05/30/06

>>

Recently I wrote an email to J.R. Kerr, who has the PicServo SC chip.This chip allows one to use step and direction signals to control a DC motor with encoder. Very similar to Rutex or Gecko, 29 bucks for the chip, build your own support circuits.Here is the text of that email, and his reply.

Hi, tech support:

I emailed a while back with a question about the Picservo SC, inquiring if there was any example circuits for, oh let's say an Hbridge with irf540's and 9540's. You steered me towards an example in your existing doc's but I don't want to use l293's or 18200's, I would like to use discretes with enough power in a brushed motor example to do something useful, such as my retrofit on my old but spotless moog/bridgeport. If you had better example circuits and less technical jargonese you might,just might, sell a bunch of these. You must realize by now that step/direction to servo is becoming big business, as per Rutex and Gecko, and even the peripheral companies are starting to think about it.

Here's the honest facts:

Most motion geeks are typically machinists or hobby machinists who are by and large SELF-TAUGHT in the field of electronics. And yet.....they usually manage to cobble together their boards into enclosures with transformers and rectifiers and plugs, etc. I have been over your documentation many times and I find it difficult to understand from a "make the darn thing" perspective. (I was a recipient of a Regents Scholarship from New York State, and the winner of the Bausch and Lomb Science Award. I am no Einstein but I seem to have no problem understanding most of the documentation from other sources.)

If I do a search on let's say Google with the terms Rutex or Gecko, I get many results, and there are entire forums and groups dedicated to researching, purchasing, and utilizing these products. If I do a search based on JR Kerr Picservo SC, I get exactly 5 results.

I found at the CNCZone forum one person asking if anyone knew anything about the PicServo SC, and no replies at that time.

You can, if you wish, get very angry at what may be perceived as criticism but wouldn't it be better if you took a good look at your documentation practices and said, "Wow, we could sell a bunch of these if the consumers that normally buy these sorts of things understood what we were saying!"

Here's my take on the difference: I bought a couple of Rutex 990's from Vlad and had no problem making a high power drive circuit with the documentation he supplied.(9 pages) The only reason I don't like them is that I am not a big fan of Lock Antiphase Drive. I prefer PWM/Direction.

On the other hand, I bought your PicServo/PicEnc combination a couple of years ago and they have never left their packaging because I didn't want to wade thru the pages and pages of documentation. (Would you like to have them back? Never used.)

I am going to publish this letter on my website, and, (with your permission, of course) any reply you care to make. If you do not reply,a simple note will be made to that effect. I really believe you could serve yourselves better than you do.

With all sincerity,

Paul Jones

www.cnconabudget.com

Here is his reply:

Paul,

I was traveling when I responded to your last e-mail, so I do not have a copy of what I recommended to you. However, I should have pointed you to the documentation for our PIC-SERVO 3PH board (KAE-T0V5-BD3PHV1) which includes a schematic using IR2104 MOSFET drivers driving TO-220 power MOSFETs. The design will work with either brushed motors or 3-phase brushless motors. You might also want to look at the application note for the IRAMS driver board, which uses a 10 amp, 90v driver module -- more than enough power for a Bridgeport retrofit.

Again, this design will drive either brushed or brushless motors.

Your larger point, that our documentation is not tailored to the hobbyist or machinist, is essentially correct. While our products can be used for CNC applications, and we do have some minimal support for it, CNC is a very small part of our business. Most of our customers are either using our controller boards for light automation (manufacturing processes, biotech, etc.), or they are electrical engineers designing our chips into higher volume products. The breadth of applications using our controllers (valve controllers, filament winders, computer vision systems, etc.) requires that the operation of our chips be described in fairly minute detail.

Of course, that does not preclude us from adding application notes directed more specifically to the CNC hobbyist. In that regard, we are mostly limited by the manpower to generate them. I will take your comments as a request for more specific documentation aimed at non-electrical engineers wishing to build CNC control systems.

I don't know if this addresses your concerns, but hopefully it gives you an idea of where we are coming from.

Regards,

Jeffrey Kerr

JEFFREY KERR, LLC

So you see, some people do listen! Perhaps a benefit from this will be access to servo drive chips and the ability to roll our own drivers.

06/04/06

Decided to start with what looks like the easiest assembly of a PicServo SC, a Max232, and an LMD18200 driver in PWM/Direction/Enable mode.

06/05/06

Ordered enough supplies to get started with the PicServo SC Board. I found the LMD18200 at American MicroSemiconductor, 3 pcs for $29.67. The PicServo SC chip itself was $30.00 from Jameco, and the MAX232 chip to talk to the PicServo was $2.19 from Jameco.

Here is the schematic clipped from the VAST PicServo SC documentation. It is in *PDF format.PicServo SC schematic

06/08/06-Everything got here except the LMD18200's which should be here by the 12th.

I made a couple of boards based on Alan Garfield's 18200 Handy board, but he hasn't gotten back to me yet on putting his artwork up here. Here's a picture of the boards, though. Alan's website is www.fromorbit.com , in case you want to check it out.

Now I have to make the board for the PicServo SC chip/Max232. I found a cable that I hope will work for communication with the setup program, NMCTEST. If it works I will put the cable connections here.

I finished the design for the board for the PicServo chip and the Max232 today so I made up a quickie and it came out OK. I got the LMD18200's today, so I may stuff the board this weekend.

Alan Garfield got back to me with permission to use his LMD18200 board on the website. Here is a ZIP with his files in *.PDF format:Files for 18200 board

I would like to thank Alan for his hard work and encourage everyone to visit his site at www.fromorbit.com

08/28/06

Bill Krause loaned me his homemade Picservo boards to try out, and the first thing I have to say is that he does topnotch work. The board is chemically etched, and everything is quite well laid out. He preserved the layout of the actual JR Kerr boards as far as physical locations, which makes it possible to use their documentation. (a real lifesaver.)

I tested them with the utility/setting program for Picservo, NmcTest, and other than a broken wire on one of the encoders from shipping,(easily repaired), the boards work perfectly. I have tested them in position, velocity, and PWM modes, they go where they are supposed to and stay there. There is some deadband drift but it seems to average out over several positions. At 2000 counts per revolution, the error is less than 1%. When I get the step/direction set up, I will know more about the possibilities.