Brushed DC spindle motor power supply
Have you noticed that there are many, many suppliers for VFDs and 3-phase motors, but precious few for DC motors?
Indeed.
So why the interest in brushed DC motors?
IF you get an industrial brushed DC motor, such as a Baldor, you will find it far more powerful and far more reliable than you ever thought. Mine is technically rated at 2 - 3 amps, but it cheerfully survived a slight programming error (ahem) while drilling 6 mm holes in 6Al4V titanium. I hit the Stop button when I saw the current meter on the spindle hitting 8 amps. Eh, no worry, that did not hurt the motor. It is big and heavy and has much thermal mass.
I have been using a PWM KBWT-26 from KB for it, but it died recently. The +15 V rail went down to +0.060 V. I emailed KB asking for help, but they were somewhat reluctant to supply any circuitry for debugging. I could buy another one, but I dislike vendors who sell quite old technology with no support.
Alternatives from other American vendors seemed to be very limited, bulky, expensive, and antique. OK, they have existing customers who just want more of the same.
So I searched on ebay, and found the HQ-SXPWM-X PWM driver.
Pick your input voltage (110 or 220, etc) and your output: mine was 0-200 V and 0-8 A off the 220 V version.
It's a small unit with multiple control inputs (0-5V, 0-10V, PWM, potentiometer) in a nice little screened case.
And it IS a PWM unit rather than a noisy SCR job. The SCR units are cheap and versatile, but they put lots of mains-frequency power on the motor, which then runs very noisily. Trashes the bearing in the long run.
Cost: US$60 + US$10 postage. Amazing!
Extra features they don't tell you about: internal 7-segment display showing current and voltage output, or error messages. Adjustable ramp up/down time, etc. Several other adjustments too.
The secret sauce is, I suspect, that it uses a tiny microprocessor as the controller, unlike the extensive analog components on other older devices.
I bought mine from nanma0910 on ebay, and he was quite helpful. Yes, there IS a manual, abeit Chinglish, for it. The manual is 'adequate', but to the point.
Cheers
Roger
Re: Brushed DC spindle motor power supply
So how about an actual review? How much have you used it? Any issues so far?
I have a similar issue to you. I have a 180 VDC motor in my hercus compulathe. I had a KBWT-26 as well that went up in sparks. I'm looking for an alternative. A lot of USA places will not ship to Canada (or make it expensive as in we can't tell you how much the shipping will be we just want your credit card number)
and the drives are far more expensive in Canada than in the usa if you want a KB drive.
Re: Brushed DC spindle motor power supply
I have tested it on the bench, and that was fine. I hope to put it into the CNC today or tomorrow and to test it properly. I WILL let you know how it goes.
I have frequently found USA retailers selling the same item as I can find in ebay, but for 5 - 10 times as much. I guess they have to cover their costs, but I don't see why I have to be the bunny.
I have also found that a lot of the traditional USA suppliers are sticking with technology which is 50 years out of date, while the leading edge now seems to be Asia.
Innovate or perish.
Cheers
Roger
Re: Brushed DC spindle motor power supply
I have been using the 2HP to 3HP DC Johnson Treadmill motors and their PWM MC2100 drives.
They are micro processor controlled and have ramp up.
If you need reversal, a relay on the output is required.
The earlier MC-60 version is not too bad, also has many safety features but is SCR bridge controlled technology, so not quite as good as the PWM version. There is a source of schematics available out there on the Net.
I made a small Microchip controller for the MC2100 as it needs a PWM control signal also.
P.S. When using a T.M. motor on machines tools, the flywheel should be removed!!
Re: Brushed DC spindle motor power supply
Try as I might, I cannot find any specs for the MC2100 or equivalent.
What voltage output and what current output, please!
Also, I saw references to needing an off-board transformer with some variants: is this so?
Prices on ebay seem to be all over the place, $100 to >$200.
Fwiiw: the HQ-SXPWM-X supply has adjustments for
ramp-up time,
ramp-down time,
maximum voltage limit,
maximum current limit,
motor internal resistance compensation,
separate Enable/disable input (5 V),
and multiple input control modes.
It sells for $60.
I will be testing it (hopefully) today in the CNC.
Cheers
Roger
1 Attachment(s)
Re: Brushed DC spindle motor power supply
There is quite a few reverse engineered schematics out there, there is also about five versions, all basically the same.
No transformer needed, although there is one that has the LV PS transformer, off board, but it is an exception.
Re: Brushed DC spindle motor power supply
Thank you for that.
That doc seems to be focused on how to make a variable PWM signal using 2 off 555 timers. Rather basic, and in the examples provided the control was via a manual pot. This does not really suit a CNC machine. Anyhow, Mach3 already provides a PWM signal for spindle control, so the whole thing is superfluous to my needs.
I do use a dual 555 circuit for the misting unit which provides MQL to my CNC, so nothing new there. Independently variable pulse frequency and variable pulse width.
Chuckle: I have been using the 555 since it was first marketed, albeit with some reluctance at times. The design has copped its share of abuse over the years, but it still works.
But I could not find info in the doc on the performance of the MC2100 itself: what voltage range output and what current range output. I am still interested in that information.
Cheers
Roger
Re: Brushed DC spindle motor power supply
A problem with the unit whose circuit diagram is shown in that doc: the rail voltage cannot exceed 200 VDC. Limits to the devices. You can't get that direct off the 240 V mains. Perhaps there are other versions?
Cheers
Roger
Re: Brushed DC spindle motor power supply
There are two versions of the MC2100 the N.A. version and the 230v version.
There is at least one T.M. not working give away in the local on-line market place here for a source of DC motors and, in some cases, have a MC2100.
These require the PWM drive signal, for which I built the 10mm square microchip boards, the output is 20Hz PWM.
Re: Brushed DC spindle motor power supply
the N.A. version and the 230v version.
Figures.
How much current can one handle?
Cheers
Roger
PS: treadmill??? We use the local Fire Trails!
Re: Brushed DC spindle motor power supply
Well the motors run from 19amps for the 2HP to around 28amps for the 3HP. This is the 120vac versions.
Actually you have given me an idea of taking the PWM controller boards I made up a bit further and attempt to modify for analogue control signal from a remote source, rather than the pot. :banana:
Remote Start/Stop inputs also, at present just PB inputs.
.
Re: Brushed DC spindle motor power supply
OK, that is grunt.
My preferred method for the control signal is a 2-op amp triangle wave generator followed by a comparator for the analog signal input.
One can use a digital lockout signal via a gate for the Enable, or one can just use a transistor to pull the comparator off to inhibit. Using digital gates means also generating +5 V as well as the higher rail for the op amps.
I definitely prefer optical isolation around the power stage: the nominal 0 V rails from the power stage are anything but.
Cheers
Roger
Re: Brushed DC spindle motor power supply
Quote:
Originally Posted by
RCaffin
I have tested it on the bench, and that was fine. I hope to put it into the CNC today or tomorrow and to test it properly. I WILL let you know how it goes.
I have frequently found USA retailers selling the same item as I can find in ebay, but for 5 - 10 times as much. I guess they have to cover their costs, but I don't see why I have to be the bunny.
I have also found that a lot of the traditional USA suppliers are sticking with technology which is 50 years out of date, while the leading edge now seems to be Asia.
Innovate or perish.
Cheers
Roger
Thanks looking forward to hearing how it works out for you. I have a Hercus CNC lathe and mill (Made in Australia) I would buy two of the drives if they work out ok.
I did have a quick look at the specs they do seem versatile.
Re: Brushed DC spindle motor power supply
Did not get as much done yesterday as I had hoped. Had to first disassemble some temporary stuff inside the machine, then mount the driver onto the control plate. That is now done.
I am, right now, making up the opto-isolator unit for the PWM. That should not take long.
Your CNC machine, made in Australia - brand? Adept?
Cheers
Roger
1 Attachment(s)
Re: Brushed DC spindle motor power supply
Attachment 472448
As promised, some feedback on the HQ-SXPWM-X spindle driver.
The unit is the mesh box on the left, with all I/O at the top.
Have you ever started a program running, only to realise the spindle
is not spinning? That way lies damaged jobs and broken cutters. I
needed a 'state sensor' to tell Mach3 when the spindle driver was
powered up - or not,as the case may be. The PCB at the top, half
out of the frame, is my custom state sensor. Its operation needs
explaining.
A lot of these DC drivers have an auxiliary output of 200 VDC
(often +100 and -100 V) for the Field winding - if this is used.
I use it for the state sensor. A chain of resistors and a zener
go across the two Field outputs. An opto-isolator goes across the
zener, with the output signal going into Mach3. If that signal is
asserted, the power supply is 'live'. If not, there is no power
and the M3/M4 macros abort with an error message.
To the right there is a white terminal strip. This supports a very
low resistance current sensor which feeds a front panel meter,
just for diagnostics.
Below that there is a black terminal strip supporting a black
slab. That is an opto-coupled Triac, which drives a small 240 VAC
water pump for spindle cooling, via the black power socket at
bottom right. This is driven by a signal from Mach3 in the M3/M4
macros.
In the middle there are 4 power relays. As my machine is a dual-
spindle unit, the bottom two energise either the Mill or the
Lathe (but NOT both at once!). There are lockouts in both the
M3/M4 macros and in the control electronics to prevent this.
The upper one labeled R is the reversing relay, driven by M4. Its
function is obvious. The upper one labeled O is the On/Off relay.
It is activated by either M3 or M4. This is of course optional,
but it allows me to dump the motor back-EMF when the spindle
drive is turned off by M5. The dump is into the two big orange
power resistors to the right of the M & L relays. They work well.
Yes, there are a lot of D-connectors across the bottom. It is
amazing how the number of wires in a CNC proliferate.
I have tested the unit, and it runs fine. I have yet to fine-tune
the output inductor (not shown) to match the driver. Also not
shown is the PWM signal and a little opto-isolator unit which
goes between the wires from Mach3 and the power driver. Both are
properly grounded (of course), but RFI is RFI. Yeah, I took the
photo too early.
Cheer
Roger
Re: Brushed DC spindle motor power supply
Quote:
Originally Posted by
RCaffin
Did not get as much done yesterday as I had hoped. Had to first disassemble some temporary stuff inside the machine, then mount the driver onto the control plate. That is now done.
I am, right now, making up the opto-isolator unit for the PWM. That should not take long.
Your CNC machine, made in Australia - brand? Adept?
Cheers
Roger
My CNC lathe and one of my mills are made by Hercus. I should point out that they are not new they are probably at least 30 years old. They came without controls so one is getting an Centroid Acorn. The other one has a Mach 3 based control but it will also eventually get an acorn. If you ever have a need for a small lathe or mill and run across one they are very well made.
Here is a link to an Australian site that has information about them.
lathesPC200 | benchtopcnc
Re: Brushed DC spindle motor power supply
Quote:
Originally Posted by
RCaffin
Thank you for that.
That doc seems to be focused on how to make a variable PWM signal using 2 off 555 timers. Rather basic, and in the examples provided the control was via a manual pot. Roger
I inadvertently uploaded the wrong file, the right one is now in the link. (post#6).
I am working on a method to control the MC2100 off of Mach3
Re: Brushed DC spindle motor power supply
Unexpected happenings with HQ-SXPWM-X unit. Essential reading.
The HQ-SXPWM-X seems to be controlled by a microprocessor. Changing the various parameters is done the same way you adjust a watch: with up & down buttons. This has certain unexpected and probably undesired consequences.
I had the system fired up and running yesterday, and the voltage on the spindle 'fluttered' at about 1 Hz. It took some thought to figure out why. Note that on the bench when driven by either the supplied pot or by my sig gen there was NO flutter. But explaining this takes us into some obscure internals of Mach3.
Mach3 has an internal engine/update speed. By default it is 25 kHz, but that can be changed by the user to over 100 kHz. It also has a user selected base frequency for the PWM. Default seems to be 20 Hz, but again this can be changed by the user. Most users would probably not touch this. But the manual for the HQ-SXPWM-X recommends a PWM frequency of 1 kHz. There are consequences. For reference, my system is set to 100 kHz engine and 1 kHz PWM.
Imagine you have an engine speed of 25 kHz and a PWM speed of 1 kHz. The pulse width can only be set to one of 25kHz/1 kHz values by the SW. For a top spindle speed of 3,000 RPM, this means you can only set the spindle speed in steps of 3000/25 = 120 Hz. But Mach3 allows you to specify the spindle speed to within 1 Hz. How does it do it?
I think it dithers, swapping between the two nearest possible PWM widths at a cycle rate of about 1 Hz. Over time this averages out to give the specified spindle speed. If you are using a PWM to DC convertor with a good smoothing capacitor on the output you would probably never notice it. But I think the HQ-SXPWM-X senses the incoming width of every single PWM pulse digitally, and tries to adjust according to the latest width. As this width flutters, so does the output voltage! And so does the output current as the driver tries to vary the spindle speed.
How to solve? Actually, dead easy. I had a pulse width to DC convertor on the input to the previous driver, with a bit of a filter on the output for smoothing. I will simply switch from the PWM input to the +5 V DC input using this existing convertor. I have already tested this on the bench and it works fine.
What this means underneath all this is that Mach3 was meant for analog-type spindle controls. The PWM signal was a way for a computer to create the required analog signal - slowly. PCs have their limits.
Cheers
Roger
Re: Brushed DC spindle motor power supply
Quote:
Originally Posted by
RCaffin
Unexpected happenings with HQ-SXPWM-X unit. Essential reading.
Roger
If what you are saying is the Mach3 PWM signal fluctuates, I assume you are saying in PWM frequency, not pulse width?
If so, that doesn't bode well for what I intend to do, use the PWM signal without conversion.
Anyone 'scope it to find out what it looks like I wonder?
May have to do that next.
Re: Brushed DC spindle motor power supply
I assume you are saying in PWM frequency, not pulse width?
No, I am saying that I believe that the pulse WIDTH fluctuates.
After all, the PWM frequency is set by the user on one of the config pages.
Mach3 can only generate a small number of pulse widths due to the maths involved. For instance, if you are using a 1 kHz PWM speed and a 100 kHz engine speed or clock frequency, you can have only so many pulse widths. You might for instance have 50 clock pulses or 51 clock pulses for the width, but you can not have 50.5. So Mach3 dithers to get the average.
Yeah, rather unexpected beforehand, but it makes perfect mathematical and software sense. Mach3 was designed for drive controllers which used an analog control signal, of low bandwidth. I can supply a circuit for a very simple PWM-to-DC convertor or for a better (more linear?) one.
Cheers
Roger