What's new
What's new

Help with remote speed and direction control for VFD

M.B. Naegle

Diamond
Joined
Feb 7, 2011
Location
Conroe, TX USA
I could use some help from some of you electronics wizards on here.

I'm setting up my 1909 Hendey Engine lathe with a VFD, not for voltage or phase change, but for speed control. The lathe was a very early "electrically driven" machine tool that probably had a DC motor that was controlled by a mechanical selector on the carriage. The original electronics are long gone and were replaced with a 5HP 220v AC motor and a drum switch, but it still has most of the mechanical control parts on the carriage that I would like to put to use. So my solution is to stick with the AC motor but control it with a VFD and tie the carriage control into a small potentiometer.
control handle 2.jpg
The VFD has your standard remote control taps: three wires for a 1K potentiometer (common, high, low), and three wires for directional control (common, forward, reverse). Where I need your help is controlling it all with a single dial.:)
control concept.jpg
What I would like to do is have the lathe be off when the dial is in the middle of it's travel, speed up forward when turned to the right of center, and speed up reverse when turned to to left of center. What components would I need to do this? A dual range pot and a rotary 3 position switch? If it takes multiple switches, they'll all need to be able to connect to a single shaft. Is there a way to wire a standard pot to do this or is there a commercially available switch that does this? I'v considered making a cam with roller limit switches to control which side of a dual range pot is in use, but I wonder if I'm overthinking it?

From a practical stand-point, the goal is to have full control of the spindle at the carriage. The speed control would be helpful to adjust surface speed when cutting faces (speeding up as cutter gets closer to center).
 
A WORTHY undertaking!

I have not seen a "Center tapped" pot in all my days. Does not mean they don't exist.
Such a device stacked with a simple on off switch (with some over travel slots) would do the job.

Hmmm... a pot, a pot,,, my kingdom for a pot (of a certain type).
 
In regards to the directional cam all you'd need to do is get a single pole double throw switch to cut over right at the center of travel - switch from common-forward to common-reverse. You probably want something with a very short toggle "deadband" for this application. Would it be possible to make an eccentric to control a linear slide potentiometer fitted with a return spring?

Otherwise you could get creative with an encoder/potentiometer and a suitable microcontroller. The VFD only needs to see a 0-10V or 4-20mA signal at the common terminal for it's potentiometer, whether that originates from a pot or a filtered PWM output from a microcontroller. The uC would only need a very basic program to serve as nothing more than a mere driver between the mechanical bits on the lathe and the input on the VFD.

Needless to say with this level of fun and games it's worth installing a proper E-stop and a disconnecting means somewhere.
 
Potentiometer, 10KB rotary with center tap - Syntaur
Center tap potentiometer.

I do like the notion of putting a arduino or something of that nature in the control circuit though. Otherwise I think you would need to add the switch as stated. With the arduino or similar you could have it read the pot - if the value is past the center tap in one direction it could drive a wire high connected to your forward...if it was in the other direction it could drive the reverse wire high. I'm not sure if you could send the pot straight to the vfd in parallel for speed control or if you would need the arduino to "forward" the value so to speak...
 
What model VFD? Many can be set up so the analogue input runs both forwards and reverse; you just need to set up a dead-band in the middle.

You will probably want a hardware on/off switch somewhere too.
 
Yes, what model of VFD do you have?

‘Tis is a piece of cake with a Control-Techniques drive . . . as well as any number of other drives. You can even set the CT drive up to disable the drive when the pot is in the mid position if you like.

The CT drive has a complete analog input function block that allows you to sum an offset (negative or positive) and then scale as you see fit to Hz or RPM. This can then be directed to a deadband function block with the output sent to the speed reference.

Then set up the drive enable such that the absolute value of the speed reference needs to be just above zero to enable the drive . . . and then seal the drive enable on when actual absolute speed is greater than zero.

That is the beauty of the CT drive, lots of built in function blocks plus PLC programming functionality built in.

Unidrive M400 VFD| AC Drives | Control Techniques

You may be able to get away with an M200 series as well although I have never worked with anything lower than the M400 series.
 
Thanks for the replies!

The VFD is an new-old-stock older model Ace from Boston Gear (I posted awhile back trying to find documentation, to no avail). It does work, but I'd like to keep the design open enough that if I blow it up, I can swap it with any other 5HP VFD that has basic remote connections.

I do have a standard 1K pot to play with and plenty of micro and toggle switches. My current plan involves a couple switches and a simple cam, similar to the joy-stick pictured above, and there would be two NO/NC switches that switched reverse/stop/forward, and another NO/NC switch that flipped which side of the pot was in use. Even with a dual range pot however, the motion of the switch is such that it will only see half of it's travel in each direction. I don't think the current VFD has a way to tune the minimum or maximum speed settings, so the center tapped pot may be the best solution.

If I could find some kind of rotary dial 3 position switch with a through hole, that would make things much easier. That would save me from having to engineer a cam and switches. I can get a three position, four pole wafer board switch that would do the job. Without a thru hole in either switch, I'll need to make some kind of linkage or offset shaft to move both with the same motion.
 
Potentiometer, 10KB rotary with center tap - Syntaur
Center tap potentiometer.

I do like the notion of putting a arduino or something of that nature in the control circuit though. Otherwise I think you would need to add the switch as stated. With the arduino or similar you could have it read the pot - if the value is past the center tap in one direction it could drive a wire high connected to your forward...if it was in the other direction it could drive the reverse wire high. I'm not sure if you could send the pot straight to the vfd in parallel for speed control or if you would need the arduino to "forward" the value so to speak...

I've never worked with a center-tapped pot. I've wired pots for VFDs and DC speed controllers that had two speed setting via two pots, and a toggle to switch which was in use by either a single pole switch switching which wiper was in use, or a double pole switch switching which hi/low wires were in use.

Am I correct to assume that with a center tapped pot, if you have three pot wires for forward and three pot wires for reverse, the wiper and low wires are shared, while there are two separate high wires? If that's the case, I could use a two pole three position switch. One pole would tell the VFD which direction to go and the other would switch which side of the pot is in use.
 
You'd want the two end leads of a 2k pot connected to V+ and the center tap at V-. The wiper can then select from 10V-0V-10V. Get a DPST switch to cut over from common-forward to common-reverse halfway through the travel (or two SPST) and bob's your uncle.
 
VFD.jpg
20210119_094421.jpg
20210119_094436.jpg

Here's the VFD. I know it's old, but it does work, and if I can get the remote control right then replacing it in the future won't be a problem.

control tube.jpg

For reference, The original mechanical lever on the carriage turns a pair of mitre gears that slide along a keyed shaft. That shaft drives a sprocket and chain. The original control was mounted under the chip pan on the left side of the machine, but is now missing. I'm using a section of pipe with some dome end caps to replace it (mixing function with aesthetics). The bottom sprocket will mount on the right side of the cylinder. The sprocket, cap, and electronics will be removable from the right side while the wires will go through the left side to connect to an electrical cabinet in the back that will house the VFD.

Back-lash in the mechanics of the system will also be a limiting factor, but I plan to replace the keyed shaft as it has some wear, and I may also add a ball-detent to the mechanism on the control side so there's some firm feed-back going in and out of "off" position.
 
You'd want the two end leads of a 2k pot connected to V+ and the center tap at V-. The wiper can then select from 10V-0V-10V. Get a DPST switch to cut over from common-forward to common-reverse halfway through the travel (or two SPST) and bob's your uncle.

That makes sense. The center tapped pots I'm seeing all have four terminals, so your wiper and V- wire straight to the VFD's terminals, while the VFD's V+ is toggled between V+ (fwd) and V+ (rev).

Does it have to be a 2K pot? There dosn't seem to be many center tapped pots out there and the closest I'm seeing is a 2.5K.
 
Forgot to add too:

The machine will have a main power disconnect to turn everything on and off. For a safety, I was also going to add a foot pedal along the front of the lathe that would kill power to the VFD via a contactor. It might also be a good idea to add another switch to the VFD control that would keep that contactor from coming on unless the speed/direction selector was at the off position. That way it doesn't potentially spin-up as soon as you turn on the machine. If you hit the pedal while the lathe was running, it would stay off until the selector went back to "off", so separate start/stop buttons for the operator to use shouldn't be needed.
 
Correction, 4k. The pot needs to be approximately quadruple what the drive normally calls for since you're only connecting 1/4 of it's total resistance between V+ and V-. 1/2 of 4k is 2k between one side of the pot and it's center tap. Divide that again when you connect the other half of the pot in parallel to get 1k.

The resistance doesn't need to be exact but you want to stay fairly close to the manufacturer's spec because:

1. Too little resistance could overload the small DC power supply designed to be sufficient for a 1k pot.
2. Too much resistance will result in insufficient current to the analog input and cause wonky non-linear speed control.

A 5k would probably be perfect.

EDIT:

Another thought here...

If it would be simpler to implement, you could stack a second (normal) potentiometer into your setup and configure an op-amp circuit to perform the cut-over between forward and reverse. No need to machine cams or mounting fixtures for microswitches. Handle a very precise, repeatable center calibration electronically instead of fiddling with mechanical switches and their dead-band.
 
Another option might be to use the original control for a digital (on or off) increment/decrement speed control. This would mean having
a separet on/off switch, which I guess you are trying to avoid.
 
Correction, 4k. The pot needs to be approximately quadruple what the drive normally calls for since you're only connecting 1/4 of it's total resistance between V+ and V-. 1/2 of 4k is 2k between one side of the pot and it's center tap. Divide that again when you connect the other half of the pot in parallel to get 1k.

The resistance doesn't need to be exact but you want to stay fairly close to the manufacturer's spec because:

1. Too little resistance could overload the small DC power supply designed to be sufficient for a 1k pot.
2. Too much resistance will result in insufficient current to the analog input and cause wonky non-linear speed control.

A 5k would probably be perfect.

EDIT:

Another thought here...

If it would be simpler to implement, you could stack a second (normal) potentiometer into your setup and configure an op-amp circuit to perform the cut-over between forward and reverse. No need to machine cams or mounting fixtures for microswitches. Handle a very precise, repeatable center calibration electronically instead of fiddling with mechanical switches and their dead-band.

Ok, I think I've given some mixed info. My 1000ohm potentiometer figure was old notes and I'm not sure where i figured that from. I went back to the VFD connection drawing (sorry it's so blury) and it notes a "0-5v, 0-10v, 4-20ma" potentiometer, and that there should be a Jumper between a couple terminals for a 4-20ma pot.

Looking for pots, I'm also seeing lots of Balance potentiometers, some with a center detent. These have six connections as they're effectivly two potentiometers stacked together, but with their ranges off-set so zero is in the middle. I think these would work as well, but the required Ohms would be doubled, not quadrupled, correct? Or would they have factored this in when it was rated?
 
I've never worked with a center-tapped pot. I've wired pots for VFDs and DC speed controllers that had two speed setting via two pots, and a toggle to switch which was in use by either a single pole switch switching which wiper was in use, or a double pole switch switching which hi/low wires were in use.

Am I correct to assume that with a center tapped pot, if you have three pot wires for forward and three pot wires for reverse, the wiper and low wires are shared, while there are two separate high wires? If that's the case, I could use a two pole three position switch. One pole would tell the VFD which direction to go and the other would switch which side of the pot is in use.

Looks like you got your answer further down the thread - I'm interested in how it turns out from a curiosity point of view :)
 
Ok, I think I've given some mixed info. My 1000ohm potentiometer figure was old notes and I'm not sure where i figured that from. I went back to the VFD connection drawing (sorry it's so blury) and it notes a "0-5v, 0-10v, 4-20ma" potentiometer, and that there should be a Jumper between a couple terminals for a 4-20ma pot.

Looking for pots, I'm also seeing lots of Balance potentiometers, some with a center detent. These have six connections as they're effectivly two potentiometers stacked together, but with their ranges off-set so zero is in the middle. I think these would work as well, but the required Ohms would be doubled, not quadrupled, correct? Or would they have factored this in when it was rated?

Hoping someone else knows because you went beyond my skill level - you lost me on 4-20ma rating on a pot not sure what that is....I'm used to ohms, k, and mega ohms...
 
Same here, lol. I'm researching how to convert what's written on the VFD into a "xxxx ohm" rating. I think I had gotten the 1000 ohm figure from a similar sized VFD, but now I want to be certain it's the correct pot.

At this point, I think the theoretical design and wiring is pretty straight forward. I just need to confirm and purchase the actual components and make the brackets to link it all together.

When it comes to electronics, most of my experience is in terms of creating a simple circuit, or building/repairing things based off of a shopping list of materials. I'm still filling a lot of holes in my knowledge.
 
There's three types of analogue signal common in automation (and a few others somewhat less common):

Voltage source: Something outputs 0-10V to be read by your controller. Pots kind of act like this as long as the current through the pot (determined by the pot's resistance and the voltage across it) greatly exceeds the current drawn by the input pin. Many generic sensors (e.g. pressure, flow, gas detectors) output 0-10V.

Current source: Sensor outputs typically 4-20mA regardless of voltage; the controller is expected to act as a near-short or low-ish resistance. Much more robust against noise and long cable lengths. More $$$.

Resistance - Thermistors generally. Cheap as chips, middling accuracy. Generally found in the HVAC world not in VFDs. The controller needs to output a voltage and see how much current the thermistor draws or vice versa. Pots are normally not read this way.
 








 
Back
Top