Mach 3 VFD speed control

[JpBinc]

Hello all
Can anyone tell me where I can find a good schematic of a vfd spindle speed interface controler for Mach 2 or 3. I have converted an X3 to a belt drive ac motor and this is the last part of the conversion. Now I know you can buy them but am kinda a homebrewer. I have burned quite a few boards with pics, dac's and adc's etc.. but am not sure how to go about converting the digital io's / pwm to my hitachi sj200 vfd. It looks like there are two different methods for controling output speed/freq. 0 to 9.8 volts @ 10k or 4 to 19.6ma @ 250 ohm. Since I am not use to working with variable current designs. I think that the variable voltage route is the way to go, but I am open to any suggestions. Maybe even a microchip pic pwm capture and compare to dac would work. Any ideas...
Thanks
JpBinc

Mills specs...
X3 cnc'd, steppers, closed loop, ballscrews, gecko's, converted to belt drive spindle 3 phase 1.5hp metric 4 pole motor. 480ipm xy 300ipm z

[dareposte]

I am working on a circuit for my Hitachi X200 VFD. The X200 takes a 0-10v signal input to accelerate across a programmed range of speeds. I am using EMC2, which like Mach3 only takes digital pulse IO.

My plan is to use a PIC to take a pulse train (PWM) out of EMC2, and use the IO on the PIC to program a variable resistor. With a 10k variable resistor, a 14.7K fixed resistor, and a 24VDC power supply then you can get 0-10v across the variable resistor in a fairly fine step. The other option would be an el-cheapo D/A chip like one made by Analog Devices, or even a RC filter and Op-Amp has been used.

The RC filter and Op-Amp is very simple, basically converting a PWM duty cycle into a fractional voltage by charging a capacitor, then buffering the output using an Op-Amp follower. If you search around on Google there are a couple of examples if I recall.

[JpBinc]

Thanks for the input. I have looked at digital pots like the one offered by dallas semi before. Not sure of the load they can drive. The input impedance of the sj200 is 10k. I'll have to check the specs. I like the op amp idea. I seems like it would be smoother output since it does not have the overhead of a pic.. I have never found a good way to use a loop command to keep checking if the driving signal (input) has changed without a small interuption in the output wave form. I guess a higher end pics may have hardware pwm input but not again not sure..

[dareposte]

Here's a link to the Op-Amp circuit I was recalling.

http://controls.ame.nd.edu/microcont...in/node40.html

It shouldn't be too hard to breadboard and get it working reliably before making it permanent with just a trial/test approach using standard components. Obviously you'd have to adjust the op-amp gain settings pretty carefully to get a full-scale swing from 0-10v, and watch carefully for any jitter in the output as it would most definitely not be good for surface finish. Maybe add in a low-pass filter on the output stage of the Op-Amp to help tame any positive feedback that develops.

I was reading my Hitachi X200 manual, and noticed it can receive MODBUS commands. This brought up the idea to maybe use a PIC to decipher a PWM output from the control software, and generate a modbus compatible command for the X200 using the usart. I'm not sure what the latency would be, but it might not be too bad. Modbus is pretty easy to use, too.

I just got a 4U rackmount case in, so I will begin the new wiring soon. I have the parts on hand to build an op-amp follower type circuit, so I will probably start with that to see how it goes.

Best of luck, and thanks for the ideas.

[JpBinc]

I found this on the net today. It opto isolates and uses r/c to output 10v looks good. The author says it works too. I am going to investagate a little more and maybe build one. Then set it up on my oscope to see how well it works.. It just can not be that simple.
http://www.9w2bsr.com/2008/01/pwm-to-analog-update.html

[JpBinc]

I like the op amp schematic.. I just wonder how stable it is. I have had some op amps do weird things without any reason. ( maybe bad design on my part ) I think I will bread board it with a opto front end to protect my computer. Thanks for the suggestions..

[JpBinc]

I will have to look into that. It would solve some of the inherent problems of converting pwm..

[JpBinc]

I now have a pwm to analog schematic that has proven to work. It is based a design built around a 4011 Nand gate ic. The op amp idea was good but finding one that really went rail to rail proved to be a problem. I will post a schematic on my web site.

[Al_The_Man]

You could try feeding the op amp output into a H11F1 opto, the ouput of this opto is a variable resistance that varies with input voltage.
Al.

[srikrish]

Hi
I am Srikrish R from India, I am new to Mach3, I would like to know the following , Please send me the details for the following :

1) Whether I can use VFD drives to control the X and Z axis for my lathe machine, as the VFD drive accepts 0 - 10 VDC analog voltage to control the drive, and i am having the circuit to convert STEP pulse from the LPT port to 0-10VDC analog voltage. in this case whal will be the accuracy.

2) Whether I need to use any feedback device like encoders or Linear scales.

3) If i use stepper motor with stepper amplifier, whether i will get good accuracy.

please send me the above details immediately my email id is : rsrikrish1984@gmail.com