to discharge the cap, and that makes it non-linear.

Or get a Mesa SPINX1 which is meant for exactly this job

Without braking resistors the VFD will still brake to a halt, but the rate at which is can do so without a bus-overvolt fault will be limited.
For many machine tools the deceleration rate without additional resistors will be perfectly adequate.

There is no reason for a VFD to care about mains frequency at all. They can run OK on anything from DC to whatever frequency the switching speed of the rectifier diodes on the input stage an handle.