[Bob La Londe]

My mini lathe started out as an HF 7x8 (billed as a 7x10), and I converted it to a 7x14. I never felt the screw that came with the 7X14 bed mated up properly with the grip nut on the carriage, so the other day in frustration I started my ballscrew CNC conversion. I'ld been accumulating misc parts for it for a while. It will still be a while before I have it going, but I do have a larger lathe to use in the mean time.

Anyway, while I was playing and figuring (and cuttign a part) my speed controller quit working. Rather than waste a bunch of time playing with it I checked the obvious. Fuse, power, switches, etc, and jerked it off the machine. I had a part on the lathe, and no way to run the motor. Not even at full speed. I tried running AC straight to the motor and got a beautiful demonstration of the 60 cycle vibration of running a DC motor on an AC power source. LOL. I had already decided the speed controller was going to go bye bye since I would want one I could control with an external pulse source when I proceeded with my CNC conversion so I unsoldered the bridge rectifier from the controller and hooked it directly to the incoming AC and to the motor. The motor spun up to full speed nicely. I finished my part, and then went in search of a new speed controller.

I found one from a Chinese Hong Kong vendor pretty cheap. Claims to be pulse width modulated for speed control and will use either a 100K POT or an external pulse signal at 3-12V. It also appealed to me because the selection seemed to be by way of a jumper. This would easily allow me to switch between manual or machine control of spindle speed. Given that they were cheap, and that replacements might be slow I bought two of them a few weeks ago. They finally arrived last week.

As is often the case with low price Chinese electronics there was no documentation. By putting on a pair of high power reading glasses (2x) I was able to make out power in, and motor out. The POT was already connected to the board. I soldered some jumper leads to the bridge rectifier and hooked it up to my new speed controller. I hooked the motor up to the proper terminals, plugged it in, and turned up the POT. The motor ran beautifully backwards. Not unexpected.

As is often the case with DC motors I found that there was a point near the top of the control were it actually dropped off slightly in speed as I turned up the control. That is unexpected. I have found many DC motors that seem to run better in one direction than in the other. I was curious. The control had claimed in the advertisement to be either AC or DC input and non-polarity. I swapped hte input leads and hte motor still ran backwards. Cool. It truly is nonpolarity for input. I swapped the leads to the motor, and the motor ran perfectly in forward all the way to top speed at the full travel of the POT. When I have some time I may try it on straight AC in and see if its any better, but for now its working fine.

I left it spinning for half an hour at top speed no load and the motor did not get noticeably hotter than it ever did. I pulled the motor cover and I could put my hand on it. Then I let it run for another half hour at a fairly low speed. Maybe 2-300 RPM. I did not get out my optical TACH and check it. I'm not to worried about it though as those little motors don't have much power at that low of a speed. I just set it at the slowest speed that it seemed to run smoothly. Anyway, the motor did get noticeably warmer, but not enough to raise any alarm bells. Also not unexpected with a "cheap" DC motor.

At that point I had it working as well as before, although electrical was all exposed across my work bench in front of the machine. I wanted to "know" how it was supposed to be connected and exactly what the three pin jumper did so I emailed the vendor asking for documentation, and I proceeded to be a search. I found a picture of a similar speed controller with inputs and outputs labeled. The jumper and pot were different, but similar. Yesterday the vendor emailed me the exact same picture. I guess I'll have to try the jumper and see, but it does seem pretty straight forward. Its really only got two positions plus no jumper. The picture showed the motor and power source connections just as I had them hooked up, and it shows the other terminals as the external pulse source, and shows the polarity. Maybe in the next few days I'll take one of my Mach 3 setup computers and try doing external speed control to see if it works as well.

IF (note the big if) everything works as I think, I'll use a relay to reverse polarity to the motor so I can trigger it from MACH for forward or reverse, I'll put in a toggle switch so I can swap jumper positions from manual to machine speed control, and that should be about it. I still need to make bearing mounts for my ball screws, and do a lot of other work, but I thought some other mini lathers might appreciate knowing about a cheap substitute speed control that works.

Here is the link for the Ebay vendor where I bought them. Later when I get out to the shop I'll post a picture of my board(s) and the terminal connection picture I have.

New 5V 110V Max 10A DC Motor Speed Control PWM Mach3 Speed Control | eBay

[Al_The_Man]

That controller, although PWM is not very sophisticated, it may have been worth more to pick up a KB or Baldor DC drive off ebay, they are mostly SCR but the odd one is PWM.
There is probably no current limiting circuit or other features the KB has.
The other feature they have when reversing the motor is to reset the drive so that the motor goes through the accell properly, reversing a DC motor without making sure that the supply has come up from zero can damage the drive or motor if the RPM is high.
A normal DC motor should be the same in either direction, unless it has offset brushes for uni-directional use.
Al.

[Bob La Londe]

The thing that got me about this particular speed controller was AC or DC in, non-polarity in. Huge voltage range, and 10 amp current limit. Since the original speed control had a 6.3 amp fuse it seemed to fit all the criteria to run. I can't argue about the price either. This controller has a fuse holder on the board with a 10 amp fuse on it. I may pull that fuse and solder on some leads so I can use a chassis mount fuse holder to make the fuse more accessible once I box everything up neat. The spindle will get a simple pulse tach to feedback to Mach once I get everything going.

I have nothing against KB except the price. I have a KB VFD that runs the spindle motor cooling fan (two motors one case) on my Hurco mill.

Yeah, stop and run up from zero is a biggy I have thought about. I may run some type of centrifical switch and latching relay configuration so power can not be reapplied to the motor if it is still spinning. I may not since I will be the only one to use this machine anyway. I might just make it only reversible in manual mode. In CNC mode I probably won't have a lot of use for reverse anyway. Its not powerful enough for rigid tapping or die thread cutting. For regular thread cutting the tach keeps it in sync so no need for reverse. I have considered puttign a pneumatic or electromechanical disk brake on it for certain types of indexing operations, but then I would probably change everything and put a regular servo motor on the spindle instead using an encoder with an index pin for the tach input. Seems a bit much for a cheap mini lathe though.

Oh, yeah... I think the center pin of the three pin jumper is common so I should be able to just switch it from one outer pin to the other for manual/machine control, and put the safety cover switch in series on the common pin.

Anyway, here are some pictures of the quick and dirty test layout, and a copy of the info picture sent from the vendor.

[doorknob]

Although your post is primarily about the speed control replacement, it would also be interesting to see your progress on the CNC conversion, either here or in a different thread.

[Bob La Londe]

Although your post is primarily about the speed control replacement, it would also be interesting to see your progress on the CNC conversion, either here or in a different thread.

I think the conversion will go fairly fast except for money when I really get into the swing of it. The only thing I am really struggling with is how I want to configure the X axis screw. I have some new 12mm ballscrews, but there is not enough room on such a small machine to mount the ball nut in a conventional manner. I may make a new cross slide table for it with a T-slot, and if I do I'll probably make it with an extended assembley behind the machine with the nut and screw completely behind the machine. If not I'll have to use an acme setup with a custom cut acme nut and mount.

[BAMCNC.COM]

I have a similar lathe and speed controller. How old is your lathe?

[Bob La Londe]

Probably about 5 or 6 years old. In the first picture above the card on the right is the old controller, and the one on the left is the new controller I am using. So far in testing it seems to work fairly well.

[x69rzx]

I'm working on a 9x20 convertion of my own. I see there have been several nice ones through here, though some of the threads have fallen silent. I snagged a 2.25HP motor w/ controller from a treadmill. The original controller has a analog slider style resistor for motor speed control. The control unit has a sensor and a magnet that lives in the pulley to determine actual speed. I'm not finding any documentation on integrating the controller and the cnc software to be able to acurately set the spindle speed. Do you know of any documentation on this procedure?

* I should add I'm using EMC2(LinuxCNC). There is some documentation, but hardly what I would consider complete. I also have Mach3 available, but the EMC2 came with the CNC kit.

[Al_The_Man]

Find out what is the DC voltage across the outer ends of the slider, I ~12v then there are boards out there that convert PWM to the analogue required for the drive, If you are using something like Mach that used PWM.
The advantage of the Tach, is that it will maintain an rpm based on the position of the slider.
The calibration for the spindle command to rpm ratio is done in the controller configuration, depending on what control you are using?
Al.

[109jb]

Bob,

Did you ever have a chance to cut any parts after putting this controller on? If so, did the lathe seem to have the same power as previously?

Thanks,

John B.

[Bob La Londe]

Al,

1. This control has an input for and external pulse source. I primarily use mach, although I was thinking about using LinuxCNC for the lathe just to learn something different.

2. I am pretty sure using a tach input with Mach does NOT provide for closed loop RPM control. You need a controller that does that on its own.

John, I have cut parts on the lathe since I put on new control. I have done some hot rolled steel parts, and some brass parts. It seemed to have plenty of power. I've done both rough (as much as such a flimsy machine can really rough), and I have done threading with it. I have also used it to drive stock into a die. Small dies need no precutting at all. Larger dies I thread 90% with single poitn threading first. It seems to have plenty of power. At lower speeds I need to turn it up slightly when it starts a heavy cut, but at higher speeds I think it has more power than it did before.

[109jb]

Thanks Bob. I ordered one and will give it a try on my G0704 mill.

I use LinuxCNC now and was reading about spindle integration. It sounds like LinuxCNC can handle the soft start and ramp the motor from 0 RPM up to the set speed. I'll give it a try when the controller arrives.

[Bob La Londe]

Are you sure it will handle your motor? Its fused at 10 amps, but I like to be conservative. The motor on my little lathe is only rated at about 600 watts (~ 3/4 HP). Your G0704 is rated at 1 HP and I would be more inclined to believe its rating than the rating on my mini lathe.

Doesn't your G0704 already have a speed control? I thought there was a hack to externally speed control it with a 0-10V input board.

I only changed speed controls on the lathe because mine died.

[109jb]

Yep, the mill has a speed control on it right now. This is more of an experiment for me. As for power, if the controller specs are accurate it should be good for about 1100 watts. The G0704 is 750 watts so should be well within those specs. I plan to run some cutting tests with the stock drive, then with the Chinese drive and see what the comparative RPM drop is while in the cut. If the rpm drop is comparable then I'll figure the drive is ok. If so, I will continue to use the Chinese controller with pwm hooked to linuxcnc. There is a cnc4pc product that can provide an analog signal for the stock controller, but it is more than twice the price of the Chinese controller and I am always up for an experiment anyway. If it doesn't work, I used to have a mini lathe and sold it to a buddy. I talked to him about a month ago and the controller for the lathe crapped out, so if it won't work for me I will just give it to him to get the mini lathe working again.

[Bob La Londe]

Good. You are approaching it with the right attitude. The controller is cheap enough. Took about 10-12 days for mine to arrive. Yours might arrive faster now that the Christmas package rush is over.

[inciteman]

Hey Bob, thanks for the information! By the way have you seen the thread at Anyone used one of these motor speed controllers off Ebay??? - Page 2 - CNCzone.com-The Largest Machinist Community on the net!

He's using the same controller, but states the heatsink gets REALLY HOT VERY FAST at low RPM do you know if thats so? Do you know if this is an SCR type of speed controller?

I just ordered one, so I'll be working on the microcontroller part for now (Arduino).

Thanks again!

[Bob La Londe]

Looks like virtually the same controller. I just don't think they work all that well at really low (relative) speed. Even with the SPID (which is a superior controller (and more expensive)) I doubt you are going to get reliable control out of a PC690 below about 8000 RPM.

[HPB]

Hey Bob, thanks for the information! By the way have you seen the thread at Anyone used one of these motor speed controllers off Ebay??? - Page 2 - CNCzone.com-The Largest Machinist Community on the net!

He's using the same controller, but states the heatsink gets REALLY HOT VERY FAST at low RPM do you know if thats so? Do you know if this is an SCR type of speed controller?

Hi inciteman,

These speed controllers don't use SCR, it uses a high current Mosfet to
PWM the -V line to the motor.

From looking at the controller picture, the Mosfet is driven by the output of
the opto coupler, this is so to enable you to select an external PWM source
such as Mach3.

The primary source of PWM an NE555 or LM7555 CMOS version.

Reason heatsink is going "HOT VERY FAST" is the lack of suitable
Mosfet driver.

Looking at the controller picture, I can't see any in there!.

When you do a PWM with a Mosfet you have to charge the gate capacitor
very rapidly during the ON phase, and discharge it rapidly during the OFF
phase. This should bring the Mosfet to it's safe operating range.

As the case with these Chinese PWM controllers, the opto transistor must
be switching On the Mosfet fast, but due to a pull-down resistor at the gate,
the Gate Capacitor will not discharge fast enough, leaving you with the
situation where mosfet drain-source channel goes from RDS(on) to like a potentiometer then Off.

This transition will cause the Mosfet to go "HOT VERY FAST"!!

You can fix this by driving the Mosfet with a PMD3001D.

See the image.

Kind regards,
Hanspeter.

[inciteman]

Hello HPB, thank you for that explanation! I had no idea it needed a mosfet driver. I will put one on there; however, for driving an AC Inductive Load (my CNC router) I get the feeling using an SCR would be more efficient then using a Bridge Rectifier with a Mosfet like in this setup, what would you suggest?

[Bob La Londe]

Hello HPB, thank you for that explanation! I had no idea it needed a mosfet driver. I will put one on there; however, for driving an AC Inductive Load (my CNC router) I get the feeling using an SCR would be more efficient then using a Bridge Rectifier with a Mosfet like in this setup, what would you suggest?

Is your CNC router AC inductive? What spindle are you using?