Brushed DC spindle motor power supply

[Al_The_Man]

[I]
I can supply a circuit for a very simple PWM-to-DC convertor or for a better (more linear?) one.

Cheers
Roger

Yeah, I have one Schematic I believe is the Gecko version, but in mine there is no analogue needed at all.
Will have to make a few tests to see what transpires!

[RCaffin]

Not, mind you, that linearity really matters: the Mach3 option to 'Calibrate Spindle' can look after that.

Cheers
Roger

[upnorth]

I'm leaning towards ordering one of these to give it a try. I should not have the same issues you have. My acorn board puts out a 0-5 or a 0-10VDC signal. I also will have the spindle drive relays and drive in a separate box from the main control. I can get 2 or 3 of these for the price of a replacement KBWT-26.

[Al_The_Man]

Also See page 5 of this attachment.

[RCaffin]

An interesting and useful doc, but not without some errors. For instance:

Pulley Ratio (an accurate ratio is required for Mach3 to be able to calculate the correct PWM to
control the spindle motor closely).
This only applies if sensing the motor RPM rather than the spindle RPM. The distinction should be made.

SmoothStepper has a bug in that the G4 Dwell command is expressed in Seconds regardless of what you
set it to in Mach3 / General Config.
My understanding is that this bug may be in Mach3, not in the SS. That needs to be checked though: I might be wrong.

The use of a motor speed sensor when a V-belt drive is used is not recommended. Such leather V-belts do not have a fixed ratio. OK, it will be close, but that is all. A toothed belt is different.

an inexpensive laser digital optical tachomete
Of what accuracy I wonder? That would depend on the clock inside the tacho. It may be of course that the tacho had been calibrated against a good reference. I don't know.

Figure 16 shows a CD4011N package with 3 gates unused, That is perfectly OK. It shows the inputs to the unused gates taken to 0 V, and that too is normal. BUT it shows the outputs of those 3 gates also being taken to 0 V, and that is quite definitely NOT OK. It may not destroy the chip, but it will certainly stress it.

Cheers
Roger

[upnorth]

I just put in an order for one of these. Hopefully it last's longer than the kbwt-26.

[RCaffin]

I 'updated' the original doc into English, and attached a lot of my notes to it.
I tried to attach my version FYI, but 'invalid file'.
Email me or PM me or whatever and I will send it to you.

Cheers
Roger Caffin

[upnorth]

I 'updated' the original doc into English, and attached a lot of my notes to it.
I tried to attach my version FYI, but 'invalid file'.
Email me or PM me or whatever and I will send it to you.

Cheers
Roger Caffin

Thanks for the manual and notes. Mine arrived today. It did not have a manual with it so I'll use yours.

What method are you using to turn on the spindle? Are you using the inhibit terminals or switching the AC on and off? Any issues so far?

[RCaffin]

My spindle driver continues to work very smoothly with no issues. I have tweaked some of the internal parameters, which process also seems to work just fine. I have been running the machine most days.

Spindle turn on:

I most definitely do not cycle the AC power. There are two reasons for this.
* The most obvious is that cycling the AC power to any gear is probably the single most likely source of a failure. That applies anywhere, not just here. OK, maybe I am a bit old-fashioned in this.

* The second reason is that my system needs the driver turned on anyhow.. There is a secondary output from the driver for the field winding needed for some DC motors (not mine), and I use this to create a signal back into Mach3 to say that the unit is powered and can run. Trying to run a program with the spindle not spinning almost always results in a broken cutter, which is 'not good'.

Why might the spindle be not spinning one might ask - the answer is that there is a circuit breaker in the AC line which protects 'things' against disaster, but the breaker is a bit twitchy and has been known to flip at power-up. I could remove the breaker, but I choose to leave it there. Yes, I am a bit 'belt and braces' in the safety area.

There is an 'Enable' input on the driver, but I am not sure what it does internally. It IS used to flip from a Run state to a Program state, but it seemed to me better that I should leave that input for that function alone. Instead I do two other things.

* The first is that the M5 command turns off the PWM signal from Mach3 to the driver. This PWM signal is converted to a DC voltage through optical isolation so the control signal into the driver goes to zero volts, which seems to turn off the driver anyhow. The driver is reliable in this: it seems to have a small threshold on the input which has to be exceeded before it will start.

* The second thing is that the output from the driver goes through a Forward/Reverse power relay and then through an On/Off power relay. If neither M3 nor M4 have been asserted this relay is 'Off', which really means it is hanging a dump resistor across the motor winding. So turning the spindle off (M5) dumps the back-emf into this resistor and kills the rotation within a few revolutions, which makes it fairly safe.

Cheers
Roger

[upnorth]

Thanks for the fast and detailed answer. Now I'll be able to get the spindle going tomorrow morning. I think I will skip using the enable switch. Mine will be wired a little bit different from yours. I have the choice of a 0-5 or a 0-10 VDC control signal with my control board. You indicated it shuts off with low voltage and I can set it to output no less than 10% of full spindle speed that should work.

On my previous drive a KBWT-26 it was acceptable to control on/off by turning on and off the AC input to the drive. I decided to change that on a new install and hooked up the inhibit switch and the drive blew. I'm 90% sure this was caused by a stray wire.

Good idea to repurpose the field wires. I'll look at that if I have any issues with the spindle not being reliable.

I found one reference to the HQ-SXPWM drives that mentioned short brush life with this drive. I hope that is not the case. My lathe is an old made in Australia with weird parts that are rare or unavailable. The motor is from England and I have so far not found replacement brushes.

Do you ever actually run your lathe in reverse? My plan was to leave that unconnected for now. The way I had it set up before did have a Forward reverse relay and stopping was fast in reverse and a normal coast down when being driven forward.

Once again thanks for your answers. Saves me a lot of googling time. Nice that people on opposite sides of the world can discuss things like this.

[Al_The_Man]

On my previous drive a KBWT-26 it was acceptable to control on/off by turning on and off the AC input to the drive. .

How come you did not make use of the Inhibit terminals?

[Al_The_Man]

* The second thing is that the output from the driver goes through a Forward/Reverse power relay and then through an On/Off power relay. If neither M3 nor M4 have been asserted this relay is 'Off',
Cheers
Roger

Are these relays easily replaceable or are they PCT versions?

[upnorth]

How come you did not make use of the Inhibit terminals?

If there is ever an issue with an open wire in the inhibit circuit it would turn on the spindle. Disconnecting the AC input with a relay made the spindle 100% safe. I did try to use the inhibit switch in the rewire and the drive blew on the first try after that.

[RCaffin]

For upnorth:

I think the Inhibit switch terminals on the KBWT-26 are at -100 VDC. Could be a problem there! That fact was not all that clear from the tech data.
The DC control voltage into the KBWT-26 was also (I think) at -100 VDC. I thought the idea of having inputs at -100 VDC was a rather cheapskate idea which did not endear it to me.

For Al

The power relays I am using are VERY stock standard 2-pole change-over (DPDT) 10 A ones, with 24 VDC windings: JQX-13F, in sockets (PYF08A). AgCd02/AgSn02 contacts for high switching loads. Electrical operation life of 10^5 operations, maximum switching current 20 A, contact voltage rating of 150 VDC. Readily available with sockets on ebay for very reasonable prices (<US$3 with socket), may also be called LY2NJ. Mine come with little LEDs across the coil winding so you can see at a glance what state they are in. I recommend using the socket: if a relay burns out you just unplug/plug.
Finder 55 series (for example) would work as well I think. It is a very common spec.

Brush life: That remains to be seen. My Baldor industrial DC motors have large brushes and they do not seem to be wearing (yet). I suspect that a rough rectified AC voltage drive may be more harsh than a nice smooth SMPS drive on both the brushes and the bearings. Certainly I noticed a big improvement in noise when I converted to an SMPS.

Cheers
Roger

[Al_The_Man]

For upnorth:

The power relays I am using are VERY stock standard 2-pole change-over (DPDT) 10 A ones, with 24 VDC windings: JQX-13F, in sockets (PYF08A).
Finder 55 series (for example) would work as well I think. It is a very common spec.

Brush life: That remains to be seen. My Baldor industrial DC motors have large brushes and they do not seem to be wearing (yet). I suspect that a rough rectified AC voltage drive may be more harsh than a nice smooth SMPS drive on both the brushes and the bearings. Certainly I noticed a big improvement in noise when I converted to an SMPS.

Cheers
Roger

OK I was under the impression the unit came with the relays.
As to PS, I have always used either linear style or direct off the mains, as is done with the typical 3Hp TM versions, they use fairly large smoothing capacitors and with PWM control are Very quiet down to zero RPM's.

[upnorth]

Early results are not promising. I have some issues to solve. I do have a long voltage input cable because the drive is in a box that is separate from the control. Possibly this is the cause of the issues. It's about a metre long. High quality shielded twisted pair wire. Properly grounded at one end.

The issues. Motor cogs when idle. I can eliminate it by placing my hand on the wire. When I do that I can bring the display down to zero at idle then the motor stops cogging. I can also adjust the rpm by manipulating the position of the wire. With the drive connected I can't get up to full voltage on the acorn control board.

I'm going to shorten the control wire tomorrow and see what happens. I'm thinking it might be a good idea to set up a relay that grounds the positive control voltage input at the control when the drive is idle.

[upnorth]

Looks like I will not be able to use the drive. I was able to eliminate the cogging at idle. Shielded wire to the motor fixed that. I am unable to get full voltage out of the acorn when full rpm is selected. In addition it seemed like the spindle was always hunting. A voltmeter confirmed that the voltage was varying a little. I think maybe the voltage input of the drive needs a few more milliamps than the acorn puts out. I tried it using the 0-5 and 0-10 VDC inputs to the drive. I see that the acorn can also put out a PWM signal for the spindle speed control so I may give that a try tomorrow. If it does not work it's back to a KBWS or KBWT drive.

[RCaffin]

Helpful ideas to assist:

* Try using the pot they supplied to see whether the output is stable. If it is stable with the pot, then the control signal or cabling is the problem. These are solvable.

* Can you use a soldering iron for some simple electronics? I can supply the circuit I am using if you wish. Parts from any of the normal electronics suppliers. It will drive the unit properly.

Cheers
Roger

[upnorth]

I did already try with the pot because I had the same idea. It does not seem to be as smooth as it was with the KBWT drive. It does not seem smooth but it better than when driven by the acorn. I also have a small power supply that I can use for an input to the drive for testing purposes.

I'm pretty well equipped for doing electronics. I can build almost anything from a schematic. I'm no expert but have some basic test equipment. I have a couple of oscilloscopes but was not sure if it was ok to hook them up to a motor or not because I usually only use them on electronics. I have other test units also like frequency generators, counters breadboards etc.

I also have a KBIC drive that I think I'll wire back in as a quick test for smoothness.

[RCaffin]

I have attached a JPG of my circuit. Maybe it can help?

Note: it does have an Enable input, but this does not go to the Enable of the HQ-SXPWM driver.


Cheers
Roger