The 'almost-dead' servo driver;
+ inside the servo box
5 Bears Home Homebrew CNC bench mill

Why did my servo drive unit enjoy a near-death experience? Here's what happened... The goal was to wire and test one of my deluxe Parker motors (see bottom of the eBay page). I followed the guidelines, quadruple-checking the connections, both DB15 (encoder and hall) and the power connectors, and I am confident that the wiring was done correctly.

I then plugged the motor into an axis. The Flashcut servo drive box is powered via a big switch in the rear, but enabled with a pushbutton on front. When pressed, you can hear a relay click, and power is then applied to the motor windings. When this occurs, the motor freezes in place until motion is commanded. I was gratified to see the Parker motor "go static", with no evil smoke smell. So far so good.

I then ran a circular-interpolation g-code, which produces a distinctive motor motion. The Parker motor was going fine for about 10 seconds, indicating the encoding, hall, and commutation wiring was probably OK. At that point, everything "let go", meaning the whole system went "limp", no power to the windings. The motor was cool, no smoke. Just dead. The Flashcut system would power on, (leds illuminated) but the front enable button no longer produced the satisfying relay click. I wasn't sure if I had cooked the servo, or if the axis driver itself had fried.

To check, I hooked up the stock Flashcut motors, and found that all three axes from the driver had no output. No power to the windings, no motion. The Flashcut user's guide states (referring to the rear panel) "In here is a 6A. fuse... note that this should be sized for your total current draw, etc" Of course, there is a fuse on the back! And it must have been too small! Joy!

I popped the primary fuse out of the back of the unit. It was totally intact. And I should have realized that it would be, because the unit still powered up, and this fuse protects the 115VAC input line. If it had blown, the entire drive box would be completely dead... no led's.

Pain. Sadness. Wailing.

 

At this point, I actually hoped that all I had done was fry a single axis, and that perhaps replacement of that axis driver might restore everything. I decided to investigate further by opening the servo driver unit. The lid to the drive unit is nicely fitted, and took me a while to get it off. The cover is brushed aluminum, well-made and attractive.

This view shows the power-supply side. Note the big toroidal coil to generate the three-phases for the brushless motors. And to my absolute delight, there was another fuse here, a 10Amp job, right below the blue arrow. It was cooked.

I looked through my box for a 10 amp fuse, located only a 5, and decided to try it just to make Mr. relay close again, and make Mr. Kurt happy once more. When I tried the 5A job, it powered up monentarily but blew within a second or two. With NO motor attached. This indicated that possibly something on the board had shorted internally. There was only one way to find out, and that was to get a correct fuse.

I used the lack of a 10Amp fuse as an excuse to grab the kids and drive to Taco Bell, right next to the local Radio Shack. While they munched burritos, I looked for the fuses. Radio Shack did have the 10A jobs, and at home, one of them was quickly installed and tested (shown here).

AND IT WORKED. pant pant, damn that was scary. All the axes worked fine with the Flashcut servos. Feeling brave now, I repeated the test of the single Parker motor, with the same result. Dead intrnal fuse.

Now, here's what I'd like help with... why did the power supply fuse blow? I don't want to mess with the Parker motors anymore if they keep cooking fuses. From the Flashcut web site, and my own manual, this servo driver can power brushed or brushless servos with an output of 18 to 90V.

From the Parker web site, some links to this servo, model SM232BE-NFLN
Parker servo basics (PDF)
SM motor overview
And the most important page, the technical data. Look at the SM232B column.

This motor should be capable of handling this driver with ease. Any suggestions? Thanks guys, it'd be a shame not to put them to work.

While I had the lid off, I thought I'd photograph the guts. Here are the 3 servo driver units, which feed the connectors (blue strips) at the rear

Rear View:

Blue = DB15 (Hall and Encoder data)
Red = 24V brake output terminals (these will be used for the Z axis without a doubt.)
Green = Brushless/brushed motor power output
Yellow = DB25 IN from signal Generator
Violet = DB9 input for DCN servo autotuning utility

The bottom network jack is not connected internally in the box. Some basic observations of the unit itself - The construction is high-quality. The sheet metal guage is heavy enough to give the unit a solid feel. I like the panic button on the front, and it has already seen use. It also doesn't look like it would be tough to upgrade from 3 to 4 axes; all the necessary stuff is there except the amplifier itself. A nice touch... the cooling fan is filtered. How often have we opened an old PC or other unfiltered, fan-cooled box only to find tons of choking dust inside?

So far, in all the use of the PC, servo amp and the signal generator, there has not been a single instance of communication problems between the three, and that has been at 115 kbaud. Very reliable.

This picture is of the side of one of the servo-drive units. I am unaware of the brand. The label is printed Servo Drive FC-182D-2010. These drivers are made by Logosol, and are nice, capable units.

I highly recommend using a dedicated Windows XP PC with Flashcut. Get a dual serial port of high quality so you can simultaneously connect both the Flashcut 2.0 motion software and the servo-tuning utility, DCN. It is helpful to have both programs running together. Cleanse the PC of all unneeded software, especially crap that runs in the backgound like all messenger-type programs, etc. If you cannot connect the PC to the internet (I keep my FC PC separate from the 'net for a number of reasons), use one of the new key-fob sized flash disks which plug into a USB port for transfer of G-coding. I plan on using my P4 desktop for CAD/CAM, then transfer the G-code to the shop on the flash-disk, which is smaller than a Bic lighter.