PMDX 126 & 108 - Spindle Encoder

[spumco]

Having a problem as I go through the process of designing the control system on a 4-axis mill (retrofit).  As I'm sure you're used to hearing - I'm a novice at this stuff so electronic terms may go sailing over my head.

Current hiccup is that I'm trying to get the spindle encoder signal to both my VFD and the control software.  The VFD needs feedback for low-speed accuracy and improved torque, and the software needs it to do rigid tapping (via electronic gearing between spindle and Z-axis motor).  I'd like - but don't need - to send an index signal to the software as well in case I fit a tool changer that has to be homed.

The WJ200 VFD accepts a 2-channel single-ended signal (A, B, GND).  The A-terminal can be 5-24Vdc, but the B-terminal must be 24Vdc only.

The software and motion controller (UCCNC/UC400ETH) can monitor 2 pins for a spindle encoder signal (A, B), as well as an index signal.

My spindle motor encoder is a BEI quadrature with index (A+, A_, B+, B_, Z+, Z_) and operates at 5-24Vdc.

Other inputs (not including estop, resets, enables, etc):
3/ea NPN-NC homing (10-30Vdc)
1/ea NPN-NC limits (6/ea 10-30Vdc, series)
1/ea 24Vdc VFD alarm signal
2/ea touch probes (future)

Although I already have a two-port BOB, it is restricted to 5Vdc only inputs.  This is going to be enough of a hassle to attach the 24Vdc inputs that I'm happy to spend money on a PMDX BOB that makes wiring/integrating easier.

It appears that a PMDX 126 plus a 108 input board will provide me with the inputs I need/want for current and future expansion, as well as making the wiring easier for homing/limits.  If not, please let me know. 

Spindle control will be via a 106 board I bought from you a couple months ago based on a plan to try to integrate the OEM control hardware with the VFD.  That plan has been abandoned, but the 106 appears to be the ticket for spindle 0-10Vdc control and I have it in hand.  I don't see a need to buy a 107 (or something else) if the 106 will do.

My questions
1. How do I split the quadrature signal from the encoder to both the VFD and the 126/108, and is there some reason I shouldn't do this?  I don't think any small timing difference between the two signals will cause issues as the VFD and UCCNC shouldn't be competing.

2. Can the 105 translator board take the quadrature signal and change it to single-ended?  It appears that this is the case, but I'm trying to double-check to compensate for my total lack of experience with this stuff.

3. Assuming that splitting the signal isn't a bad idea, should I split the quadrature signal, or split the single-ended signal?  i.e. two 105 boards and some sort of splitter upstream, or one 105 and a single-ended signal splitter? My guess is that the less single-ended wiring I have, the less chance of noise interference and placing the 105(s) as close to the VFD and 108 is preferred.

[The VFD and low-voltage hardware are in different enclosures, so there will be 24"-36" of cable from the encoder termination (105 or splitter) in the low-voltage enclosure to the VFD.]

4. I'd like to keep as much in the control system at 24Vdc as possible.  Can the 108 take a 24Vdc input (presumably from the 105) rather than me trying to figure out how to drop it to 5Vdc? Or does the 105 drop the outputs down to 5Vdc? In which case, how do I maintain the 24Vdc signal to the VFD?

Enough for now.  Thank you in advance for any assistance you can give me with this.

Regards,
Spumco

[Bob at PMDX]

The WJ200 VFD accepts a 2-channel single-ended signal (A, B, GND).  The A-terminal can be 5-24Vdc, but the B-terminal must be 24Vdc only.

I can't find where the "A" terminal (I presume this is the "EA" terminal) is 5V only.  The example wiring diagram on page 20 of the "Quick Reference Guide" that I found says "EA" is 24VDC, and the table on page 21 says "EB" is 27V max (so I would presume 24V is OK).

Something to watch out for is the "EB" input is 2 KHz max.  This may be an issue depending on what your encoder resolution is and what your max RPM is.

The encoder connections to the PMDX boards must go in the PMDX-108 if they are 24V signals.  The PMDX-126's inputs can handle 24V, but only if the 24V is from a PNP-style interface where it is either floating or connected to +24V.  The encoder's outputs are most likely totem pole outputs which will not work with the PMDX-126 when they switch from 0V to +24V.  That means they will appear on pins 2-9 on the 2nd parallel port.

Now on to your questions:

(1)  You don't *need" to use the differential signals from the encoder.  You can simply wire it up as a single ended interface and leave the "A-", "B-" and "Z-" terminals unconnected.

If you are worried about loading down the encoder outputs, you may be able to connect the "A+", "B+" and "GND" to the VFD inputs, and the "A-", "B-" and "GND" to the PMDX-108 inputs.  If you do this, you may need to invert the encoder direction or encoder input polarity.  You can also connect either the "Z+" or "Z-" (either one but not both) to the PMDX-108.

(2) No, the PMDX-108 (nor the PMDX-105) cannot translate differential to signle-ended.

By the way, if you really DID mean to type "PMDX-105" in a couple of places, the PMDX-105 will not work with your encoder as the PMDX-105 has slower opto isolators on it and most likely won't pass the encoder signals through at higher RPM.

(3) I at all possible you should use a shielded cable for the encoder signals.  Use wires for the A/B/Z and GND connections, and tie the shield to GND **AT ONE END OF THE CABLE ONLY!!!**.  Since none of the PMDX boards do differential to single-ended conversion, you will have to run single-ended signals from the encoder.

(4) Yes, the PMDX-108 can take totem-poll 24V input signals and convert them to 0-5V.  The PMDX-126 cannot handle totem-pole 24V inputs as I mentioned above.

Any clearer?  Or have I confused you even more?

Bob

[spumco]

Bob,

Thanks much - quite a bit more clear.  I'm still confused, but less so.

The quick ref guide doesn't identify it, but the main VFD user manual indicates "EA" takes a 5-24Vdc signal on page 3-114 in the table.  "EB" is 18-24Vdc only.

I have a 2048ppr encoder, so I'll have to figure out something.  I plan to spin to motor up to 5kRPM, so that's something like 170KHz - too much for even the "EA" terminal.

I found that Mesa Electronics has two very inexpensive items related to this particular problem.  They have a 'ENCY' board that is a differential signal splitter.  One in, two out - that should allow me to get noise-resistant differential signals very close to both the VFD and the 108 board before converting to single-ended signal.

Second, they sell a '422ENC' board that appears to convert differential RS-422 signals to TTL.  One of the ENCY and two 422ENC boards appear to be the ticket.  I'm awaiting a response from them as I'm not sure the 422 board can handle 24Vdc inputs.

I did not realize that I could send the A+B+Z+ signal one place, and the A-B-Z- to another component.  Interesting, as the UCCNC software permits me to 'flop' the encoder inputs to ensure that spindle direction is accurate.

So... sounds like the 126 & 108 combo are on the menu, and the 105 is out. The home/limit switch convenience is worth it alone.  Order placed.

Now, if you're still willing to help with a non-PMDX issue: how do I reconcile the encoder PPR to VFD frequency limitation problem?  I know I can set the PPR in the VFD, but I assume the VFD can't even read the signals fast enough to do the PPR math (given the 36KHz or 2KHz limits).

Is there some 'counter' board in the universe that will drop the encoder pulses down to a frequency the VFD can accept?  Or... if I just hook the index (Z) signal to the "EA" input and have a 1 PPR encoder pulse would that still work?  The PMDX/ETH400/UCCNC would get the full 2048 PPR signal, but the VFD would not...

As far as using the "EA" terminal only (single phase), I'm confused by the VFD manual.  Page 3-115 illustrates three diagrams for using the "EB" terminal as a direction indicator, but I'm lost.  The first one appears to be a relay-sourced voltage input (assuming 24Vdc) where voltage in indicates FWD and 0Vdc indicates REV.  I might be able to do this using the 106/107 FWD/REV relay?

Thanks again,
Spumco

[Bob at PMDX]

I have a 2048ppr encoder, so I'll have to figure out something.  I plan to spin to motor up to 5kRPM, so that's something like 170KHz - too much for even the "EA" terminal.

WJ200 cannot handle anything even close to 170 KHz.  It can take 32KHz max on the higher speed of the 2 terminals.  And I'm not even sure the PMDX-108-Input will pass those pulses through.  170 KHz (actually, 170.667 KHz) give a period of 5.86 microseconds.  And presuming the waveform is a 50% duty cycle square wave, the "high" and "low" portions are half that: 2.93 microseconds.  We specify the PMDX-108-Input's "response time" at 5us max but we don't specify the minimum pulse width.  That signal *may* make it through the 108, or it may not, or it may make it but be greatly distorted.

Now, if you're still willing to help with a non-PMDX issue: how do I reconcile the encoder PPR to VFD frequency limitation problem?  I know I can set the PPR in the VFD, but I assume the VFD can't even read the signals fast enough to do the PPR math (given the 36KHz or 2KHz limits).

Get a different, lower resolution encoder.  Do you really need to resolve the spindle position down to 0.175 degrees?  Or 1/4 of that (0.044 degrees) if you use quadrature mode?

Is there some 'counter' board in the universe that will drop the encoder pulses down to a frequency the VFD can accept?  Or... if I just hook the index (Z) signal to the "EA" input and have a 1 PPR encoder pulse would that still work?  The PMDX/ETH400/UCCNC would get the full 2048 PPR signal, but the VFD would not...

The WJ200 does not support a 1-pulse-per-revolution encoder signal (see register P004, 32 to 1024  pulses per rev).

As far as using the "EA" terminal only (single phase), I'm confused by the VFD manual.  Page 3-115 illustrates three diagrams for using the "EB" terminal as a direction indicator, but I'm lost.  The first one appears to be a relay-sourced voltage input (assuming 24Vdc) where voltage in indicates FWD and 0Vdc indicates REV.  I might be able to do this using the 106/107 FWD/REV relay?

You *might* be able to.  In a quick scan of the manual I could not find if there is a separate "dir" input for controlled in the directions, as opposed to the EB/EA Pins to reporting current direction.  If there are two separate "DIR" inputs- one for commanding the direction and one for reporting the direction, then you would use the PMDX-107 solid-state relay outputs to control the "commanded direction" inputs signal.  *IF* there are two spindle direction signals (one for commanded dir and one for actual dir), you *may* be able to wire them in parallel.

Bob

[spumco]

Bugger.

I dont need resolution that fine, but the encoder came with the motor and I was hoping to use it.  Oh well, it's just money.

Any suggestions for an appropriate ppr to work with the VFD and the 108 board?

[Bob at PMDX]

100 line encoders should be fairly common.  At 5K RPM that would be 8333 Hz (120 microsecond period) on the "A" and "B" phase signals.  That is still too fast for the slower input on your VFD (2 KHz max), but you could run 1 line into the higher speed input.  And the PMDX-108-Input should handle that just fine.

Bob

[spumco]

Awesome.  Thanks so much for the help.

I'm sure I'll be back to pester you all when the 126 & 108 show up.

Regards,
Spumco